The Next Big Arenas of Competition Final

The next
big arenas of
competition
Arenas are industries that transform the business landscape.
Eighteen future arenas could reshape the global economy and
generate $29 trillion to $48 trillion in revenues by 2040.
Authors
Chris Bradley
Michael Chui
Kevin Russell
Kweilin Ellingrud
Michael Birshan
Suhayl Chettih
Editor
Max Berley
Data visualization
Chuck Burke
October 2024
Confidential and proprietary. Any use of
this material without specific permission of
McKinsey & Company is strictly prohibited.
Copyright © 2024 McKinsey & Company.

All rights reserved.
Cover image: Large stadium lights
© ZargonDesign/Getty Images
The next big arenas of competition ii

McKinsey Global Institute
The McKinsey Global Institute was established in 1990. Our mission is to provide a fact base to
aid decision making on the economic and business issues most critical to the world’s companies
and policy leaders. We benefit from the full range of McKinsey’s regional, sectoral, and functional
knowledge, skills, and expertise, but editorial direction and decisions are solely the responsibility
of MGI directors and partners.
Our research is currently grouped into five major themes:
— Productivity and prosperity: Creating and harnessing the world’s assets
most productively
— Resources of the world: Building, powering, and feeding the world sustainably
— Human potential: Maximizing and achieving the potential of human talent
— Global connections: Exploring how flows of goods, services, people, capital, and
ideas shape economies
— Technologies and markets of the future: Discussing the next big arenas of value
and competition
We aim for independent and fact-based research. None of our work is commissioned or funded by
any business, government, or other institution; we share our results publicly free of charge; and we
are entirely funded by the partners of McKinsey. While we engage multiple distinguished external
advisers to contribute to our work, the analyses presented in our publications are MGI’s alone, and
any errors are our own.
You can find out more about MGI and our research at www.mckinsey.com/mgi.
MGI Directors MGI Partners

Sven Smit (chair) Mekala Krishnan
Chris Bradley Anu Madgavkar
Kweilin Ellingrud Jan Mischke
Sylvain Johansson Jeongmin Seong
Nick Leung Tilman Tacke
Olivia White
The next big arenas of competition 1

Contents
At a glance 4
Introduction 6
Executive summary 8
Chapter one. The arenas of today 17

Chapter two. The arena-creation potion 37
Chapter three. The arenas of tomorrow 55
Arenas of tomorrow compendium 83

1. E-commerce 85
2. Artificial intelligence software and services 89
3. Cloud services 93
4. Electric vehicles 97
5. Digital advertising 106
6. Semiconductors 114
7. Shared autonomous vehicles 119
8. Space 125
9. Cybersecurity 132
10. Batteries 137
11. Modular construction 144
12. Streaming video 149
13. Video games 154
14. Robotics 161
15. Industrial and consumer biotechnology 167
16. Future air mobility 174
17. Drugs for obesity and related conditions 179
18. Nuclear fission power plants 185
Acknowledgments 191
Endnotes 192
Technical appendix 200

Title of publication 3
18 future arenas in detail
E-commerce EVs Shared AVs Batteries Video games Future air mobility
Arena-
AI Digital ads Space Modular construction Robotics Obesity drugs
Arenas creation Arenas of Technical
of today potion tomorrow Cloud Semiconductors Cybersecurity Streaming video Non-medical biotech Nuclear fission appendix
At a glance
— Arenas are a unique category of industries defined by two characteristics: high growth and
dynamism. They capture an outsize share of the economy’s growth, and the market shares of
players within them change to an outsize degree.
— We have identified 18 potential arenas of the future that could reshape the global economy,
generating $29 trillion to $48 trillion in revenues by 2040. These arenas range from AI
software and services to cybersecurity, from future air mobility to drugs for obesity and related
conditions, and from robotics to nonmedical biotechnology These future arenas could generate
$2 trillion to $6 trillion in profit by 2040. Their collective share of global GDP could increase from
4 percent to 10 to 16 percent by 2040.
— Twelve arenas of today showed outsize growth and dynamism from 2005 to 2020. These
industries include e-commerce, biopharma, electric vehicles, consumer internet, and cloud
services. They had a revenue compound annual growth rate (CAGR) of 10 percent and market
capitalization CAGR of 16 percent, and they tripled their global GDP share from 3 to 9 percent in
the period. By contrast, non-arenas had only a 4 percent revenue CAGR and a 6 percent market
cap CAGR over the same period.
— The many striking differences between the 12 arenas of today and non-arenas inform our
understanding of the arenas of the future. Arenas earn far greater profits than other industries
do, they spawn a disproportionate number of global giants, and they offer unusually strong
opportunities for new entrants to become powerhouses.
— Three combined ingredients in an “arena-creation potion” tend to generate the escalatory
mode of competition that characterizes arenas. The telltale elements of a forming arena
are business model or technological step changes, escalatory investments, and a large and/or
growing addressable market. The presence of these elements can lead to escalatory competition
among players, who make large investments to gain not only market share but also a product
quality edge, compounding the benefits and further setting them apart from other companies in a
race to the top.

Arena-
Introduction
This report from the McKinsey Global Institute identifies and describes a category of industries that
could account for much of the future change in the business landscape and transform the world. We
call these industries arenas of competition. To identify the arenas of tomorrow, we look back at the
arenas of today to see how they evolved. Arenas are defined by two characteristics: they capture an
outsize share of the economy’s growth, and market share within them changes hands to an outsize
degree. The presence of those two attributes indicates that a new competitive game has begun,
usually prompted by a new bundle of technologies and business models.
Understanding arenas is important for at least two reasons. First, they are where the business
world is reshaped. They account for many major shifts in investment, R&D, and value, as well as
the emergence of many new and growing global corporations. Second, once we start to recognize
the factors that could point to the potential formation of an arena, we can identify a set of arenas
that could plausibly emerge over the next 15 years. If the past is any guide, they will be centers of
competition, innovation, and value creation.
The report begins with a discussion of 12 arenas of competition that emerged between 2005 and
2020, including cloud services, e-commerce, biopharmaceuticals, and electric vehicles (EVs).
Chapter 1 identifies the arenas of today by drawing on a large, customized data set of the top
3,000 companies by market cap and explores how the arenas differ from other industries. We used
2005 to 2020 as our analytical interval to delineate a clean decade boundary and ensure consistent,
well-established data. Some of the differences are striking: in 2005, these arenas generated only
9 percent of our sample’s economic profit, but by 2019 they accounted for 49 percent of all economic
profit in the biggest companies.
In chapter 2, we examine how today’s arenas emerged and grew as companies entered a mode of
intense competition and made the escalatory investments characteristic of arenas. Understanding
this industrial logic is crucial because knowing how today’s arenas were born could help us
spot tomorrow’s potential arenas. We identified a “potion” that appears to underlie the emergence
of arenas.
In chapter 3, we describe 18 potential arenas of tomorrow and explore how they may materialize,
including their potential sources of growth and dynamism. Understanding potential arenas is
relevant for entrepreneurs and incumbent companies that want to compete directly in arenas, other
companies whose businesses would be affected by the emergence of arenas, investors looking
to allocate capital to these industries, people seeking jobs in the winning industries of the coming
decades, and policy makers looking to play a role in how and where these industries develop.
A compendium covering the 18 industries that could become the arenas of tomorrow includes further
descriptions of the growth factors and competitive dynamics that we present in the report and that
could shape each potential arena over the coming decades. To be sure, looking into the future is
always speculative, and we recognize the possibility—indeed, the likelihood—that we may be getting
some things wrong. For that reason, we have made transparent our assumptions, or what you need to
believe about each candidate arena to match our scenarios. This will allow readers to calibrate their
own views of which industries will end up becoming arenas.

Arena-
Executive summary
This report identifies a set of present and future arenas of competition, industries that could
transform the business landscape and our world. Arenas are defined by high growth and high
dynamism. These industries capture an outsize share of value growth, and market share within them
shifts dramatically, as measured by the “shuffle rate,” a metric of company market share movements
(Exhibit E1). These two characteristics signal a new era of competition and signify new technologies
and business models in play.
The arenas of today

We have identified 12 arenas of today: software, semiconductors, consumer internet, e-commerce,
consumer electronics, biopharmaceuticals, industrial electronics, payments, video and audio
entertainment, cloud services, electric vehicles (EVs), and information-enabled business services
(ranked in order of 2020 market cap). “Arenas of today” refers to the arenas that formed over the past
two decades. We used 2005 to 2020 as our analytical interval to set a clean decade boundary and
ensure consistent, well-established data. Understanding arenas is important for at least two reasons.
Not only are they where the business world is reshaped, but recognizing the elements that are usually
present in an arena and that help explain its growth and dynamism allows us to identify a set of arenas
that could plausibly emerge over the next 15 years. If the past is any guide, they will be centers of
competition, innovation, and value creation.
In 2005, arenas generated less than

10 percent of total global economic
profit. By 2019, they accounted
for half of the total.

Arena-
Exhibit E1
The 12 arenas of today exhibited outsize shuffle rates and significant growth
in share by market cap.
Market cap shuffle rates and growth share for 57 industries

Arenas of today Other industries Circle size = market cap in 2020 Market cap,
$ trillion
100
Electric vehicles, 0.9
Cloud services, 1.0
90
Payments, 1.6
80
E-commerce, 3.3
Shuffle rate
(increase Consumer
70
in market cap electronics, 2.5
share among Consumer
companies in internet, 3.5
60
each industry),
2005–20, Biopharma, 2.3
percentage Video and audio
points 50
entertainment, 1.5
Industrial
40 electronics, 2.0
Information-enabled
business services, 0.9
30
Software, 3.6
Semiconductors, 3.5
20
–8 –6 –4 –2 0 2 4
Industry share growth rate (change in share of global increase

in market cap), 2005–20, percentage points
Note: Based on McKinsey Industry Classification; Quality 4 & 5 data from McKinsey Value Intelligence, PitchBook only; subsidiaries excluded; includes only firms
with market cap >$3.5B in 2005 or >$5B in 2020; number of firms by arena varies; firms identified as did not exist/nonpublic in ’05 based on no McKinsey Value
Intelligence market cap data in 2005.
Source: McKinsey Value Intelligence; McKinsey Global Institute analysis
McKinsey & Company
Today’s arenas stand out from other industries in six ways.

— Arenas captured an increasing share of economic profit. In 2005, arenas generated $55 billion,
or 9 percent of total global economic profit, while other industries generated $549 billion,
about 90 percent of the total. By 2019, arenas were generating $250 billion—half of total global
economic profit. Comparing 2005 economic profit rankings with those from 2019 and 2020,
every arena except industrial electronics moved up.
— Arenas attracted outsize levels of investment for innovation. Arenas’ share of R&D investment
was already high in 2005 and remained high for 15 years. Sixty-two percent of US business R&D
spend went to arenas and arena-adjacent industries in 2005; that figure increased to 65 percent
by 2020. Semiconductors and electrical components accounted for the largest share, followed by
biopharmaceuticals and software.

Arena-
— Arenas enabled new entrants to grow. In 2020, 33 percent of arenas’ total market capitalization
was held by companies that had been “outsiders” in 2005—that did not exist, had market caps
of less than $3.5 billion, or existed but were not yet meaningful competitors in those arenas. In
comparison, businesses new to non-arena industries held just 15 percent of total market cap.
New players tended to enter during early stages of arena formation when competitors identified
innovations that met customer demands as targets for investment. This competition led to arenas’
characteristic dynamism.
— Arenas spawned giants. Arenas were more likely than non-arenas to give rise to the world’s
largest companies. In 2020, 74 percent of arenas’ total market cap was held by companies with
market caps greater than $50 billion, compared with 47 percent for other industries. Fifty percent
of arenas’ total market cap was held by companies with market caps greater than $200 billion,
compared with only 15 percent for other industries.
Large companies were also more likely to be in arenas. Of the companies with market caps above
$200 billion, more than half belonged to arenas, even though arenas represent only one-fifth
of the overall sample by number of companies. In 2005, just one of the top ten companies was
in a future arena—Microsoft, with a market cap of $278 billion. By 2020, eight of the ten were in
arenas, with market caps ranging from $511 billion to $1.7 trillion.
— Arenas tended to be more concentrated. Concentration was observed at certain times in
arenas; for example, the top ten players in five arenas (cloud services, consumer electronics,
consumer internet, EVs, and payments) accounted for at least 90 percent of 2020 arena
market cap and revenues in our sample. At the same time, competitive pressure to innovate
remained. Investments that improve products or take advantage of network effects can have
increasing returns, requiring leaders to continually innovate to retain prominence. Arenas’
industry structures are rarely static or stable in the long term. Escalatory investments and their
exceptional returns for arenas can inspire fierce competition in markets that already have high
levels of innovation, provoking step changes in technology and business, which can disrupt the
ranks of winners.
— Arenas were more global. On average, 50 percent of arena revenues were generated outside
companies’ home regions, compared with 42 percent for non-arena companies. Companies in
arenas were also much more likely to be multinationals. Sixty-eight percent of arena companies
derived more than 20 percent of their revenues from countries other than their own. By
contrast, about half of non-arena companies took in more than 20 percent of revenues from
other countries. The software arena is particularly global. Its four largest companies by 2020
revenues—Microsoft, IBM, Oracle, and SAP—generated almost 60 percent of their revenues
outside their home countries.

Arena-
The origins of arenas

To help us identify future potential arenas, we examined how today’s arenas originated. We observed
three elements that, when combined, were likely to result in high growth and high dynamism and to
generate an arena. The three ingredients, which we call an “arena-creation potion,” are business
model or technology step changes, escalation incentives for investments, and a large or growing
addressable market.
— Business model or technology step changes. Technology performance and adoption are
often modeled as idealized S-curves. When a technology capability undergoes a step change,
adoption starts off slowly, reaches an inflection point when it accelerates, then flattens out as
the technology reaches maturity. Although real-world technology capabilities and adoption
do not perfectly adhere to this S-curve, we observe technology step changes in our arenas,
for example the innovations in lithium-ion battery technology that enabled production of EVs
at scale. Business model step changes can also result from technology that shifts commercial
models (who pays for what and how) for products or services, thereby disrupting existing market
structures, as occurred with e-commerce and with video and audio entertainment (streaming).
— Escalatory incentives for investments. Companies that leverage investments to not only produce
more products but fundamentally change and improve their products can boost their competitive
position and rapidly gain market share. Such investments are features of specific types of
spending, such as marketing, R&D, and certain capital expenditures. As companies advance their
capabilities in this way, they tend to also improve long-term margins and see returns increase with
scale. As a result, competitors also have a strong incentive to invest, beginning an “arms race”
in which they iteratively invest to scale and scale to invest, causing a simultaneous escalation in
capabilities. This pattern accelerates growth and the kind of market share jumps that are typical
of arenas, and eventually can limit the ability for new entrants to enter the arena, unless a new
technology or business model step change again opens up the playing field.
— Large or growing addressable market. Companies tend to reach large or fast-growing markets
either by already playing in sizable markets where demand growth continues to outstrip the
rest of the economy or by displacing share of an existing large market with a superior product
or service. The companies in our arenas that played in fast-growing markets took advantage
of technology and business model step changes to accelerate value creation. These markets
typically already had revenue pools of more than $100 billion in 2005, and companies competing
in them made escalatory investments that improved efficiencies or broadened capabilities.
These arenas included biopharmaceuticals, industrial electronics, information-enabled business
services, consumer electronics, payments, semiconductors, software, and video and audio
entertainment. Together, they recorded 5 to 13 percent revenue compound annual growth rates
(CAGRs) from 2005 to 2020. Companies that displaced share of existing large markets achieved
rapid growth by launching novel categories of products or services, taking shares from existing
markets or unlocking latent demand. These companies’ industries include cloud services,
consumer internet, e-commerce, and EVs. Their revenue grew at a 13 to 33 percent CAGR from
2005 to 2020.
The three ingredients of the arena-creation potion produce an escalatory mode of competition,
which results in high growth and high dynamism. The continuous investments characteristic of
escalatory competition typically build heightened competitive capabilities and globally relevant scale.
Competition in these settings can be like a tournament with a huge prize to the winner, but it is not
a lifelong crown, because a new round of competitive escalation often begins just as the last one is
settling. This dynamic contrasts with more traditional modes of competition, which involve initial entry
costs and additional investments to increase production quantity that result in more localized and

Arena-
static industry structures. In escalatory competition, players must iteratively improve product quality
through investment and capability building to obtain or retain market share.
In addition, the arenas of today generally benefited from the overall trend toward digitization. This
last era of digitization was a fertile place for big arenas to bloom. Global internet connectivity allowed
e-commerce players to connect buyers and sellers across the globe, providing the opportunity for
customers to purchase goods anytime, anywhere. Digitization also transformed video and audio
entertainment—and streaming video in particular—by shifting media consumption from movie
theaters and CDs to homes and mobile devices.
Continual exponential improvements in the cost of processing and communicating information
created a long S-curve. Digitization enabled global markets and scaled distribution platforms.
Software-driven business models with low variable costs and strong network effects created a
powerful battleground for escalatory investment. It is no surprise that most of today’s arenas are
rooted firmly in the digital economy.
The arenas of tomorrow

With these insights on existing arenas and their characteristics as a guide, we have identified 18
potential future arenas that together could yield $29 trillion to $48 trillion in revenues and $2 trillion
to $6 trillion in profits by 2040 (Exhibit E2). In terms of impact on the economy, we estimate that
they could grow from about 4 percent of GDP in 2022 to 10 to 16 percent by 2040. This translates
to a 18 to 34 percent share of total GDP growth. This shift in GDP share is a hallmark of arenas: if we
perform a similar analysis on our arenas of today, we find that the sample companies increased their
equivalent share of GDP from 3 percent in 2005 to 9 percent in 2020.
Some arenas of today will

continue as arenas of tomorrow.
Others will no longer qualify
for arena status.

Arena-
Exhibit E2
The 18 potential arenas of tomorrow could generate $29 trillion to

$48 trillion in revenues and $2 trillion to $6 trillion in profits.
18 potential arenas Revenue, 2040 estimate (CAGR, Profit, 2040

of tomorrow, by 2040 2022 2022–40, %) estimate,¹ $ billion
revenue estimate, $ billion (profit margin, %)
14,000–
E-commerce 4,000 20,000 280–
(7–9) 1,000
(2–5)
1,500–4,600 230–920
AI software and services 85
(17–25) (15–20)
1,600–3,400 160–510
Cloud services 220
(12–17) (10–15)
2,500–3,200 100–320
Electric vehicles 450
(10–12) (4–10)
2,100–2,900 320–580
Digital advertisements 520
(8–10) (15–20)
1,700–2,400 340–600
Semiconductors 630
(6–8) (20–25)
20–460
Shared autonomous vehicles n/a 610–2,300
(4–20)
Space 300 960–1,600 (7–10) 50–160 (5–10)
Cybersecurity 160 590–1,200 (8–12) 90–240 (15–20)
Batteries 98 810–1,100 (12–14) 40–110 (5–10)
Modular construction 180 540–1,100 (6–10) 20–220 (4–20)
Streaming video 160 510–1,000 (6–11) 50–150 (10–15)
Video games 230 550–910 (5–8) 80–180 (15–20)
Robotics 21 190–910 (13–23) 20–180 (10–20)
Industrial and consumer biotech 140 340–900 (5–11) 10–270 (4–30)

Future air mobility n/a 75–340 10–70 (10–20)
Drugs for obesity and related conditions 24 120–280 (9–15) 30–100 (25–35)
Nuclear fission power plants 18 65–150 (7–13) 5–50 (5–30)
Total 7,250+ 29,000–48,000 (8–11) 1,900–6,100
¹Defined as net operating profit less adjusted taxes (NOPLAT). NOPLAT share based on most closely mappable industries from our database of 3,000 companies
analyzed in chapters 1 and 2.
Source: Company annual reports; McKinsey Value Intelligence; McKinsey Global Institute analysis
McKinsey & Company

Arena-
The potential future arenas can be divided into three groups: arenas of today that are likely to
continue developing into arenas of tomorrow; subsegments of current arenas that may grow
sufficiently large and fast to become spin-off arenas; and emergent arenas that are not as closely
linked as the potential spin-offs to any of today’s arenas. While at different stages of their evolution,
each displays early signs of the three arena-creation potion elements.
Continuing arenas
1. E-commerce. Companies that sell goods through digital channels and fulfill them directly
2. Electric vehicles. Manufacturers of battery, plug-in hybrid, and fuel-cell EVs
3. Cloud services. Companies that deliver on-demand cloud infrastructure and platforms as
a service
4. Semiconductors. Designers and manufacturers of semiconductors, microchips, and integrated
circuits as well as providers of tools for semiconductor manufacturing
Spin-off arenas
5. AI software and services (spin-off from software). Companies that provide software and
services incorporating AI, excluding the hardware necessary to operate AI
6. Digital ads (spin-off from consumer internet). Platforms that enable advertisers to reach
consumers digitally
7. Streaming (spin-off from video and audio entertainment). Providers of on-demand video
entertainment over the internet
Emergent arenas
8. Shared autonomous vehicles. Operators of shared autonomous vehicle services
9. Space. Providers of space-related infrastructure and services to the commercial and state-
sponsored segments
10. Cybersecurity. Companies that provide protection for computer systems from unintended
and unauthorized access, modification, or destruction
11. Batteries. Manufacturers of rechargeable batteries used for EVs and other technologies
that are mostly linked to the energy transition
12. Video games. Producers and distributors of games played on dedicated consoles, PCs, and
mobile phones
13. Robotics. Manufacturers of robots and providers of robotics solutions
14. Industrial and consumer biotechnology. Providers of biotechnology-enabled products in markets
like agriculture, alternative proteins, biomaterials and biochemicals, and consumer products
15. Modular construction. Companies that operate in the modular construction value chain, from
design to assembly, with volumetric modules
16. Nuclear fission power plants. Players that construct nuclear fission power-generation facilities
17. Future air mobility. Operators of air mobility transportation services, such as electric vertical
takeoff and landing vehicles (eVTOLs) and delivery drones
18. Drugs for obesity and related conditions. Companies that sell glucagon-like peptide-1 (GLP-1)
drugs and other treatments for obesity and related conditions, such as diabetes

Arena-
Five arenas of today—biopharmaceuticals, consumer electronics, information-enabled business

services, industrial electronics, and payments—could lose arena status. They are unlikely to maintain
the scale of growth and dynamism that would propel them into being arenas of tomorrow.
We also examine almost-emergent arenas, industries with some elements that could make them
arenas but that also have uncertain growth or dynamism prospects and a relatively lower probability
of evolving into arenas. We include these because predicting the development of arenas is not an
exact science, and we recognize that we could be wrong. These potential arenas could represent
significant opportunities. The almost-emergent arenas include clean hydrogen, lower-carbon
materials, products and services for older adults, nuclear fusion, renewables generation equipment
and infrastructure, sustainable fuels, virtual reality and augmented reality, and Web3 (including
decentralized finance).
Our analyses of present and future arenas reveal three key swing factors that go beyond
uncertainties in the modeling and could be fundamental to the evolution of the arenas of tomorrow
in technology, investment patterns, and sources of demand. These factors are developments in
geopolitics affecting regulation of innovation and technological regionalization, advances and
adoption of AI technology in a range of industries, and the pace of the green transition, which aims to
alter the course of climate change and could drive demand in various parts of the market.
The 18 arenas of tomorrow we have identified could be even more materially transformative than the
12 arenas of today, shaping how we consume and process data, approach health and wellness, and
interact and communicate with one another. They could introduce new options for our lives as well as
new questions about our social progress, from the morality and ethics underpinning data and privacy
to imperatives for businesses to be inclusive and sustainable. Recognizing how and when arenas
originate, understanding how they evolve, and anticipating the way they could change society can
offer a unique view of the arc of society’s progress.
As we look forward, this report provides an initial view of where to expect the most growth and
dynamism and how to update that view as the future takes shape. At the end of the report, we include
a compendium that sketches the quantitative possibilities for the range of growth and the dynamism
prospects of each of the 18 potential arenas of tomorrow.

Arena-
CHAPTER ONE
The arenas of today
In 2005, there was no iPhone or App Store, no one had done a cloud migration, and a mass-market
EV would have sounded like science fiction. Today, these technologies are commonplace, part
of a wave of innovation that radically changed businesses, the economy, and the way we live. The
economic growth from these innovations was concentrated in a relatively small number of industries.
The numbers tell the tale. Of the world’s top 20 companies by market cap at the end of 2005, only
four remain in the top 20 as of September 2024. The combined market cap of the top 20 grew more
than sixfold from around $4 trillion at the end of 2005 to $25 trillion by September 2024. Of the
new entries, almost all were in a handful of industries we have designated as arenas. This picture
looks remarkably similar for the interval of 2005 to 2020, the period we study in detail. Of the
world’s 100 largest companies by market cap in 2005, only 46 remained on the list in 2020. Of the
54 newcomer companies, almost half were in a handful of industries we have designated as arenas,
lifting the total number of top 100 companies in arenas from ten in 2005 to 33 in 2020. Arenas are
defined by two characteristics: their revenues and market cap increase at a far faster pace than the
rest of the economy, and shares of revenue and market cap within them shift substantially over time.
We identified 12 arenas that formed during the 2005–20 period, ranging from cloud services to
e-commerce to consumer internet to biopharmaceuticals, that have reshaped the business world.
Arenas also transformed our everyday lives, including the way we interact with one another, access
information, complete tasks, entertain ourselves, and receive medical treatment. Take, for example,
the simple experience of watching a movie with friends. In 2005, the process was straightforward:
check schedules in the newspaper, call your friends, drive to the nearest theater or perhaps to the
local video store in hopes that your DVD of choice is in stock. Today, a moviegoer would use their
phone to check schedules online, decide if they want to buy a ticket in advance, message friends in
a group chat, and proceed to the theater, perhaps in an EV hailed ride. Alternatively, you can decide
to host a watch party, hop on a Zoom call, and stream a movie on Netflix. That activity alone involves
a multitude of arenas, including semiconductors, software, and consumer electronics that enable
mobile phone features; e-commerce, payments, and consumer internet, which allow the purchase of
tickets online or buying streaming subscriptions; and EVs, industrial electronics, cloud services, and
video and audio entertainment, which innovated ways of enjoying a service.
In this chapter, we describe the 12 arenas of today and explore how they differ from other industries.
Arenas are exceptional industries

For the purposes of this report, we have defined industries as groups of companies that compete for
the same addressable market.1 Because these companies sell similar products and target the same
customers, their strategies inevitably overlap, making them both competitive and interdependent.
Companies can be part of more than one industry when they sell products and services in more than
one market.

Arena-
Arenas are a unique set of industries in which the competitive tussle is heightened, resulting in the
two characteristics we have identified: extraordinary growth and exceptional dynamism or shifts in
shares. Players in arenas tend to innovate in technology and business models, invest at escalating
levels in competitive capabilities, and expand possible addressable markets. Arenas show far greater
levels of growth and dynamism than their non-arena peers because of these characteristics.
Industries can be analyzed on many levels. To identify potential arenas, we selected definitions
based on supply and demand considerations. The supply perspective required some similarity in
the technology of competing companies, as arenas are often launched through technology step
changes. The demand perspective ensured that the products and services were sufficiently similar
or substitutable and could define an addressable market. Together, these two criteria allowed us
to define potential arenas at comprehensible levels of granularity (but we also recognize that these
criteria are subjective, and alternative delineations are possible). In some cases, arenas were defined
as a fast-growing portion of an existing industry. To name one such case, the automotive industry
can be partitioned between internal combustion engine (ICE) vehicles and EVs. A similar pattern
applies to other instances where a new arena forms adjacent to and draws demand from a large
existing industry, as is the case with e-commerce and traditional retail, and biopharmaceuticals and
traditional pharmaceuticals.2
The arenas of today were at different stages of development on different timelines and followed
different trajectories in the 2005–20 time frame. It’s easy to imagine the differing trajectories of the
arenas of previous eras: oil and gas and mining, driven by rapid expansion of energy and materials
needs; automobiles and commercial air travel, fueled by demand from a rising global middle class;
and personal computers, powered by fast-growing new technologies, to name a few.
We suggest, however, that today’s business landscape is influenced by the overlapping emergence
of many big arenas. The common cause is the exponential force of digitization that has enabled both
novel business models and rapid technological improvement on a global scale; we discuss these
factors of critical importance later in this report.
Now we turn to the arenas of today to determine how they differ from other industries.
Today’s 12 arenas grew quickly and were dynamic

To identify today’s arenas, we relied on a McKinsey data set covering the world’s 3,000 largest
companies. For the purposes of our analysis, we reclassified some of them or parts of them
from broad industry groups (such as retail) into more specific competitive categories (such as
e-commerce). In some cases, this meant dividing a very large company (for example, Microsoft or
Amazon) into two or three pieces, each in a different competitive category.
We then identified industries that exhibited outsize growth and market dynamism. Specifically, these
industries grew quickly from 2005 to 2020, dramatically outpacing overall economic growth. That left
them with a larger share of the world’s revenues and market cap in 2020 than in 2005, a difference
we call the “industry share growth rate.” This metric captures the disproportionate value created in
those industries. In addition, their dynamism, measured by company market share movements—a
metric we call the “shuffle rate”—was relatively high during the period. In other words, there was more
shifting of market share within the industries we selected to be arenas than in other industries (see
sidebar “Industry dynamism measured by share shifts” and the technical appendix).

Arena-
Sidebar: Industry dynamism Consider some of the ten biggest players exercise for the entire market, we end up with
measured by share shifts in the mobile and consumer electronics market share percentage-point changes
industry. In 2005, Sony held 27 percent for each player. Adding up just the positive
To measure industry dynamism, we of market share by revenue in this group, percentage-point changes gives us the
considered its degree of market share but by 2020, its share had declined to industry’s market shuffle rate as measured by
shifting by seeing how much either 13 percent, a 14-percentage-point drop. revenues: 53 percentage points (Exhibit A).
revenues or market cap changed hands Apple, by contrast, had a 4 percent market We applied the same process for both
among companies. share by revenues in 2005. That had revenues and market cap to all companies,
grown to 30 percent by 2020, an increase not just the top ten, in each industry in our
of 26 percentage points. Continuing this analyzed data set.
Exhibit A The top ten companies in the consumer electronics industry had a 53-point
shuffle rate.
Company ranking by market share of revenue
Change in ranking Revenue share, % Positive shifts,

2005 2020 2005 2020 2005–20
1 1 Apple 4 30 +26
2 2 Huawei 2 21 +19
3 3 Samsung 14 14
4 4 Sony 27 13
53
5 5 LG 17 6 Percentage-
point sum =
6 6 Xiaomi 0 6 +6 shuffle rate
7 7 Microsoft 11 4
8 8 Sharp 10 3
9 9 Lenovo 0 1 +1
10 10 Shenzhen Transsion 0 1 +1
Outside
top 10
McKinsey & Company

Arena-
Sidebar: Industry dynamism which had notably less dynamism. Following industries that had no revenues or market
measured by share shifts the same approach, that industry had a cap at the start of a period would necessarily
16 percentage-point shuffle rate, much have an aggregate share shift of 100 percent,
(continued)
lower than the shuffle rate of the mobile and as is the case with the nascent EV industry.
consumer electronics industry (Exhibit B). Because high levels of dynamism were a
Now consider the top ten players in a criterion for qualifying as an arena, any
The shuffle rate indicates the degree of
different industry, aerospace and defense, nascent industry made the cut.
dynamism in an industry. We assumed that
Exhibit B The top ten companies in the aerospace and defense industry had a 16-point
shuffle rate.
Company ranking by market share of revenue
Change in ranking Revenue share, % Positive shifts,

2005 2020 2005 2020 2005–20
1 1 Lockheed Martin 13 16 +3
2 2 Airbus 14 15 +1
3 3 Boeing 18 14
4 4 Raytheon 14 14
16
5 5 General Dynamics 7 9 +2 Percentage-
point sum =
6 6 Northrop Grumman 10 9 shuffle rate
7 7 Honeywell 9 8
8 8 BAE Systems 6 6
9 9 Thales 0 5 +5
10 10 Safran 0 5 +5
Outside
top 10
McKinsey & Company

Arena-
Our method for measuring industry-share growth rates and shuffle rates yielded rankings of
industries by the extent of their growth and dynamism between 2005 and 2020—one set as
measured by market cap (Exhibit 1) and another set as measured by revenues (Exhibit 2). We used a
ranking system that considered both measures to settle on a single list of 12 arenas that placed high
in combined ranking scores and had a 2020 market cap of at least $800 billion (for more information,
see the technical appendix).
Exhibit 1
in share by market cap.
Market cap shuffle rates and growth share for 57 industries

Arenas of today Other industries Circle size = market cap in 2020 Market cap,
$ trillion
100
Electric vehicles, 0.9
Cloud services, 1.0
90
Payments, 1.6
80
E-commerce, 3.3
Shuffle rate
(increase Consumer
70
in market cap electronics, 2.5
share among Consumer
companies in internet, 3.5
60
each industry),
2005–20, Biopharma, 2.3
percentage Video and audio
points 50
entertainment, 1.5
Industrial
40 electronics, 2.0
Information-enabled
business services, 0.9
30
Software, 3.6
Semiconductors, 3.5
20
–8 –6 –4 –2 0 2 4

in market cap), 2005–20, percentage points
McKinsey & Company

Arena-
Exhibit 2
in share by revenue.
Revenue shuffle rates and growth share for 57 industries

Arenas of today Other industries Circle size = revenue in 2020 Revenue,
$ billion
90
Cloud services, 109
Consumer
80
internet, 403
Consumer
70 electronics, 648
Shuffle rate E-commerce, 888

(increase 60
in revenue Software, 341
share among
Electric vehicles, 101
companies in
50
each industry), Payments, 144
2005–20,
percentage Biopharma, 343
points 40
Video and audio
entertainment, 407
30
Industrial
electronics, 987
20 Information-enabled
business services, 154
10 Semiconductors, 574
–6 –5 –4 –3 –2 –1 0 1 2 3

in revenue), 2005–20, percentage points
McKinsey & Company
Our arenas displayed outsize growth compared with other industries from 2005 to 2020. Arena
revenues grew 10 percent year-on-year in that period, while revenues in other industries grew just
4 percent. The 12 arenas collectively more than doubled their share of revenue (from 6 percent in
2005 to 14 percent in 2020) and nearly tripled their share of our data set’s total market cap (from 12
percent in 2005 to 34 percent in 2020). The biggest industry share shifts were in categories created
by the new tech giants: cloud services, consumer internet, and e-commerce, driven by the growth
of leading technology companies like Alphabet (Google), Amazon, Meta (Facebook), and Microsoft.
For instance, e-commerce, which had the largest growth rate for both revenues and market cap,
accounted for only $15 billion of revenues in 2005, or 0.1 percent of revenues across all industries.
By 2020, this number had grown to $890 billion, or 2.4 percent of revenues across all industries. In
market cap, our e-commerce sample was an $87 billion industry in 2005, or 0.35 percent of market

Arena-
cap across all industries. That number grew to $3.3 trillion, or 4.2 percent of all industries, by 2020.
Alibaba, Amazon, and JD.com contributed $547 billion, or 63 percent, of this revenue increase and
$2.1 trillion, or 64 percent, of the market cap increase.3 Similarly, consumer internet represented
$215 billion in 2005, or 0.87 percent, of total market cap, which grew to $3.5 trillion in 2020, or
4.4 percent of market cap, in our data set. Alphabet, Meta, and Tencent were the largest players.
Cloud services companies represented $54 billion in 2005, or 0.2 percent, of total market cap, which
grew to $1 trillion and 1.3 percent of market cap by 2020, driven by growth of the cloud-services
businesses of Amazon Web Services (AWS) and Microsoft.
These arenas similarly demonstrated exceptional competitive dynamism, measured by the shifts of
revenue and market cap shares within an industry, or the shuffle rate. Average revenue shuffle rates
for arenas were 49 percent, compared with 32 percent for other industries. Average market cap
shuffle rates were 66 percent for arenas, compared with 47 percent for other industries.
Here we can see new industries transformed, as was the case with mobile and consumer electronics,
in which the arrival of smartphones contributed to a large shift in the competitive landscape for
players. Total industry revenues of $238 billion in 2005 were carved up as follows: Sony had the
highest revenues, $64 billion (27 percent of the total), followed by LG with $41 billion (17 percent) and
Samsung with $32 billion (14 percent). By 2020, total revenues had grown to $648 billion but had
reshuffled, with smartphone players leading the way: Apple had revenues of $192 billion (30 percent
of the total), and only Samsung continued to hold a top-three spot, with revenues of $88 billion
(14 percent of the total). Sony and LG, which previously had the most revenues, now occupied the
fourth and fifth spots, respectively.4 Nokia occupied the fifth spot in 2005 with 10 percent of revenue
market share, but by 2020, it held only a 0.3 percent revenue market share in this segment. There is
also dynamism happening in these industries beyond just their players in 2005 and 2020. The iconic
Blackberry brand was still relatively small in 2005, with roughly $2 billion in revenues. This grew
quickly to almost $20 billion by 2010, but then fell again to under $2 billion by 2015.
A similar analysis yields the same results for market cap, with high shuffle rates between players.
Together, these shifts put consumer electronics high on the reshuffling axis. Cloud services,
e-commerce, and EVs also experienced this internal reshuffling, to a large extent due to the large
market share movements of the players mentioned above (Exhibit 3).
The share shifts in market capitalization are also more volatile compared with revenues. For example,
the semiconductors arena had the lowest market cap shuffle rate among arenas at 41 percent
through 2020. By June 2024, its shuffle rate had more than doubled to 87 percent, driven by Nvidia
market cap growth, effectively giving semiconductors the third-highest market cap shuffle rate
among arenas. This dynamism, even in a short time, is typical of an arena. As we explain in chapter 3,
semiconductors may continue as a future arena, driven by technological step changes such as the
advances of artificial intelligence that can enable rapid reordering of companies in the race to the top.

Arena-
Exhibit 3
Arenas exhibited exceptional growth in value creation and scale.
12 arenas of today, by 2020 market cap

Companies, # Market cap, $ billion Revenue, $ billion
2005 2020 2005 2020
Software 110 259 3,636 64 341
Semiconductors 79 642 3,495 194 574
Consumer internet 31 215 3,460 24 403
E-commerce 73 87 3,308 15 888
Consumer electronics 16 362 2,502 238 648
Biopharma 91 416 2,289 57 343
Industrial electronics 111 395 2,000 389 987
Payments 22 68 1,643 34 144
Video and audio entertainment 35 256 1,500 135 407
Cloud services 20 54 1,026 10 109
Electric vehicles 13 1 941 — 101
Information-enabled 36 140 888 54 154

business services
Total 637 2,894 26,686 1,241 5,100
Growth for arenas of today: 9.2× 4.2×

Growth for all other industries: 2.4× 1.7×
Note: Based on MIC (McKinsey Industry Classification); Quality 4 & 5 data from McKinsey Value Intelligence, PitchBook only; subsidiaries excluded; only firms with
market cap >$3.5B in 2005 or >$5B in 2020 are included; # of firms by arena varies; firms identified as did not exist/non-public in ’05 based on no market cap
data in 2005 within McKinsey Value Intelligence.
Source: McKinsey Global Institute analysis
McKinsey & Company

Arena-
Following is a brief description of the 12 arenas of today, ranked by their market cap in 2020.
Software. Companies such as Microsoft, Oracle, and Adobe that develop, maintain, and distribute
software, increasingly as a service.
Semiconductors. Companies such as Intel, TSMC, and Nvidia that design and manufacture
semiconductors. This arena also includes the providers of tools used to manufacture semiconductors,
such as ASML.
Consumer internet. Companies such as Alphabet, Meta, and Tencent that provide consumers with
internet services, including search, social media, and email.
E-commerce. Companies such as Amazon, JD.com, and Alibaba that only sell products online or
provide online marketplaces. This arena also includes the divisions of traditional brick-and-mortar
retailers that engage in online sales.
Mobile and other consumer electronics. Companies such as Apple and Samsung that engineer
and manufacture personal mobile and other consumer electronics, such as smartphones, tablets,
wearable devices, and televisions.
Biopharmaceuticals. Companies such as Amgen and Regeneron that develop and manufacture
drugs based on biological materials like proteins and nucleic acids instead of producing traditional
pharmaceuticals, which are typically drugs consisting of simpler chemical compounds. This arena
also includes the biopharma segments of traditional pharma players, such as Pfizer and AstraZeneca.
Industrial electronics. Companies such as the contract manufacturers Foxconn, Jabil, and Flex that
handle outsourced manufacturing for hardware designers (such as Foxconn for Apple). Also in this
arena are original equipment manufacturers (OEMs) such as Panasonic, Siemens, and ABB that
manufacture electrical equipment, testing devices, and components.
Payments. Companies such as Visa, Mastercard, and American Express that provide payment and
transaction services like the networks that process credit-card payments.
Video and audio entertainment. Companies such as Paramount, Disney, and Netflix that provide
video and entertainment via broadcasting, video streaming, and audio streaming, as well as record
labels such as Warner Music.
Cloud services. Companies and their divisions such as AWS, Microsoft Azure, and Google Cloud
Platform that provide IT infrastructure or platforms as online services, such as cloud computing and
cloud storage.
Electric vehicles. Companies that manufacture EVs, ranging from those specializing only in EVs, such
as BYD and Tesla, to the divisions of traditional automotive OEMs like Toyota, GM, and Mercedes-
Benz, which have added EVs or plug-in hybrid EVs to their product lines.
Information-enabled business services. Professional services firms such as Deloitte; information,
data processing, and analytics providers such as ADP and S&P; and consumer-credit-reporting
agencies such as Experian, Equifax, and TransUnion.
Six key metrics that make today’s arenas stand out

We highlighted six metrics in which the arenas of today outpaced other industries: share of economic
profit, share of R&D growth, proportion of new entrants, proportion of companies with a market
cap of more than $200 billion, proportion of market cap held by top ten players, and proportion of
companies that generate at least 20 percent of revenues from outside their home country (Exhibit 4).

Arena-
Exhibit 4
Arenas of today stood apart from

non-arenas in terms of … Arenas of today Other industries
549
250 264 290
55
increasing Economic profit,¹
$ billion
economic profit
–221
2005 2019 2020
requiring more Distribution of increase

in US spending
R&D investment on R&D, 2005–20, % 67 33
enabling new Share of market cap

held by new entrants,
entrants to grow 2020, % 33 15
spawning giants Share of total market cap

held by companies with
market cap of more than 50 15
$200 billion in 2020, %
tending to be Share of market cap

held by top 10 players,
more concentrated 2020, % 75 56
operating Share of companies

with at least 20% of
more globally revenues from outside 68 55
their home country, %
¹We cite both 2019 and 2020, acknowledging 2020 was an exceptional year because of the COVID-19 pandemic.
McKinsey & Company
Arenas capture an increasing share of economic profit. Today’s arenas captured a much larger share
of economic profit in 2019 than they did in 2005. Economic profit is what is left over after subtracting
the cost of capital from net operating profit. Put differently, it is revenue minus the explicit costs of
doing business and the implicit opportunity costs. In 2005, the industries that went on to become
arenas generated $55 billion, or 9 percent of total economic profit in our data set, while the other
industries we studied generated a combined 91 percent of the total, or $549 billion. By 2019, the
arenas were generating $250 billion, or 49 percent of total economic profit. And in the exceptional

Arena-
pandemic year of 2020, the economic profit of arenas continued to increase, while collectively, the
economic profit of other industries was negative. Underpinning this increase is a persistently higher
return on invested capital of 28 percent in the arenas and 12 percent outside of them (Exhibit 5).
Comparing 2005 economic profit rankings with those for 2019, every arena other than industrial
electronics moved up (Exhibit 6).5
Exhibit 5
Arenas’ share of economic profit grew from 9 percent in 2005 to 49 percent

in 2019.
Economic profit, $ billion

Arenas of today Other industries
900
9% 49%
800
share share
700
600
500
400
300
200
100
–100
–200
–300
2005 2007 2009 2011 2013 2015 2017 2019 2020
Arenas 24 26 25 28 31 33 31 30 30
Others 13 14 11 12 11 10 11 12 10
Return on invested capital,¹ %
¹Computed as net operating profit after tax/invested capital for a given year.
Source: McKinsey Value Intelligence; McKinsey Global Institute
McKinsey & Company

Arena-
Exhibit 6
Arenas rose in rankings of economic profit.
Industry ranking by economic profit

2005 2019
1 Consumer electronics Arenas
2 Consumer internet of today
3 Pharmaceuticals Other
4 Semiconductors industries
5 Biopharma
6 Hardware
7 Everyday and general retail
8 Medical technology
9 Conglomerates
10 Software
11 11 Insurance
12 12 Home and personal goods
13 Telecom
14 Automotives
15 Apparel and luxury
16 Retail and commercial banks
17 Beverages
18 18 Payments
19 IT solutions and services
20 Industrial machinery
21 Tobacco
22 Packaged food
23 Industrial equipment
24 24 Restaurants and food service
25 Travel and leisure
26 Gaming
27 Air services
28 Healthcare
29 29 Information-enabled business services
30 30 E-commerce
31 Consumer durables
32 Aerospace and defense
33 Cloud services
34 34 Recruitment support
35 35 Other business services
36 Leisure products
37 Building materials and products
38 38 Agriculture
39 Consumer services
40 40 Video and audio entertainment
41 Media
42 Wholesale trading
43 Industrial electronics
44 Real estate (excluding China)
45 Freight and logistics
46 Other utilities
47 Forest products and packaging
48 Electric vehicles
49 49 Rail and transport
50 Construction engineering and services
51 Mining
52 Chinese real estate
53 Chemicals
54 Oil and gas
55 Life insurance
56 Electric power
57 57 Nonbank financials
McKinsey & Company

Arena-
In 2019, consumer electronics led the arenas in economic profit with $58 billion, followed by
consumer internet with $48 billion and semiconductors with $45 billion. In 2020, the same three
arenas were in the lead. Many developments—including rising global smartphone penetration,
companies’ widespread shift to cloud computing, and the increase in microchips embedded in a
multitude of physical devices such as automobile components—boosted demand for the products in
these arenas and fueled profits.
The decrease in economic profit among non-arena companies during the same period resulted from
a number of factors, including a long-term decline of performance in commodity-driven energy and
material sectors, lackluster performances by companies based in Europe, and lower profitability
in the world’s “next largest” companies (those that rank beneath the 500 largest companies by
revenue).6 The industries showing the biggest reduction in economic profit from 2005 to 2019 were
oil and gas, which dropped $143 billion; retail and commercial banks, which fell $69 billion; non-bank
financials, which dropped $58 billion; and mining, which fell $45 billion. The COVID-19 pandemic
exacerbated these effects in 2020. Of course, these dynamics change with economic cycles,
and some of the industries that trended downward from 2005 to 2020 may see a resurgence; for
example, energy has experienced an upswing.
Arenas are where investment and innovation happen. Today’s arenas received a disproportionate
amount of R&D investment from 2005 to 2020. In the United States, 62 percent of all R&D spending
was already allocated to arenas and arena-adjacent industries in 2005.7 That share increased to
65 percent in 2020. Despite arenas’ larger base, their R&D spending grew more than R&D spending
in other industries over the same period (Exhibit 7). As a share of revenues, arenas also spent more
Exhibit 7
Arena companies account for a disproportionate amount of R&D spending

in the US.
R&D spending among US companies, $ billion
538
CAGR,
188 2005–20, %
Other industries 6.2
Arenas of today 7.0
204
77 350
128
2005 2020
Source: National Center for Science and Engineering Statistics and Census Bureau, Business Enterprise Research and Development Survey, 2005 and 2020;
McKinsey Global Institute analysis
McKinsey & Company

Arena-
on R&D: in 2020, 10 percent of the revenues of arenas and arena-adjacent industries went to R&D,
compared to 5 percent for other industries.8
In data from the US National Center for Science and Engineering Statistics, semiconductors and
electric components led in R&D spending, followed by biopharmaceuticals and software. The
semiconductor industry spent $43 billion on R&D in 2020, up from $19 billion in 2005. This increase
reflected intense competition resulting from the disaggregation of the semiconductor value chain.
Under the new model, companies made a series of competitive moves across the value chain,
escalating investments in increasingly expensive efforts to stake out the technical frontier of one
product segment or a single step of the value chain. Nvidia, for example, increased its annual R&D
spending by a factor of 11 between 2005 and 2020 to become the leading player in the market for
graphics processing units (GPUs).9
Biopharma followed with an estimated $92 billion in R&D spending in 2020, up from $35 billion
in 2005. This reflected the increase in pharmaceutical companies’ average R&D spending as
a percentage of revenues from about 17 percent in 2005 to 25 percent by 2020.10 In addition,
biopharmaceuticals R&D grew at a 14 percent average annual growth rate in the same period,
compared with 4 percent for traditional pharmaceuticals. Software followed with an increase in R&D
spending from $17 billion in 2005 to $35 billion in 2020.
Arenas attract “outsiders.” Arenas are fertile ground for new entrants. In 2020, 33 percent of arenas’
total market cap was held by companies that had been “outsiders” in 2005—that didn’t exist, had
market caps of less than $3.5 billion, or existed but were not yet meaningfully competing in those
arenas. In other industries, just 15 percent of the total market cap in our data set in 2020 was held by
businesses new to those industries. The entry of new players often happens during the early stages
of arena formation, when competitors make investments as they discover which innovative products
and services meet customer demand. Naturally, this means higher dynamism observed in arenas as
competition plays out.
The consumer internet arena is a good example. About half of that industry’s companies—including
giants such as Meta, Meituan, and Shopify—either did not exist or did not report any public market
cap in 2005.11 That said, it would be wrong to think that only young companies belong in arenas.
Microsoft and Apple are nearly half a century old.
Arenas spawn giants. Arenas are more likely than other industries to include the world’s largest
companies. In 2020, 74 percent of arenas’ total market cap was held by companies with market caps
greater than $50 billion, compared with 47 percent for other industries. Fifty percent of arenas’ total
market cap was held by companies with market caps greater than $200 billion, compared with only
15 percent for other industries (Exhibit 8). The same pattern is visible from a perspective of number
of firms. In 2020, 15 percent of arena companies had market caps greater than $50 billion, compared
with 9 percent of non-arena companies. Four percent of arena companies had market caps greater
than $200 billion, while only 1 percent of non-arena companies met that benchmark.

Arena-
Exhibit 8
Most of arenas’ total market cap is held by ‘giant companies.’
Distribution of market cap by company size, 2020
Market cap, Distribution of market cap within each industry grouping, %

$ trillion Companies by size: >$200 billion $50 billion–$200 billion <$50 billion
Arenas of today 26.7 50 24 26
Other industries 52.5 15 32 53
Total 79.2
McKinsey & Company
While arenas are more likely to see most of their market cap driven by giant companies, it is also true
that giant companies are more likely to be in arenas. Of the companies that have market caps above
$200 billion, more than half belong to arenas, even though arenas represent only one-fifth of the
overall sample by number of companies (Exhibit 9). In 2005, only one of the top ten companies was in
a future arena—Microsoft, with a market cap of $278 billion. By 2020, eight of the ten were in arenas,
with market caps ranging from $511 billion to $1.7 trillion.12 The market cap lead was even more
pronounced by 2024.
Exhibit 9
‘Giant companies’ are more likely to be in arenas than in non-arenas.
Distribution of 3,075 sample companies by market cap, 2020
Market cap Distribution Distribution of companies within each market cap grouping, %
of companies, Companies in: Arenas of today Other industries
#
>$200 billion 44 52 48
$50 billion–
261 27 73
$200 billion
<$50 billion 2,771 19 81
Total in sample 3,075 20% of companies in overall sample are in arenas
McKinsey & Company

Arena-
Information technology played an enormous role in the growth of arena giants. In 2005, the ten
largest companies by market capitalization included several oil and gas firms as well as other well-
known brands, such as GE and Walmart. By 2020, the top ten were led by tech companies: Alibaba,
Alphabet, Amazon, Apple, Meta, Microsoft, and Tencent. By 2024, there was only one non-arena firm
in the top ten.
Arenas tend to be more concentrated. Market cap and revenues are more concentrated in the
largest arena companies than in non-arena companies. Consider the ten largest players in each of
our arenas in 2020. The top ten in five arenas—cloud services, consumer electronics, consumer
internet, EVs, and payments—accounted for at least 90 percent of market cap and revenues in our
sample in 2020. Similarly, the top ten players in e-commerce and video and audio entertainment had
about 80 percent of both market cap and revenues in 2020. That is, more than half of the arenas met
this 80 percent threshold. By comparison, only a quarter of non-arena companies met this cutoff for
market cap and less than two-fifths did so for revenues.
But the industry structure is often not static or stable in the longer term. Even when arenas are
concentrated, the competitive pressure to innovate persists, as investments that improve products
or take advantage of network effects can have increasing returns, requiring the leaders to continually
push to remain on top (discussed further in chapter 2). In the consumer internet arena, for instance,
the three largest companies—Alphabet, Meta, and Tencent—accounted for 72 percent of market
cap and 78 percent of revenues in 2020. In 2012, those same top three accounted for 73 percent
of market cap. During that time intense competition shifted among the three companies, with
Alphabet’s share dropping from roughly 50 percent to 30 percent and both Meta and Tencent
increasing from roughly 10 percent to 20 percent.
Furthermore, arenas are often markets with high levels of innovation, leading to step changes in
technology and business models, which can disrupt the existing group of winners (see “Arenas are
where investment and innovation happen” in this chapter). In some instances, the step changes also
result in new spin-off arenas forming, which we explore in chapter 3. For example, the emergence
of chatbots powered by generative AI (gen AI) threatens the advertising-supported web search
business model, and platforms with short-form videos recommended by algorithms have grown
quickly to capture user engagement in the social media space.
Arenas are more global. Arena companies tend to be more global than non-arena companies.
On average, 50 percent of revenues in arenas were generated outside companies’ home regions,
compared with 42 percent for non-arena companies. Moreover, companies in arenas were much
more likely to be multinationals.13 Sixty-eight percent of companies from arenas derived more than
20 percent of their revenues from countries other than their own, compared with about half of
companies in other industries.
Software is a particularly global arena. Fifty-three percent of its revenues came from outside its
component companies’ home regions in 2020. The four largest companies by 2020 revenues—
Microsoft, IBM, Oracle, and SAP, which together accounted for more than half of the software
industry’s revenues—took in almost 60 percent of those revenues outside of their home countries.
The global spread of these technology companies is unsurprising, given the digital nature of their
products and services: they are not limited by shipping costs, physical plants, or input limitations, at
least not directly (see sidebar “Companies based in the US and Greater China are disproportionately
represented in arenas”).

Arena-
Sidebar: Companies based in for a substantial majority, 65 percent, of Companies based in Europe accounted
the US and Greater China are 2020 market cap. US companies accounted for substantial market cap in biopharma,
for the majority of market cap in every arena industrial electronics, and information-
disproportionately represented
except industrial electronics, where China enabled business services. However,
in arenas took the lion’s share. China also participated European companies had a much higher
materially in the consumer internet, EVs, share of global market cap in non-arena
Companies based in the United States were e-commerce, and semiconductors arenas, industries (22 percent) than they did in
highly represented in arenas and accounted with 22 to 30 percent of market cap. arenas (9 percent) (exhibit).
Exhibit The US and Greater China account for the majority of market cap across all
arenas.
Regional distribution of market cap by arena, 2020,¹ %

US Greater China Europe Rest of world
0 10 20 30 40 50 60 70 80 90 100
Payments
Video and audio

entertainment
Software
Consumer electronics
Cloud services
Electric vehicles
Information-enabled
business services
E-commerce
Consumer internet
Biopharma
Semiconductors
Industrial electronics
All 12 arenas
All other industries

0 10 20 30 40 50 60 70 80 90 100
¹Companies’ regions defined by their country of incorporation.

McKinsey & Company

Arena-
Four side-by-side examples

Some arenas are carved out of large established industries that don’t exhibit outsize growth and
dynamism. This gives us a natural set of comparable examples of traditional industries and arenas
to illustrate these differences. We describe four of them below: payments and banking, biopharma
and pharmaceuticals, EVs and traditional ICE vehicles, and e-commerce and brick-and-mortar retail
(Exhibit 10). The four relatively new industries are examples of arenas that have sufficiently novel
technologies or business models to merit separating them from the industries they disrupted during
this period.
Exhibit 10
Our sample arenas surpassed their traditional counterparts in economic

growth, market concentration, and opportunities for new entrants.
Comparison of 4 arenas and their traditional counterpart industries

Investor Value Players More New entrants

attraction creation growing in size concentration Market share
Industry Total invested Total economic Average Top 5 in 2020 held
matchups capital, profit, market cap, companies’ by companies
CAGR, CAGR, CAGR, market share, that had none
2005–20, % 2005–19, % 2005–20, % 2020, % in 2005,¹ %
Payments 18 11 24 77 33
vs
Banking –2 –10 3 18 3
Biopharma 11 14 12 28 33
vs
Pharmaceuticals 3 1 4 33 14
Electric vehicles n/a² n/a 56 94 100

vs
Internal 4 10 5 30 16
combustion
engine vehicles
E-commerce 27 11 27 73 52
vs
Retail 2 3 4 44 11
¹Defined as non-incumbents from 2005 (ie, firms that didn’t exist, had market cap of less than $3.5 billion, or existed but were not yet meaningfully competing in
those arenas).
²From $0 in 2005 to $46 billion in 2020.
McKinsey & Company

Arena-
Consider how the payments industry, which is one of the arenas of today, compares with the closely
related banking industry, which is not an arena. Banking showed economic profit of $88 billion in
2005, but by 2019, economic profit had shrunk to $21 billion. During the same period, economic profit
in the payments industry rose from $4 billion to $17 billion. Payments was also far more concentrated
than banking: the top five companies in payments held 77 percent of market cap in our data set in
2020, while the top five in banking held 18 percent. Also in 2020, new businesses held 33 percent of
all market cap in payments; in banking, the share was just 3 percent.
Or compare the biopharmaceuticals arena with traditional pharmaceuticals. Economic profit in this
arena increased more than sevenfold, from $6 billion in 2005 to $42 billion in 2019. By contrast,
economic profit in traditional pharmaceuticals increased by only 10 percent, from $42 billion in 2005
to $46 billion in 2019. The biopharmaceuticals industry also offered more opportunities to new
entrants, which had 33 percent of market cap in 2020—a much larger share than the 14 percent new
businesses held in traditional pharmaceuticals.
And consider the EV industry alongside the traditional auto industry. In 2005, the fledgling EV
industry had very few players, but by 2020, it received all of its $101 billion in revenues from new
businesses. It was also highly concentrated: the five largest EV players accounted for 94 percent of
market cap, far more than the 30 percent share held by the five largest traditional auto companies.
Finally, consider the e-commerce and retail industries. Between 2005 and 2019, e-commerce’s
economic profit grew from $1.1 billion to $5.6 billion, compared with retail’s growth from $20.2
billion to $32 billion. E-commerce also experienced explosive growth during the COVID-19 crisis,
generating $16 billion in economic profit in 2020. In 2020, the five biggest e-commerce companies
held 73 percent of the industry’s market cap, and new businesses controlled 52 percent. In retail, the
five biggest companies accounted for just 44 percent of market cap, and new businesses had only
11 percent.
§ § §
In this chapter, we identified industries that became arenas from 2005 to 2020, and we explored
the ways in which they differed from other industries. We have shown that arenas outperform their
counterparts in R&D investment inflow and value creation, provide more opportunity to new entrants,
and tend to become more concentrated. In chapter 2, we examine how the arenas of the present
emerged and grew from 2005 to 2020 and discuss how that understanding can help us spot the
emergence of the arenas of tomorrow.

Arena-
CHAPTER T WO
The arena-creation potion
To understand how our present-day arenas came into being, we need to take a time machine to go
back about two decades. What would we have seen? Let’s look at four industries that became arenas
from the early to mid-2000s.
In 2005, e-commerce had a market cap in our data set of only $87 billion. The industry reported
revenues of $15 billion, accounting for only 2.5 percent of overall US retail sales. Amazon had
recorded rapid revenue growth, from $16 million in 1996 to $8.5 billion in 2005. It was investing in
its merchant network, started in 2000, and in its fulfillment services, launched in 2006. At the same
time, the world was becoming more digital, as US internet penetration increased from 9 percent in
1995 to 68 percent in 2005.14 Other players were also making moves: eBay acquired PayPal in 2002,
and Shopify launched its alternative to the Amazon ecosystem in 2006. By 2020, e-commerce had a
market cap of $3.3 trillion and revenues of $888 billion in our data set.15
The video and audio entertainment industry, which had a market cap of $256 billion in 2005,
also began to transform. At the beginning of this period, content producers, such as Disney and
Paramount, distributed their productions through traditional movie theaters and through DVD
rentals offered by the likes of Blockbuster and Netflix, as well as broadcast and cable TV. The
subscription model began to emerge in 2003, when Netflix filed a patent for subscription rental
services for physical DVDs. In the next few years, the company grappled with Blockbuster over
pricing and delivery models. The Netflix subscriber base continued to grow, from 1.4 million in 2003
to 7.3 million in 2007. Revenues grew, too, from $272 million in 2003 to $1.2 billion in 2007.16 In 2007,
Netflix launched online streaming, now its primary business model. By 2020, the video and audio
entertainment industry’s market cap reached $1.5 trillion.
In 2005, the semiconductor industry’s market cap was just $642 billion. That year, Intel launched
its first multicore processor, and in 2006, it announced the construction of a production facility in
Vietnam, its biggest to date. Today’s widely used Intel Core i7 processor debuted in 2008. Meanwhile,
the semiconductor manufacturer TSMC started to strengthen its leadership in the foundry segment.
From 2000 to 2005, the company had started production in four new fabrication plants, with an
associated initial investment of about $14 billion. TSMC started to spend 20 percent more on R&D
than its three biggest competitors combined in 2008. By 2020, the semiconductor industry was a
$3.5 trillion arena by market cap.
The EV industry also started to coalesce into an arena in the early 2000s. The first modern lithium-
ion batteries were commercialized in the early 1990s in Japan, and by the early 2000s, battery packs
capable of powering a car were assembled. Honda and Toyota entered the hybrid EV market in this
period. Tesla began development of its Roadster in 2004. It launched in 2008 and was the first pure
EV, with a range of more than 240 miles per charge. By 2020, the EV industry had a market cap of
$941 billion.
In the early 2000s, perfectly predicting the path of these four industries would have been an
impossible task. In hindsight, we observed interesting developments related to growth, investment,
and new competitors happening in these arenas. How did the primordial ooze of these early
indicators bubble up into an arena?

Arena-
In this chapter, we describe the inner processes transforming industries into arenas that generate
growth and dynamism and examine how those characteristics manifested themselves in the 12
arenas of today.
The arena-creation potion and escalatory competition

What are the elements that give rise to an industry that displays high growth and high dynamism?
We have identified three that, when combined, become what we call an “arena-creation potion.” The
three ingredients are: technology or business model step changes, an escalation mechanism for
investments, and a large or growing addressable market (which investors often refer to as the total
addressable market, or TAM).
The three ingredients emerge in a certain order. We observed that arena formation starts with a
technology or business model step change that encourages players to invest rapidly in quality. These
escalating investments constitute the second ingredient. The players then undertake the cycle of
continual investment to tap a large or growing addressable market. However, while there is a causal
order to the three ingredients, there is also an unpredictable element of time as players churn, invest,
and innovate until the latent demand of a large TAM is unlocked and an arena is formed.
Technology and business model step changes
The first potion ingredient is a technology or business model step change that fundamentally
transforms how products and services are developed or delivered. Performance for a given
technology (for example, how many transistors can fit into a square centimeter of silicon, or how much
energy per kilogram can be stored in a battery) and adoption of a technology (what percentage of a
given market is using that technology) are often modeled as idealized S-curves. When a technology
capability undergoes a step change, adoption starts off slowly, reaches an inflection point when it
accelerates, then begins to plateau as the technology or business model reaches maturity. At the top
of the S-curve, the competitive game tends to settle into mature markets competition (contests with
a few large competitors and relatively high barriers to entry), but the intensity of competition among
them can persist, as we describe later in this chapter.
In practice, technology capabilities and adoption don’t always follow this idealized S-curve form
perfectly, but the conceptual pattern is a useful framework. We observe technology step changes
in our arenas when a technology innovation is of a substantial magnitude and puts the trajectory of
performance on a new S-curve, as we saw with EVs.
We also observed step changes in business models in our arenas. These were often enabled by
technology and caused by innovations that shift the commercial models (in other words, who pays
for what and how) related to products or services, disrupting existing market structures, as with
e-commerce and video and audio entertainment (streaming). You could also think of the adoption of
business model innovations as following S-curves.
In the semiconductor industry, we saw both technology and business model step changes. This
industry regularly sees technology step changes in the form of “node resets”—broadly, the process of
building new generations of semiconductors with smaller transistors and other components. Over the
first two decades of the 2000s, semiconductor technology jumped to smaller and smaller node sizes,
effectively resetting the technology curve about every five years, in line with the exponential increase
in the number of transistors per integrated circuit predicted by Moore’s law.
These regular industry-wide shifts to smaller and smaller node sizes caused the markets for larger
node sizes to decline over time. In 2000, production spanned the 0.13-to-0.5-micron (130-to-500-
nanometer) range. In just ten years, only a small segment of the market was producing chips with
nodes larger than 90 nanometers (Exhibit 11).17

Arena-
Exhibit 11
Spending on wafer fabrication equipment by semiconductor makers

has increased since 2000, focusing on decreasing node size.
Distribution of spending on semiconductor wafer fabrication equipment, by node size, %

Node size (smaller = more advanced technology),
nanometers
>500 Projected
500
350
250
180
130
90
65
45
32
22
14
10
<7
0 40 0 40 0 40 0 40 0 40 0 40
2000 2005 2010 2015 2020* 2025*
33 26 32 33 65 113
Wafer fabrication equipment sales, $ billion
*Values for 130 nanometer node size include spending on equipment for all node sizes 130 nanometers and larger.
Source: Gartner (for full information, see endnote 17)
McKinsey & Company
Over the same time span, much of the industry adopted a “fabless” business model, with some
companies handling the design and sale of semiconductors while outsourcing fabrication of
the physical microchips to other companies called foundries. Some players, including Nvidia,
Qualcomm, and Broadcom, specialized in designing microchips that served specific purposes.18
Others, such as TSMC, specialized in manufacturing chips designed by fabless players, while ASML
and others focused on the specialized equipment used to make chips (for instance, for lithography).
As a result, the industry underwent a dramatic shift from vertically integrated business models
to disaggregated specialization. However, some players, notably Intel, maintained a vertically
integrated design and manufacture model for microchips, while sourcing the tools in their factories
from specialized suppliers.

Arena-
Escalation mechanism for investments

The second potion ingredient can be observed in the type of investments players have
incentives to make, namely escalatory investments. Escalatory investments have two mutually
compounding consequences.
First, the returns of these investments increase with scale. These outlays not only grow output but
fundamentally change and improve a company’s production function. In other words, they increase
quality, not just quantity. As a result, these investments can boost margins and rapidly expand market
share, because customers want the best product. By contrast, so-called ticket-to-play investments,
such as simply opening new factories and branches that aren’t fundamentally different from existing
ones, do not influence market position as much.
Second, players begin an arms race in which they iteratively invest to scale and scale to invest. When
one player improves quality and starts to gain more profits and market share, other companies
respond and invest more to improve the quality of their product to compete. The more each player
invests, the more competitive the race becomes, creating an escalatory cycle. We expand on this
intense mode of competition in the next section.
Escalatory investments are features of specific types of spending, such as marketing, R&D,
and certain capital expenditures. These are the kinds of investments that advance a company’s
capabilities and tend to improve long-term margins. For example, when e-commerce platforms invest
in marketing to attract more customers and increase the platforms’ value for merchants, the cost of
mass marketing to customers shrinks relative to revenue potential, expanding per-customer margins.
Biopharma companies often invest in novel R&D methodologies, such as AI-driven bioinformatics, to
improve clinical success rates and return on investment (ROI). Once a drug is produced and marketed,
the cost of R&D investment is spread across that drug’s sales. Acquisitions can be a form of R&D
investment when they provide access to the capabilities and proprietary assets of the acquired
companies. In many instances, an underlying network effect makes investments attractive for arena
companies (a process we discuss later in this chapter). Furthermore, these types of investments are
not one-offs when they are escalatory—they are continual and increasing.19
Of course, not all investment-intensive industries exhibit this escalatory dynamic. In the steel
industry, building a new plant might increase overall profits, but only by increasing capacity and sales
within the same production function. However, e-commerce players like Amazon, as they optimize
last-mile delivery to the home, fundamentally change not just their scale but their unique capabilities,
too (see sidebar “Tech giants made massive escalatory investments”).

Arena-
Sidebar: Tech giants made comparison, US capital expenditures rose 2005, while the company reported spending
massive escalatory investments 3 percent and R&D spending increased $6 billion on R&D. By 2020, its capital
7 percent in the same period.1 This pattern expenditures had reached $15 billion, not
If we look at six of the biggest players of investments has continued since 2020, far behind the $19 billion for R&D. Alphabet,
in several arenas, the magnitude of with all six companies maintaining 15 to on the other hand, went from reporting
escalatory investments is enormous. These 23 percent CAGRs on their R&D expenses roughly equal amounts for R&D and capital
companies—Alphabet, Amazon, Apple, Meta, and capital expenditures from 2020 to 2023. expenditures in 2005 to reporting nearly
Microsoft, and TSMC—collectively invested 30 percent more on R&D than on capital
Beyond the total size of investments,
$13 billion in capital expenditures and R&D in expenditures by 2020, and nearly 40 percent
the allocation between R&D and capital
2005. By 2020, that number had increased more by 2023 (exhibit).
expenditures also showed how investment
20-fold, reaching almost $250 billion, a strategies evolved. For example, Microsoft’s 1
US capital expenditures cover US nonfarm businesses;
22 percent year-on-year growth rate. By capital investments were only $1 billion in 2020 Annual Capital Expenditures Survey, US Census
Bureau, December 2021.
Exhibit Investments by big tech players escalated 20-fold from 2005 to 2020.
Select companies’ spending on capital and R&D, $ billion

Capital R&D
TSMC Apple Microsoft Meta¹ Alphabet Amazon²

Semiconductors Consumer Cloud services, Consumer Cloud services, Cloud services,
electronics consumer internet consumer e-commerce
electronics, internet, video
consumer and audio
140 internet, entertainment
software
120
100
80
60
40
20
0
’10 ’15 ’23 ’10 ’15 ’23 ’10 ’15 ’23 ’15 ’23 ’10 ’15 ’23 ’10 ’15 ’23
2005 2020 2005 2020 2005 2020 2010 2020 2005 2020 2005 2020
Multiple, 2005–20
8× 33× 5× 77× 35× 127×
CAGR, 2005–20, %
15 26 11 54 27 38
¹Meta’s multiple and CAGR numbers are indexed to 2010 instead of 2005 due to data availability.
²Amazon’s R&D expenditure uses the company’s “technology and infrastructure” expense, reported under the GAAP requirement ASC 730 for research and
development expenses.
McKinsey & Company

Arena-
To illustrate this dynamic within an arena, we looked at four of the largest players in our sample of
cloud-services companies from 2015 to 2020 to determine whether escalating investments—as
indicated by the level of capital expenditures—correlated with outsize revenue market share
(Exhibit 12). In 2015, AWS was the market leader with 23 percent of revenue market share, followed
by Microsoft at 16 percent. In 2020, AWS still led in revenues with 37 percent market share, and
Microsoft followed with 21 percent. The continual escalation of these investments is clear: AWS and
Microsoft pulled further and further ahead to first gain and then at least maintain their market share.
Exhibit 12
In the cloud services arena, capital expenditures continued to escalate amid

intensifying competition.
4 companies’ cloud services revenue and capital expenditures, 2015–20
2015 2020
40
Amazon Web
35 Services (AWS)
In 2020:
• 37% share
• $11.3 billion
30 cumulative
spending
Microsoft
Share of 25
• 21% share
industry • $10.4 billion
revenue,¹ cumulative
% 20 spending
Alibaba
• 8% share
15
• $2.8 billion
cumulative
spending
10
Alphabet
• 6% share
• $5.0 billion
5 cumulative
spending
0
0 2 4 6 8 10 12
Estimated capital expenditures, cumulative,² $ billion

¹Revenue is proportionally adjusted to relevant business units (eg, AWS for Amazon).
²Capital expenditures are estimated by applying the revenue split of each company into business units to each company’s reported capital expenditures,
cumulative starting 2015, 5-year straight-line depreciation applied.
Source: McKinsey Value Intelligence; McKinsey Global Institute analysis incorporating data from IDC
McKinsey & Company

Arena-
Large or growing addressable market

Technological and business model step changes open demand pools that are conducive to escalatory
investments. As a result, these changes unlock large and often rapidly growing addressable
markets—our third potion ingredient. Generally speaking, there are two ways for companies to reach
large or fast-growing markets: either they play in already sizable markets where demand growth
continues to outstrip the rest of the economy or they displace share of an existing large market with a
superior product or service.
Companies in the first category took advantage of technology and business model step changes
to accelerate value creation in fast-growing, established industries, which typically had revenue
pools of more than $100 billion in 2005 but were subject to escalatory investments that improved
efficiencies or broadened capabilities. These arenas included biopharmaceuticals, industrial
electronics, information-enabled business services, mobile and consumer electronics, payments,
semiconductors, software, and video and audio entertainment. Together, they recorded 5 to
13 percent revenue CAGRs from 2005 to 2020. By comparison, the global economy expanded about
3 percent a year during the same period.20 For these arenas, the growth driver of the market was
often an expansion of existing demand brought about by digitization and associated globalization.
In 2020, 60 percent of revenues earned by the companies in these arenas came from outside their
home countries.
Companies in the second category achieved rapid growth by launching novel categories of products
or services, taking shares from existing markets.21 While companies in the first category used
technology and business model step changes to accelerate growth in existing industries, those in
the second category created new industries, which captured growth by eating into the demand from
existing markets. These arenas include cloud services, consumer internet, e-commerce, and EVs.
Their revenue grew at a 13 to 33 percent CAGR from 2005 to 2020.
In this group, digitization often created a superior value proposition. The worldwide adoption of
internet connectivity allowed e-commerce players to more seamlessly connect buyers and sellers
across the globe, providing the opportunity for customers to purchase goods anytime, anywhere
with their electronic devices, rather than through traditional physical retail channels. Digitization
also transformed video and audio entertainment, and streaming in particular, by shifting media
consumption from movie theaters and physical CDs to homes and mobile devices. Players were able
to offer the added flexibility of content options and scheduling to consumers. We further discuss the
importance of digitization later in this chapter.
We have discussed the three elements of the arena-creation potion: a technological or business
model step change, escalatory investments, and a large or growing addressable market. We
observed that the combination of these three ingredients, over time, leads to the formation of arenas.
Next, we expand on the new competitive game unique to arenas when these three ingredients
coalesce, which we call an escalatory mode of competition.
The three ingredients result in an escalatory mode of competition
Arenas produce a unique competitive tussle—an escalatory mode of competition—which results in
high growth and high dynamism.
Escalatory competition is the industry-level consequence that occurs when the three potion
elements come together, resulting in individual companies facing escalatory investment incentives.
As alluded to above, it is characterized by arms-race-style contests in which players continually
invest in their products, advertising, and operations because a technological or business model step
change has disrupted the industry and unleashed latent demand.

Arena-
To compete for or even maintain market

share, companies need to continually
increase investments and improve
their capabilities.
As arena companies invest more in capabilities and the quality of their products improves, their
profit eventually rises. They achieve this by increasing customers’ willingness to pay (with product
improvements or advertising, for example), by reducing costs (by making step-change improvements
in customer acquisition costs, for instance), or by growing volumes (for example, by leveraging
network effects). Further, advancing capabilities tends to allow companies to expand their economic
catchment zones: they can now sell products and services beyond old boundaries of geographies or
categories.
The competitive landscape changes as an arena moves up the S-curve of technical capability and
adoption. Initially, the possibility of translating innovation to better capabilities and rapid market
share can pull in new entrants at the beginning of the S-curve. Escalatory competition then tends
to occur during the phase of rapid growth. This movement of an industry along the S-curve marks
high levels of growth and competitive dynamism with major market share shifts, both of which are
hallmarks of an arena. Eventually the escalatory investment dynamic can put leaders out of reach
in the plateauing phase of the S-curve as new market players may find it increasingly challenging
to compete with the capabilities that incumbents have developed, limiting the number of players,
although they could continue to compete vigorously with one another.
There are two other general modes of competition. Simple competition features low barriers to
entry and minor competitive advantages. Mature markets competition occurs when a few large
players capture a significant portion of market share. What distinguishes escalatory competition
from the other modes is that players are required to keep improving product quality, enabled by new
technologies and rewarded by large addressable markets, intensifying the battle for market share. To
compete for or even maintain market share, investments and capability-building must be continual;
players that do not match the velocity of investments would eventually be unable to compete. Other
modes of competition do not result in this kind of escalatory dynamic (see sidebar “The differences
between the general modes of competition”).

Arena-
Sidebar: The differences model step changes, or both can disturb a result, current players advance their
between the general modes that equilibrium. capabilities and can even expand to other
markets. The most dynamic competitively
of competition The third mode, escalatory competition, is
escalatory industries—and hence arenas—
characterized by an arms-race dynamic in
are often found at the beginning of the
Three general modes of competition which players continually invest to advance
S-curve, when technological or business
demonstrate why escalatory competition is their capabilities. In his 1991 book Sunk Costs
model step changes occur, or in the
different and is characteristic of arenas. and Market Structure, the economist John
succeeding, middle phase of rapid growth in
Sutton presented a theory that helps explain
In the first mode, simple competition, barriers the S-curve.
the difference between mature markets
to entry are low and individual players cannot
competition and escalatory competition. He For example, we observed competitive
unilaterally move prices. This is the case with
asserts that company investments can be escalatory dynamics in the consumer
a neighborhood laundromat or locksmith:
described as either exogenous sunk costs or internet arena, especially in the early years
a minimal investment is required to enter,
endogenous sunk costs. Exogenous sunk- of Meta as the company continued to invest
the technology is well understood, there are
cost investments are those that any player in capabilities and expanded demand from
many small players, competition is high, and
needs to make to enter the market. Examples 2005 to 2020. Acquisitions made by large
the competitive advantages, if they exist,
include the cost of building a factory or players are sometimes signs of mature
are minor.
of setting up a logistics and distribution markets competition, so Meta’s acquisitions
The second mode, mature markets network, or the heavy capital investments of WhatsApp and Instagram might have
competition, is a model in which a few large required in the commercial airline suggested that the consumer internet arena
companies have captured a significant industry. Endogenous sunk costs are fixed was reaching an S-curve plateau. But we also
portion of the market share. This often investments that improve long-term profit see innovations that create new dynamism:
occurs at the plateauing end of the S-curve, by reducing cost or increasing customers’ TikTok, for example, started to meaningfully
where technological performance or willingness to pay. These investments are compete in the consumer internet arena after
adoption reaches maturity. New rivals classified as endogenous as they tend to it launched in China in 2016 and expanded
are limited by high barriers to entry like depend or build upon investments already internationally in 2017. By 2020, TikTok had
a minimum efficient scale or high capital made. Think of product improvements or more than 1 billion users, while Facebook had
investment. In this mode, the parameters marketing campaigns that allow companies almost 2 billion. New business models and
of the competitive game seem settled for like Apple or Samsung to maintain the price redefined industry lines may produce the
some time, with established business models of their products, or a process innovation arenas of tomorrow. We discuss this further
as well as physical or intangible assets that optimizes production performance (for in chapter 3 and in the compendium that
and capabilities. A prominent example is example, investments by Intel, TSMC, or follows it.
the long-haul commercial airline industry, Nvidia that permit the company to efficiently
Other industries, including long-haul
where high levels of up-front investment in produce at smaller nodes). The cost savings
commercial air travel, also underwent
aircraft are necessary to operate at scale, that companies derive from economies
shifts in competitive modes before our
while regulatory requirements and limited of scale are typically the returns from
2005–20 focus period. This market featured
landing slots constrain the entry of new endogenous sunk-cost investments.
rapid innovation, high investment, and
players. As a result, competition tends to
Escalatory competition occurs when substantial market share reshuffling before
revolve around pricing shifts. Under mature
markets are suited to endogenous sunk- a series of mergers and acquisitions led to
markets competition, the market tends
cost investments. In such cases, incumbent the current stage of quality, especially for the
toward equilibrium, although competitive
companies will keep investing in R&D, largest players.
escalations, technological or business
marketing, or capital expenditures. As

Arena-
Observing the potion in the arenas of today

In this chapter, we’ve described the three ingredients of the arena-creation potion and explained
why they result in high growth and high dynamism. In this section, we explore the importance of
digitization in enabling the rise of the three potion ingredients and how these ingredients coalesced
to create exceptional growth and dynamism in each of our arenas of today. We also explore metrics
that could have signaled the emergence of these arenas (see sidebar “Early markers of arenas”).
Digitization has been a primary

force in expanding the number of
industries where an escalatory
competitive dynamic could develop.
Sidebar: Early markers of arenas — Heightened value expectations. We These investment markers were telling
might expect valuations (measured here indicators for several candidate arenas and
Beyond the arena-creation potion, we also as the ratio of enterprise value to net helped guide us in estimating the arenas that
explored metrics, such as capital flows, operating profit less adjusted taxes, or might eventually emerge. However, like all
valuations, and revenue growth, that could NOPLAT) to predict arena emergence. predictive markers, they had limitations.
signal the presence of arenas in their We did find evidence that in any given
early stages. We discovered that these year from 2003 to 2007, the majority
early markers varied in reliability but were of arenas were in the upper half of the
nevertheless helpful in pointing to potential distribution of industries for valuation
areas where arenas could emerge when they multiples. Only biopharmaceuticals had
supplemented the three potion ingredients. multiples that were consistently lower
What would these metrics have told us in than industry medians. We also looked
2005 about arenas in their infancy? We at the overall EV/NOPLAT multiples of
examined the following (Exhibit): arenas, which were generally higher
than those of other industries in
— Venture capital. A majority of VC flows—
this period.
60 to 70 percent—went to arena-linked
companies, but just four arenas drove — Revenue growth. By definition, arenas
this large share: biopharmaceuticals, had higher revenue growth than the
consumer internet, semiconductors, economy over the 2005–20 period. We
and software. Most arenas received also examined year-on-year revenue
less than 3 percent of the total flows of growth for each arena in the years around
venture capital in any given year from the beginning of our period of analysis,
2003 to 2007. from 2003 to 2008. As expected, we
found that a majority of them grew faster
than our overall sample.

Arena-
Sidebar: Exhibit
Early markers Markers may point to a high likelihood of arenas forming.
of arenas
(contiunued) Investment and revenue metrics during years when 12 industries emerged as arenas
Greater than or equal to benchmark Less than benchmark
Venture capital investment,¹ Valuation,² Revenue growth through

$ billion multiple following year, %
2003 2005 2007 2003 2005 2007 2003 2005 2007
Benchmarks³ 1.1 1.3 1.8 2.0 2.5 21.3 19.8 20.7 20.9 19.8 14 6 14 17 1
Biopharma 4.5 6.3 5.6 7.2 9.8 19.5 17.9 16.4 16.7 15.1 23 9 23 25 6
Cloud services 0.8 1.1 1.1 1.1 1.0 26.8 31.1 23.2 24.3 27.4 16 3 28 24 14
Consumer
3.0 3.6 4.4 5.1 5.5 20.3 24.2 20.9 20.2 22.1 12 5 14 21 –5
electronics
Consumer internet 2.9 4.2 11.4 7.4 7.0 25.6 36.2 30.1 27.7 31.3 26 19 29 26 2
E-commerce 0.2 0.4 1.9 1.4 1.0 22.9 21.6 20.0 19.5 19.7 39 34 112 69 36
Electric vehicles 0.1 0.2 0.2 1.0 0.8 n/a n/a 204 74
Industrial
0.8 1.3 1.2 1.6 2.0 25.3 21.6 22.7 20.9 17.8 21 6 16 22 –1
electronics
Information-enabled
0.8 0.9 1.5 1.5 2.1 19.7 19.6 25.4 24.4 18.7 11 –14 8 20 1
business services
Payments 0.9 0.7 1.6 1.5 2.8 56.3 48.6 48.8 24.4 28.2 16 –5 15 11 11
Semiconductors 3.0 3.1 2.8 3.0 3.2 41.8 20.0 20.0 18.4 18.7 35 3 12 10 –10
Software 11.7 13.7 18.9 20.0 26.5 25.4 27.4 21.5 23.1 24.3 15 9 20 24 16
Video and audio

0.5 1.0 1.2 3.6 3.4 24.6 23.7 23.2 23.7 19.8 2 –9 10 12 7
entertainment
¹Double counting of flows occurs in several instances where multiple arenas apply to one deal (eg, a payment start-up for e-commerce platforms would be counted
in both payments and e-commerce).
²Defined as ratio of enterprise value to net operating profit less adjusted taxes (EV/NOPLAT).
³Benchmark for venture capital is at least 3% of total flows; for valuation, the EV/NOPLAT multiple for 57 industries; for revenue growth, the average of the total
sample rate for that year.
Source: PitchBook; McKinsey Value Intelligence; McKinsey Global Institute analysis
McKinsey & Company

Arena-
Digitization enables network effects and globalization

Digitization has been a primary force in the simultaneous rise of the three potion ingredients by
expanding the number of industries where an escalatory competitive dynamic could develop. For
example, brick-and-mortar retailers face high hurdles to reducing unit costs through investments. A
traditional retailer’s marketing spend is only effective in the physical catchment of the stores, but an
e-commerce retailer can spend into a much less constrained addressable market
Network effects, which increase the value of a product or service as more people use it, were
accelerated by digitization in many arenas, including consumer internet, e-commerce, consumer
electronics, and software.
For example, social media companies’ networks become more valuable to users as they add more
contacts on the platform. Search engines become more accurate and more valuable as more users
engage with them and generate data that can be used to improve search results. E-commerce
marketplace platforms have a two-way network-effect dynamic as more merchants increase the
platform’s value for customers and vice versa. For consumer electronics and the underlying software
that runs them (such as iOS and Android), a mutually reinforcing network effect connects developers
and users: more users entice developers to produce more applications, which in turn attracts
more users.
Globalization of some industries was also a consequence of digitization. Products and services that
would traditionally have local or regional reach are now accessible anywhere. This is particularly
obvious for internet-dependent arenas, such as cloud services, consumer internet, e-commerce,
and software. In the video and audio entertainment arena, even before streaming became
commonplace, globalization had already expanded consumers’ ability to access media, as seen with
K-pop’s rise to international prominence. Globalization also uniquely influenced the information-
enabled business services arena by sparking accelerated growth in the knowledge economy, which
prompted companies in this area to boost their investments in capabilities to address a newly global
addressable market.
The three potion ingredients in the arenas of today
Here we examine how the three potion ingredients have contributed to creating today’s arenas,
discussed in descending order of 2020 revenues.
Industrial electronics. This arena consists of two groups of companies: contract manufacturers,
such as Foxconn, Jabil, and Flex, and OEMs, such as Panasonic and Schneider Electric.
For contract manufacturers, the large-scale outsourcing of manufacturing was the significant
business model step change that made this industry an arena. High-revenue players, such as
Foxconn, benefited from the manufacturing outsourcing of big brands, such as Apple and Sony. As a
result, contract manufacturers often escalate investments in capital expenditures in line with growing
manufacturing capacity and in R&D to build capabilities as they meet demand.
We can see the escalatory nature of the capital investments by the largest contract manufacturers
(Exhibit 13). From 2005 to 2020, Foxconn’s outlays were associated with higher revenue growth
than was the case for its two closest competitors. In 2005, Foxconn had more revenues than Jabil
and Flex, and it devoted a high share of revenues, 5 percent, to capital expenditures. Jabil and Flex
spent 3 and 2 percent, respectively. Foxconn’s continued customer commitments allowed it to
continue this high allocation of capital expenditures as its competitors trailed behind: Jabil’s capital
expenditure allocation reached 5 percent of revenues only in 2015. Foxconn’s revenues continued
to outpace those of competitors, and by 2020, the company spent only 1.2 percent of revenues on
capital expenditures (capital expenditures still grew, but not as fast as revenues), while Jabil spent
3.6 percent of revenues.

Exhibit 13
Foxconn’s capital investments have been accompanied by rising revenues.
3 contract manufacturers’ revenue and capital expenditures, 2005–20

2020
2005
200
Foxconn
In 2020:
175 • $190.7 billion
revenue
• $11.5 billion
cumulative
150 spending
Jabil
Revenue, 125 • $27.3 billion
$ billion revenue
• $4.9 billion
cumulative
100 spending
Flex
75 • $24.1 billion
revenue
• $2.8 billion
cumulative
50 spending
25
0
0 2 4 6 8 10 12
Capital expenditures, cumulative,¹ $ billion

¹Capital expenditures cumulative starting 2005, 10-year straight-line depreciation applied.
Source: IDC; McKinsey Value Intelligence; McKinsey Global Institute analysis
McKinsey & Company
To some extent, the capital expenditure investments by the contract manufacturers were as much
about boosting quantity as quality, and less escalatory in that sense.22 For OEMs, the increasing
digitization of physical devices and equipment was a technological step change that spurred growth.
In this case, OEM players escalated both capital investment and R&D to develop novel manufacturing
processes and the software embedded in these digitally enabled products. For example, Panasonic
announced in 2006 that it would stop manufacturing analog televisions and concentrate on digital
televisions. Siemens invested about $10 billion throughout the 2010s to strengthen capabilities in its
Digital Factory.23 Increasing demand for downstream products raised revenues for these OEMs as
well as the revenues for upstream electronic components produced by contract manufacturers. The
industrialization of developing economies accounted for a large part of the demand growth in this
arena. In our sample, this arena was the largest by revenues in both 2005 ($389 billion) and 2020
($987 billion).

Arena-
E-commerce. A business model step change enabled the e-commerce arena as retail sales
increasingly took place online instead of in physical stores. This was made possible by the growing
penetration of personal computers, mobile devices, and internet connectivity. E-commerce players
escalated investments in physical distribution networks to improve delivery cost margins and delivery
times, increasing customer satisfaction and their value.
Players also escalated investments in marketing to acquire more customers, as scale gives
companies more purchasing power to negotiate favorable terms with suppliers. Marketplace and
platform players further escalated these marketing investments to acquire both customers and
merchants, which enables a mutually reinforcing network effect: customers find more value in a
platform with many merchants and vice versa. This network effect is an important part of Amazon’s
“flywheel effect,” which incorporates escalatory investments in pricing and customer experience.24
In market size and growth, the e-commerce market disrupted the already large traditional retail
market, whose revenues grew from $1.2 trillion in 2005 to $2.5 trillion in 2020, a growth rate of 5
percent. Meanwhile, e-commerce’s revenues grew from $15 billion in 2005 to $890 billion in 2020, a
31 percent growth rate. E-commerce’s share of total retail sales in the United States was 2.5 percent
in 2005. By 2020, it was 15 percent.
Consumer electronics. The evolution of mobile phones into smartphones was a technology step
change in this arena that resulted from escalatory investments in R&D aimed at integrating multiple
new technologies—including semiconductors, digital displays, cameras, batteries, and software
operating systems—into one compelling solution.
The shift to smartphones also expanded mobile phones’ value beyond simple communication tools
to multipurpose devices, creating software ecosystems. The emergence of these software platforms
created a business model step change: the arrival of application stores and application revenues.
Apple launched its application store in 2008, and Google introduced its own in 2012. To acquire
customers for their platforms and build brand loyalty, players escalated investments in marketing
(for example, Samsung spent $3 billion, or 3.8 percent of total revenues, on marketing in 2005, and
$9.7 billion, or 4.8 percent of revenues, in 2020).
Players also escalated R&D investments intended to constantly integrate software and hardware
innovations. Investments in underlying operating systems (iOS and Android, Mac OS and Windows)
also escalated as players sought to accelerate network-effect dynamics for developers and users:
mobile operating systems capabilities attract more developers to develop applications, and more
applications attract more users. This creates an incentive for developers to produce apps for
operating systems with more users. This arena accessed the large and growing demand for mobile
devices. In 2005, only 34 percent of the global population had any kind of mobile phone; by 2020,
67 percent of the global population had a smartphone. In addition, the global population grew
20 percent over this period.
Semiconductors. As discussed above, this industry experienced constant step changes of
technology S-curves in the form of node resets. For semiconductor manufacturers, including
foundries and integrated design and manufacturing players, investments escalated to stay on
the cutting edge of each successive node. Tool suppliers also escalated R&D to create tools to
manufacture smaller and smaller features; for example, extreme ultraviolet lithography required
billions of dollars over decades. Fabless players that designed semiconductor products escalated
R&D investments in the same ways as software companies: constant innovation was required to
create new products and enhance existing ones to remain competitive. Semiconductors’ revenues in
our sample nearly tripled from $194 billion in 2005 to $574 billion in 2020, driven by rising demand
for computing systems (PCs and servers) and mobile devices.

Arena-
Video and audio entertainment. This arena’s business model step change was the development
of on-demand streaming of video and audio, which offered increased accessibility of content
unbounded by time or physical constraints, disrupting the way traditional media was consumed. To
attract more customers, players escalated investments in content development, such as licensing
material and in-house production. With scale, players also had more purchasing power to negotiate
terms with content creation ecosystems, whether licensing content or producing it in-house. Further,
R&D investments escalated to improve the customer experience, for instance by leveraging customer
data—using recommendation algorithms, for example—to drive customer satisfaction and retention.
Global consumption of media grew over this period, with digital entertainment as a driving force. This
arena’s revenues more than tripled, from $135 billion in 2005 to $407 billion by 2020.
Consumer internet. The global rise of the internet and broader digitization created the opportunity
to serve consumer needs through services like social media and web search. Though membership-
based internet access providers such as AOL were still common in 2005, consumer internet players
began to implement a step change in their business model. This change enabled these services to
rapidly acquire large user bases that were monetized through advertising and sold through online
auctions. Consumer internet players were also able to offer marketers a new business model of ads
that link fees to clicks rather than impressions. Another business model innovation was the new
ability to hyper-target customer segments that traditional platforms like print publications, cable
TV, and billboards could not achieve. In terms of investments, players escalated marketing spending
to accelerate user acquisition and enable network effects. More users and more collected data
increased the value of platforms by helping search engines refine results. For social media players,
each additional user increased the value of the network for other users. For search engines and social
media platforms, network effects amplified this virtuous cycle of consumers using players’ platforms.
Biopharma. The introduction of drug discovery and production advancements, such as
recombinant DNA technology, genome sequencing, and other platform technologies, presented
new developmental paths for drugs that provide more targeted and effective therapies for patients.
The escalation of investments in R&D was driven to a large extent by a need to discover and develop
drugs that either were more effective than existing therapies or treated additional indications. For
example, therapeutic-area-focused biotech companies continually invested to take advantage of
technological curves, including AI-driven bioinformatics. Many of these companies had higher ROI
and clinical success rates than traditional pharmaceutical businesses’ R&D and in-house discovery
engines. Due to the long runway of R&D-to-market cycles in pharmaceuticals, one could argue
that this competitive dynamic may have begun as early as the 1990s, with revenue growth from
biopharmaceuticals becoming materially observable from 2005 to 2020.
Global pharmaceuticals revenues were $601 billion in 2005. That figure more than doubled to reach
$1.3 trillion by 2020, with biopharmaceutical businesses’ share of the pool growing from an estimated
20 percent in 2005 to 50 percent by 2020, indicating that the lines have already blurred between
biopharmaceuticals and pharmaceuticals. Biopharma revenues grew 13 percent year-on-year from
2005 to 2020.
Software. The emergence of software as a service (SaaS) caused step changes in business models
in this arena. For example, in 2013, Microsoft launched Office 365 (after a beta that started in 2010),
and Adobe transitioned its Creative Suite into the Creative Cloud. Players escalated investments in
R&D (for example, collecting usability data or enhancing features) to retain customers and launch
new products or features to drive lifetime customer value. This trajectory was followed by both
customer relationship management software players, such as Salesforce, and enterprise resource
planning software players, such as Oracle and SAP. Salesforce spent $9.8 million on R&D in its
2005 fiscal year. By 2020, that had grown to $2.8 billion. Oracle invested $1.5 billion in R&D in 2005

Arena-
and $6.1 billion in 2020. SAP invested €1.1 billion in R&D in 2005, which increased to €4.5 billion by
2020. As with other arenas, the digitization wave spurred growth: the number of software use cases
increased, computing power available in the cloud multiplied, and higher-performance internet
connectivity made SaaS feasible for more customers. Software revenues grew 12 percent year-on-
year, from $64 billion in 2005 to $341 billion in 2020.
Information-enabled business services. The emergence of outsourcing and offshoring services,
including business process outsourcing and outsourcing IT and knowledge services, was a major
business model step change in this arena. In addition, the increasing sophistication of data analytics
capabilities, enhanced computing power, improved communications and networking technology,
and digitization encouraged players to pivot to IT asset-based service delivery to supplement
talent. Investments through acquisitions, primarily for the addition of new capabilities, talent, and
data, escalated in this arena. For example, ADP completed 28 acquisitions from 2005 to 2020,
and the credit-scoring agencies Equifax, Experian, and Transunion collectively acquired about
150 companies in that time.25 The expansion of this arena was driven by the proliferation of the
knowledge economy that followed in the wake of globalization and by the growing need to integrate
technology into global businesses. Our sample shows that revenues in this arena almost tripled, from
$53 billion in 2005 to $154 billion by 2020 at a 7 percent year-on-year growth rate.
Payments. Digitization caused a step change in payment systems by enabling them to be faster,
more convenient, and often cheaper than offline payment solutions. As with e-commerce, players
escalated investments in marketing to acquire merchants and customers, which created mutually
reinforcing network effects. There was also escalation in R&D to enhance product features such
as better security, fraud detection, and customer experience that enhanced the value proposition
for merchants and customers. The payments industry was already large and growing, with global
revenues that rose from $493 billion in 2005 to $1.6 trillion in 2020, representing growth of 8 percent
year-on-year. The majority of the growth was from emerging markets, such as China. In 2020, more
than half of global payments revenues were from the Asia–Pacific region.
Cloud services. Cloud services exemplify a business model step change, because the companies
offered computing and storage as operating expenses, freeing customers from rigid capital
expenditures based on purchasing hardware. This business model provided flexible IT infrastructure
solutions on a pay-for-what-you-use basis. The widespread deployment and penetration of the
internet helped hasten the adoption of cloud services. Cloud-services players escalated investments
in capital spending to build more data centers. They also improved economics of scale and price
competitiveness and offered efficiencies, such as proximity and energy efficiency, to their customers.
There was some R&D investment escalation as these players offered more value-added services,
including advanced security options and analytics, as well as APIs for developers to ease integration
with customers’ systems. This arena disrupted the pool of corporate spending in physical IT
infrastructure: global IT spending mostly came from traditional on-premises solutions in 2005, but by
2020, cloud-services spending accounted for $370 billion, or more than a third of the $1.1 trillion in
global outlays.26
EVs. Rapidly improving battery and electric power train technology, as well as the ability to
integrate these new technologies into one product, provided the technology step changes that
led to an increasingly competitive market for EVs. The R&D investments for batteries, power train
technology, and other production processes escalated to lower costs and increase performance.
R&D investments also went toward improving other automotive technologies, including the
ability to download software to vehicles, similar to the way application stores operate for mobile
phones. Escalating outlays for marketing to convey competitive value to customers supplemented
these investments.

Arena-
The EV arena cut into the revenue pool of traditional ICE vehicles. That industry had $1.7 trillion
in revenues in 2005, which grew to $2.4 trillion by 2020. By then, EVs had become a $100 billion
revenue arena, equivalent to almost 5 percent of the established traditional automotive industry.
While other important industries stood out from 2005 to 2020, they did not show the outsize growth
and dynamism that would qualify them as arenas. Our potion can help us understand why.
The apparel and luxury industry showed a relatively high industry share growth rate during this
period—comparable to the biopharma and payments arenas—but displayed limited dynamism among
the largest companies. This industry lacked a technology or business model step change of the same
degree as the 12 arenas of today, one of the key ingredients of the potion. This meant that the same
small group of players could vie for the growing demand pool in mature markets competition.
Conversely, oil and gas showed high dynamism with large share shifts among players while the
industry share growth rate in this sector was far lower than the average in our overall sample: the
market caps of oil and gas players grew at a 4 percent CAGR from 2005 to 2020, compared with
average growth of 16 percent in our arenas. The industry experienced observable technology
step changes, such as the advent of horizontal drilling and hydraulic fracturing in shale and other
geological formations. However, oil and gas did not display an escalation mechanism for investments,
another key potion element, and the huge capital outlays in the sector mostly go toward exogenous
sunk costs, such as exploration and drilling of wells. This meant that growth was more linked to price
changes from global events affecting supply and demand—such as the 2007–08 financial crisis, the
COVID-19 pandemic, and the rapid proliferation of North American shale—as well as non-economic
factors, than to an escalation of investments made by players.
§ § §
Arenas do not start from scratch. While early signals may appear noisy, they can still be helpful
in helping us identify the primordial ooze from which arenas emerge. The arena-creation potion
described in this chapter and evident in the arenas of today can provide an initial guide for identifying
potential arenas of tomorrow. We can begin by looking for early signs of S-curve step changes, the
starting gun of a new escalatory competitive race. For this race to eventually lead to the creation of an
arena, the three ingredients usually must be in place.
In chapter 3, we look at the potential arenas of tomorrow and how they might evolve. Given the
dynamic nature of arena competition, our ability to speak with certainty about which arenas will
emerge is limited. But using our potion and the theoretical framework we set out in this chapter, we
can identify certain aspects of arenas and present potential scenarios for their size and competitive
dynamics in the coming years.

Arena-
CHAPTER THREE
The arenas of tomorrow
The arenas of today have brought sweeping change to our lives over the past two decades. The
arenas of tomorrow are likely to bring even greater change in the years ahead.
In chapters 1 and 2, we looked at the past and explored how the arenas of today grew and evolved,
propelled especially by digital technologies. In this chapter, we look toward the future to identify
18 potential arenas of tomorrow that could exhibit high growth and high dynamism (see sidebar
“How we defined the arenas of the future”). These arenas have the potential to transform our
world in a number of areas, including transportation, healthcare, digital connectivity, energy,
and entertainment.
In 2022, the industries we identified as the potential arenas of tomorrow took in $7.2 trillion in
combined annual revenues. In our modeled scenarios, their revenues could grow to between
$29 trillion and $48 trillion by 2040, at a CAGR of 8 to 11 percent from 2022 to 2040. For reference,
the revenues of our arenas of today listed in chapter 1 climbed at a CAGR of 10 percent from 2005 to
2020, while revenue in other industries increased at a 4 percent CAGR. When we apply the typical
after-tax profit margins observed in these arenas, our estimates show that potential future arenas
could generate $1.9 trillion to $6.1 trillion in profits by 2040 (Exhibit 14).27
The revenue of arenas could

represent a growing share of
global GDP: from 4 percent in
2022 to 10 to 16 percent by 2040.

Arena-
Exhibit 14
The 18 potential arenas of tomorrow could generate $29 trillion to

$48 trillion in revenues and $2 trillion to $6 trillion in profits.
18 potential arenas Revenue, 2040 estimate (CAGR, Profit, 2040

of tomorrow, by 2040 2022 2022–40, %) estimate,¹ $ billion
revenue estimate, $ billion (profit margin, %)
14,000–
E-commerce 4,000 20,000 280–
(7–9) 1,000
(2–5)
1,500–4,600 230–920
AI software and services 85
(17–25) (15–20)
1,600–3,400 160–510
Cloud services 220
(12–17) (10–15)
2,500–3,200 100–320
Electric vehicles 450
(10–12) (4–10)
2,100–2,900 320–580
Digital advertisements 520
(8–10) (15–20)
1,700–2,400 340–600
Semiconductors 630
(6–8) (20–25)
20–460
Shared autonomous vehicles n/a 610–2,300
(4–20)
Space 300 960–1,600 (7–10) 50–160 (5–10)
Cybersecurity 160 590–1,200 (8–12) 90–240 (15–20)
Batteries 98 810–1,100 (12–14) 40–110 (5–10)
Modular construction 180 540–1,100 (6–10) 20–220 (4–20)
Streaming video 160 510–1,000 (6–11) 50–150 (10–15)
Video games 230 550–910 (5–8) 80–180 (15–20)
Robotics 21 190–910 (13–23) 20–180 (10–20)
Industrial and consumer biotech 140 340–900 (5–11) 10–270 (4–30)

Future air mobility n/a 75–340 10–70 (10–20)
Drugs for obesity and related conditions 24 120–280 (9–15) 30–100 (25–35)
Nuclear fission power plants 18 65–150 (7–13) 5–50 (5–30)
Total 7,250+ 29,000–48,000 (8–11) 1,900–6,100
¹Defined as net operating profit less adjusted taxes (NOPLAT). NOPLAT share based on most closely mappable industries from our database of 3,000 companies
analyzed in chapters 1 and 2.
Source: Company annual reports; McKinsey Value Intelligence; McKinsey Global Institute analysis
McKinsey & Company

Arena-
While the revenue and after-tax profit margin estimates broadly followed the same ranking as shown
in Exhibit 14, the orders do deviate slightly due to business model differences. For example, if current
profit margins persist, the semiconductor arena would generate profit margins of 20 to 25 percent,
making it the third-largest arena by potential 2040 profit (when ranked by the higher range of
estimates) though it is only the sixth-largest arena by 2040 revenue. By contrast, the EV arena, which
ranks fourth by 2040 revenue (also ranked by the higher range of estimates), would rank seventh by
potential 2040 profit if the industry’s profit margins continue at 4 to 10 percent.
We also estimate that the revenue of arenas, converted into GDP terms, could represent a growing
share of global GDP: from 4 percent in 2022 to 10 to 16 percent by 2040.28 This translates to a 18
to 34 percent share of total GDP growth (Exhibit 15). This shift in GDP share and value creation is a
hallmark of arenas. If we perform a similar analysis on our arenas of today, the revenues of our sample
companies also increased their corresponding share of GDP, from 3 percent in 2005 to 9 percent
in 2020.29 Capital flows also indicate the appeal of the arenas of tomorrow for investors. By our
estimates, about a third of venture capital flows from 2020 to 2023 went to industries linked to our
potential future arenas.30
With an internet penetration rate of

just 35 percent, developing economies
have large untapped potential for
digital products and services.

Arena-
Exhibit 15
The 18 potential arenas of tomorrow could contribute a third of global

GDP growth.
Global GDP, $ trillion

Arenas of tomorrow Other industries
Lower scenario, 2022–40 Higher scenario, 2022–40
168 168
44
54
23
101 12 101
141
151
97 97
4 4 27
16
2022 2040 2022 2040
Arenas’
share of
increase,
18 34
%
Note: Figures may not sum to 100%, because of rounding.

Source: OECD; Bureau of Economic Analysis; McKinsey Value Intelligence; McKinsey Global Institute analysis
McKinsey & Company
In chapters 1 and 2, we showed how digitization was the main underlying force in the formation of
the arena-creation potion for most of today’s 12 arenas. For our list of 18 potential future arenas,
we identified four thematic “mega-forces” that will likely drive their future growth and dynamism:
continued digitization, innovations in the physical world, the global energy transition, and a growing,
more prosperous world.
First, the continually improving capabilities of digital technology could accelerate growth when
the product is information or a digital interface, driven by companies’ continuing efforts to move
workloads to the cloud, the growing complexity of cybersecurity needs, and the rapidly expanding
capabilities of AI. The continued expansion of the internet in developing economies may also drive
growth for arenas like e-commerce and digital advertising. For instance, internet penetration is
estimated at 35 percent in developing economies today, compared with 80 percent in developed
economies—a disparity that illustrates the large untapped potential for digital products and services
in developing economies.31 Second, innovations are bringing new technologies to the physical world

Arena-
with advancements such as robotics and shared autonomous vehicles. Third, the global energy
transition and the underlying momentum to decarbonize are likely to boost demand for cleaner
energy and products, including EVs, batteries, and nuclear fission power plants. Fourth, as the global
population grows and becomes more prosperous, demand will grow for industries that address
quality-of-life needs in areas ranging from healthcare to homes.
We have identified three groups of arenas, differentiated by the way they become potential arenas
of tomorrow (Exhibit 16). The first group, continuing arenas, is made up of four of the 12 arenas of
today that we have identified as potential future arenas because they will likely continue to exhibit
high growth and high dynamism. The second group, spin-off arenas, consists of specific portions of
current arenas that are likely to have their own trajectories of high growth and high dynamism. The
third group, emergent arenas, is made up of new industries that have high potential for growth and
dynamism and have exhibited early signs that the arena-creation potion elements are coalescing.
Together, these three groups of arenas make up our 18 potential future arenas. Other industries
were also considered, such as those we designated as “ceasing” arenas and a set of industries
that almost qualified as emergent arenas. We chose not to analyze these potential arenas in depth
because of uncertainties about size and dynamism, the likelihood of the scenario, and the time
frame for scaling.

Arena-
Exhibit 16
The 18 potential arenas of tomorrow include arenas of today, spin-off

arenas, and emergent arenas.
How certain industries might emerge as arenas of tomorrow

Arenas of today Subsegments of arenas of today Candidate arenas of tomorrow
4 arenas of today
Cloud Electric Semi-
E-commerce could continue
services vehicles conductors
as arenas of tomorrow.
Consumer Video/audio
Software
internet entertainment 3 spin-offs
of arenas of today
could become
Digital arenas of tomorrow.
AI software Streaming
advertise-
and services video
ments
Drugs Future
Batteries Cybersecurity
for obesity air mobility
Industrial Nuclear 11 emergent arenas

Modular could exhibit high growth
and consumer fission power
construction and high dynamism
biotech plants
through 2040.
Shared
Video
Robotics autonomous Space
games
vehicles
Why other industries considered might not emerge as arenas of tomorrow

5 arenas of today could lack the growth 8 almost-emergent arenas might show high growth and dynamism
and dynamism through 2040 to through 2040 but were not analyzed because of uncertainties about
continue as arenas of tomorrow. the likelihood of the scenario and the time frame for scaling.
Biopharma Clean hydrogen Renewable energy generation
Consumer electronics equipment
Lower-carbon materials
Industrial electronics Sustainable fuels
Nuclear fusion
Virtual reality and augmented reality
Information-enabled business services Products and services
for older adults Web3, including decentralized finance
Payments
McKinsey & Company

Arena-
The following consists of thumbnail

sketches of the industries that could
become the 18 arenas of tomorrow,
along with our assessment of the factors
influencing their growth and dynamism.
We also outline swing factors, potential
future developments that could greatly
affect the growth or dynamism of the
arena. The compendium that follows this
chapter describes these aspects of the
arenas in more depth.

The 18 potential arenas of tomorrow
Below and on the following pages are thumbnail sketches of 18 industries that could become the arenas of tomorrow,
arranged in descending order of 2040 high-case revenue scenario. Each includes a very brief summary of the
growth factors that could drive the expansion of each industry’s revenues, the dynamism factors that could
generate shifts in market shares between players within these industries, and the swing factors that could push
an industry towards the higher or lower end of the range of scenarios.
E-commerce AI software and services Cloud services

Companies that sell goods through Companies that provide software and Companies that deliver on-demand
digital channels and fulfill them services incorporating AI, excluding cloud infrastructure and platforms as
directly hardware necessary to operate AI a service
Growth factors Growth factors Growth factors

New formats, such as social and Analytical AI and generative AI Continued migration from
quick commerce, expansion of enabling enterprise use cases and on-premises computing and storage
categories in developed economies, boosting individual productivity to the cloud, renewed demand from
and higher penetration in developing computational requirements of newer
Dynamism factors
economies technologies like AI
Massive investment and scale
Dynamism factors allowing a few large players to Dynamism factors
Maintaining a few large players develop among frontier foundation Likely to continue to have a few large
with some disruption from direct- models; fragmentation in the players, with possible shifts due to
to-consumer companies, growth in segment providing specialized emerging regional competitors and
the grocery segment, and regional software for targeted use cases a growing segment of cloud services
players for AI
Swing factors
Swing factors Semiconductor supply and Swing factors
Social commerce, adoption of AI , computational power, market Data sovereignty regulations, gen AI
innovation of physical retail, pick- fragmentation due to geopolitical adoption, cloud migration costs,
and-deliver economics developments, evolution of value responses to a cybersecurity event
distribution among value chain, and
$14 trillion– impact of open-source models $1.6 trillion–
$20 trillion $3.4 trillion
Revenue in 2040 $1.5 trillion– Revenue in 2040
$4.6 trillion
Revenue in 2040

Arena-
Electric vehicles Digital advertisements Semiconductors

Manufacturers of battery, plug-in Platforms that enable advertisers to Designers and manufacturers
hybrid, and fuel-cell electric vehicles reach consumers digitally of semiconductors, microchips,
and integrated circuits, as well as
Growth factors Growth factors providers of tools for semiconductor
Improving range and value for money, Growing middle class, more manufacturing
consumer willingness to pay for consumer time on digital media, new
sustainable products, and a sufficient ad products and placements Growth factors
supply and distribution of electricity Rising demand for computing
Dynamism factors
to EV charging infrastructure and data storage, wireless
Degree of globalization of
communication, and industrial
Dynamism factors walled gardens, new ad formats,
electronics; increasing number of
Existing large global OEMs with convergence of commerce and
semiconductors used in automotive
new EV offerings competing with advertising, regulatory and data
new EV-focused entrants, as well privacy actions, shifting consumer Dynamism factors
as increasing competition from eyeballs Some share-shifting possible
Chinese-based EV manufacturers as players move from general-
Swing factors
purpose to domain-specific chip
Swing factors Interoperability between platforms,
manufacturing or as tech players
Ramp-up of sufficient charging impact of gen AI, competition in new
vertically integrate into design or
infrastructure affecting consumer media formats, price and margin
manufacturing
preference, regulatory support of impacts of new ad inventory
EV adoption, raw material needs, Swing factors
sufficient generation of clean $2.1 trillion– Geopolitical dynamics, competition
electricity $2.9 trillion in AI-specific chips, slowdown in
Revenue in 2040 investments, price erosion, slowdown
$2.5 trillion– in Moore’s law
$3.2 trillion
Revenue in 2040 $1.7 trillion–
$2.4 trillion
Revenue in 2040

Arena-
Shared autonomous vehicles Space Cybersecurity

Operators of shared autonomous Providers of outer-space-related Companies that provide protection
vehicle services infrastructure and services to the of computer systems from
commercial and state-sponsored unintended and unauthorized
Growth factors segments access, modification, or destruction
Technological progress, financial
feasibility, consumer acceptance Growth factors Growth factors
Growing demand in commercial Improving capabilities of
Dynamism factors
services and end-user equipment, attackers, growing data sets, new
Nascent but likely to have a few large
lower cost of satellites, introduction business models vulnerable to
players given high R&D costs and
of reusable heavy rocket launchers, attacks, cybersecurity regulatory
barriers to entry
increased government spending requirements
Swing factors
Dynamism factors Dynamism factors
Regulatory framework for liability,
Commercial infrastructure and state- High fragmentation due to multiple
consumer safety concerns,
sponsored defense and intelligence segments and local regulations, but
overcoming of technological hurdles,
players see higher barriers to entry could have fewer players if platforms
different business models
due to R&D requirements; higher emerge
potential for entrants in end-user
$610 billion– Swing factors
equipment and services and state-
$2.3 trillion Emergence of new cybersecurity
sponsored civil segments
Revenue in 2040 technologies, adoption of AI,
Swing factors breakthroughs in quantum
Evolving consumer demand in the computing, market fragmentation
commercial segment, geopolitics caused by geopolitical developments
affecting global market structure,
proliferation of space use cases $590 billion–
$1.2 trillion
$960 billion– Revenue in 2040
$1.6 trillion
Revenue in 2040

Arena-
Batteries Modular construction Streaming video

Manufacturers of rechargeable Companies that operate in the Providers of on-demand video
batteries used for EVs and other modular construction value chain entertainment over the internet
technologies that are mostly linked to from design to assembly, with
the energy transition volumetric modules Growth factors
Rising number of streaming
Growth factors Growth factors households, especially in developing
Improving energy density, price Partnering across the value chain, economies; increasing spending per
reduction as production capacity selection of projects with consistent household; increasing ad revenues
scales, increasing demand from EVs demand, successful on-site from streaming
and battery energy storage systems execution
Dynamism factors
Dynamism factors Dynamism factors Benefits from scale and
A few large players benefiting from Possibly a few large players if first verticalization, moderated by regional
advantages of scale, concentration movers achieve scale; regional and players and new entrants benefiting
of production in Asia–Pacific, policy vertical fragmentation likely to persist from gen AI to create content
incentives outside Asia to foster local given nature of construction industry,
Swing factors
battery supply chains possible entry of large development
Evolution and role of user-generated
companies
Swing factors content, adoption of advertiser-
Innovations in battery technology, Swing factors supported offerings, future platform
lower prices, localization of Ability to overcome industry inertia bundles, relative success of
manufacturing, changes in regulation subscription and ad-revenue-based
$540 billion– models
$810 billion– $1.1 trillion
$1.1 trillion Revenue in 2040 $510 billion–
Revenue in 2040 $1.0 trillion
Revenue in 2040

Arena-
Video games Robotics Industrial and

Producers and distributors of games Manufacturers of robots and consumer biotech
played on dedicated consoles, PCs, providers of robotics solutions Providers of biotechnology-
and mobile devices enabled products in agriculture,
Growth factors alternative proteins, biomaterials
Growth factors Continuing development of robotic and biochemicals, and consumer
Growing game-playing population technology to automate physical products markets
and spending as a result of capabilities, increasing economic
increasing global digitalization, feasibility of solutions that drives Growth factors
growth of free-to-play games and adoption of general-purpose and Commitments to decarbonization,
microtransactions, user-generated trainable robotics, pace of mass new tech such as AI that expedites
content, cloud gaming, and adoption R&D, cost and quality of replacement
increasing advertising revenue product, efficient commercialization
Dynamism factors
to scale from lab to mass production,
Dynamism factors Fragmentation in nascent specialized
public investments, consumer
A few large players in the console and autonomous general-purpose
interest
and PC market, along with vertical robotics industries, as well as a few
integration of console manufacturers players in traditional industrial market Dynamism factors
into game publishing; high High fragmentation given nascency,
Swing factors
fragmentation in the mobile-game including small science-focused
New technologies that accelerate
market start-ups and large incumbents;
productivity, new use cases beyond
a few large players may emerge if
Swing factors current physical limitations of
contenders scale and invest in R&D;
Mobile and casual gaming growth, people, ability of robots to perform
fragmentation to remain if smaller
internet expansion into emerging nonphysical labor such as caregiving,
players make easy breakthroughs
economies, and next-generation market fragmentation due to
experiences, such as augmented and geopolitical developments Swing factors
virtual reality Pace of commercialization,
$190 billion– competition with conventional
$550 billion– $910 billion consumer products, public concerns
$910 billion Revenue in 2040 about privacy and genetic testing
Revenue in 2040
$340 billion–
$900 billion
Revenue in 2040

Arena-
Future air mobility Drugs for obesity and Nuclear fission power plants
Operators of air mobility transport related conditions Players that construct nuclear fission
services, such as eVTOLs and Companies that sell GLP-1s and power generation facilities
delivery drones other drug therapies for obesity and
related conditions, such as diabetes Growth factors
Growth factors Potential to lower construction
Improving battery and propulsion Growth factors costs through new technologies
technologies, competitive pricing, Growing incidence of obesity (25% like small modular reactors,
regulatory and infrastructure by 2035), price reductions due to demand for baseload power to
support, addressing customer increasing competition and reduced augment intermittent solar and
concerns over safety and price manufacturing costs wind generation, continued global
commitments to decarbonization
Dynamism factors Dynamism factors
Scaling of passenger eVTOLs may A few large players likely to emerge Dynamism factors
attract new entrants, but economies from first movers, with potential A few large players given barriers
of scale and customer acquisition fragmentation as other branded to entry such as high R&D
may restrict market to a few major competitors and generics producers requirements; regional fragmentation
players; a few major drone players crop up driven by geopolitical scenarios;
likely to emerge regionally or globally potential for new entrants if they can
Swing factors
develop and scale next-generation
Swing factors Potential for new innovations to
technology
Regulatory framework to enable treat other chronic diseases, global
operations at scale, sustainability of recognition of obesity as a disease, Swing factors
business model, public attitudes on convenience and tolerability of drug Global energy demand shifts, price
safety and convenience delivery of nuclear power compared to
alternatives, public perception of
$75 billion– $120 billion– safety, government support, pace of
$340 billion $280 billion commercialization of next-generation
Revenue in 2040 Revenue in 2040 technology
$65 billion–
$150 billion
Revenue in 2040

Arena-
Below we show how the potion ingredients—the key indicators that an arena may be forming—are
already apparent in many cases, though depending on the arena’s stage of development some may
be harder to discern than others. The 18 arenas, sorted into continuing, spin-off, and emergent
groups and ordered by the upper end of potential revenue ranges for 2040, are as follows:32
Continuing arenas
Four of our 12 current arenas—e-commerce, electric vehicles, cloud services, and semiconductors—
are likely to become arenas of tomorrow. All four are in the middle phase of rapid growth in their
S-curve of market adoption. Demand pools still have large headroom and could produce this growth,
which is driven by the mega-forces of innovation and digitization.
E-commerce. This arena is continuing on the rapid phase of its S-curve of growth and is still
anchored in the original business model step change, which was the online availability of retail sales.
Nevertheless, other relevant business model resets could also spur growth in e-commerce, for
example retail media networks, which are adjacent to the reset in another future arena, digital ads.33
From mid-2020 to mid-2022, more than a dozen retailers debuted this new advertising revenue
stream. Social commerce, which allows consumers to make purchases on social media apps, is
another business model reset on the rise. In addition, e-commerce is expected to undergo more
business model step changes by expanding further into large product categories such as healthcare
and food.
Because e-commerce is a continuing arena, the escalatory investments in merchant and customer
acquisition are likely to continue. Capital expenditures for last-mile delivery capabilities and
investments in social media and payments integration are also likely to continue to escalate. A larger
share of sales in developing markets and expansion into new product categories in developed
markets are likely to sustain demand and drive the majority of e-commerce growth, effectively
stretching this arena’s S-curve.
Electric vehicles. EVs, an arena of today, are likely to continue to be an arena for two reasons:
continued growth as EVs capture an increasing share of the large and stable demand for passenger
and commercial vehicles, and the new rising competition in the industry. The breakthrough of battery
cell technology and its current scalability and commercialization is the technology step change in
this continuing arena. R&D investments could continue to escalate as players, especially traditional
automotive incumbents, increase their share of EV sales. EV sales could eventually take the lion’s
share of overall passenger vehicle sales—in our estimates, from 18 percent in 2023 to between 82
and 96 percent by 2040.
Cloud services. The ongoing adoption of cloud services, a continuing arena of today, could generate
$3 trillion of EBITDA increases for Forbes Global 2000 companies by 2030.34 AI cloud represents
a technological step change in this arena, because new AI models require significant amounts of
computational power, which could be supported by cloud infrastructure. Furthermore, the types of
infrastructure and platform services, including AI-enabled capabilities, that cloud providers offer
will likely continue to evolve. New data sovereignty regulations, such as requirements about where
data centers must be located, are also generating a step change in business models, fostering the
emergence of regional competitors.

Arena-
The escalation of capital expenditures is likely to continue as in past decades, and R&D expenditures
could also escalate as players try to develop more efficient cloud infrastructure. We estimate that
the share of global IT spending on cloud services could grow from 8 percent in 2022 to between 19
and 41 percent by 2040 as companies continue to migrate their computing and storage processes to
the cloud.
Semiconductors. This current arena undergoes frequent technology step changes as
manufacturers keep reducing the distance between chips’ transistors and shrinking the size of
components, including the transistors themselves. Players are likely to continue to escalate R&D
and capital expenditures for machinery in pursuit of node resets, but also to meet rising demand
for computational power. For example, the need for high-performance chips could keep growing
as cloud services companies double down on investments in graphics processing units and AI
accelerators for computationally intensive tasks.
Demand from four segments could drive the majority of the growth in semiconductors: computing
and data storage, automotive, wireless communications, and industrial electronics. The requirements
of AI and cloud services are likely to boost computing and data storage demand; driving assistance
features, electric vehicles, and infotainment system components could propel automotive demand;
the growing need for connected devices would increase wireless communications demand; and
industries such as medical technology, automation, and electricity generation are likely to fuel
industrial electronics demand.
Sidebar: How we defined the growing portion of a large industry to analyze demand may shift between them over the
arenas of the future it on its own. coming decades.
For example, we considered three future The batteries arena has several
As we explore a new set of arenas, it is worth arenas that could have some overlap, subsegments, including batteries for
revisiting the principles that guided our particularly on the demand side: digital EVs, battery energy storage systems, and
choice of the level of granularity we used to ads, streaming video, and video games. batteries for consumer electronics. While
define them. Just as we did with our selection Each has large and fast-growing markets, each of these categories could have been
of arenas in chapters 1 and 2, we used supply- distinct business models, and well-defined analyzed as an independent arena, we
and-demand considerations to determine the competitive landscapes. Some of these looked to our supply-and-demand principles
granularity level for the arenas of tomorrow. future arenas are subsets of current arenas to help identify the reasonable level of
The supply perspective determined whether (for example, the digital ads arena grew granularity. On the supply side, all three
the players and the competitive landscape out of the consumer internet arena). While subsegments have similar core technologies
were relatively well defined, with similarities they could collectively be defined from the and participating players. And on the demand
in underlying technology or business models demand side as competing in the “attention side, despite seemingly different end-use
where step changes can spark escalatory economy,” these arenas’ products are related cases there is a similar need for stored
competition, while the demand perspective replacements but not direct substitutes for energy. These characteristics are among the
ensured that products and services were one another. From the supply side, the sets reasons we chose to define batteries as a
sufficiently similar or substitutable. In some of companies that participate in each of the single arena.
cases, defining arenas as broader industries three arenas are generally distinct. For this
was sufficient; in other cases, identifying reason, we analyzed them as three separate
these arenas involved splitting off a fast- arenas, while recognizing that overall

Arena-
Spin-off arenas
Three arenas of today—consumer internet, software, and video and audio entertainment—have
subsegments that may grow large enough and fast enough to become new arenas. The consumer
internet arena of today could spin off a digital advertisement arena; the software arena of today
could generate an AI software and services arena; and the video and audio entertainment arena of
today could spawn a streaming video arena. From an S-curve perspective, a specific technological or
business model step change helped separate these subsegments, now potential arenas of tomorrow,
into their own new S-curves. For example, growth in the broader software industry is likely to slow
as it approaches the tail end of its S-curve of adoption and innovation. AI, however, is disrupting the
industry and leading to a new S-curve. This new S-curve is specific to AI software and services and
could demonstrate its own explosive growth in the next few years, as is typical in the early stages of
an S-curve.
AI software and services. Although some types of AI, such as machine learning, have been around
for decades, the technology rose to public prominence in late 2022 with the release of OpenAI’s
ChatGPT, a gen AI interface capable of producing remarkably humanlike responses to users’
prompts. The recent emergence of gen AI is the technological step change in this market that
supports our definition of AI software and services as its own arena, making it a spin-off of software
as an arena of today. Cloud services, a continuing arena, is also critical for AI software and services,
because it enables most of the computational processing required by today’s AI models. And there
are signs of momentum for adoption that indicate the beginnings of a new S-curve: in an April 2023
McKinsey survey, nearly a third of respondents said their organizations were using gen AI in at least
one business function, and 40 percent said their organizations would increase AI investments.35 In
March 2024, two-thirds of the survey’s respondents reported that their organizations were using gen
AI, doubling the level of adoption in less than one year.36
Players are escalating R&D investments in this arena as they attempt to differentiate their services
through advanced capabilities. AI systems, such as GPT-4 by OpenAI and Gemini by Google, that
compete in the “frontier” foundation models segment—that is, the biggest, most advanced models—
escalate investments in proprietary data and infrastructure to create industry-leading models.
Players that compete in the fragmented specialized AI software segment (such as AI start-ups that
have specific use cases) escalate investments by creating bespoke AI-enabled systems tailored
to their use cases. This arena’s potential economic impact and the associated revenue could be
significant as a result of a large market for both consumer as well as corporate use cases, such as
consumer research and segmentation, software engineering, and operations.
Digital advertising. More people online, a rising middle class (just ten countries could add 900
million people to the middle class by 2030), and more time spent on digital media (in the United
States, from 40 percent in 2016 to 66 percent in 2026, according to forecasts)37 are continuing to
increase the demand pool for digital advertising, which funds much of the internet through search,
social, and media. Players are continuously experimenting with business models to attract users.
For example, media players are trying out different ad-supported tiers, as well as the formats of
advertisements, such as in-line ads in social media posts or ads displayed as part of a gen AI search
query result.38 Technology in the industry also is moving quickly, with new players introducing new
experiences such as short-form social videos or building capabilities to hyper-target customers
and track the cost-effectiveness of advertisements, and increasing the use of gen AI for hyper-
personalization and to lower the cost of content creation.
Investments escalate as players attempt to sustain a leading edge in an environment with strong
network effects. Players are locked in a battle for attention across platforms and must continuously
invest to create the most engaging experience, biggest social network, or best media content to keep
people—and the advertisers that want to reach them—on their platforms. Players must also invest

Arena-
to capture new pockets of growth, especially in developing economies, where regional competitors
can emerge to challenge global leaders. We expect digital ads to capture a growing share of global ad
spending, from 65 percent in 2022 to between 80 and 90 percent by 2040.
Streaming video. The video and audio entertainment arena could be transformed because the
streaming technology developed in the past 15 years is being adopted more rapidly, resulting in a new
spin-off arena: streaming video. The uptake of streaming video is still in the steep part of its S-curve.
In the past, investments for customer acquisition and content production escalated as competition
intensified between players. However, companies now face pressure to increase profitability. As
a consequence, investments will be made in partnerships and other collaborative models, such as
content bundling. The two-way network effect in this arena will likely persist: more content attracts
viewers, and more viewers give platforms more leverage to license content or revenues to produce
their own.
The market for streaming video may also be buoyed by demand from increasing high-speed internet
access in developing economies as more households around the world gain the ability to stream
video online. This larger customer base would also increase the potential revenues earned from
advertising on these platforms.
Emergent arenas
The remaining 11 potential arenas of tomorrow, the emergent arenas, are generally novel relative to
existing industries. In other words, they are not clear subsegments, though some of their enabling
technologies may be similar to those that enable other industries or arenas. These potential arenas
vary in degree of maturity, a natural consequence of using 2022 as the starting point of our analysis.
Some of these potential arenas, such as video games, have existed for decades, while others, such
as drugs for obesity and related conditions, are in very early stages of formation. For all of them,
however, there is high growth and high dynamism potential because these arenas are in the early to
middle phases of their S-curves.
Shared autonomous vehicles (SAVs). The development of technology for vehicle autonomy
(integration of computer vision, real-time machine learning, and large-scale data engineering, such
as high-resolution mapping) is the step change in this arena, where R&D investments escalate
primarily to overcome technological obstacles that affect the cars’ safety and reliability. Despite the
increased prevalence of vehicles for hire that don’t require a human driver, a recent decline in SAV
investment suggests that commercialization may go more slowly than forecast. The large market
for SAVs could depend on regulation, technological progress, financial feasibility, and consumer
adoption of the vehicles.39 We estimate that SAVs could capture 25 to 51 percent of the shared
mobility industry’s revenues by 2040.
Space. The commercial sector of the space arena has shown robust growth and accelerating market
activity in the past decade, with recent players, such as SpaceX, Blue Origin, and Virgin Galactic,
entering the fray. Government interest has also evolved: the number of government space agencies
around the world has grown from 40 in 2000 to more than 75 today.40 This arena experiences
frequent technological step changes, especially in the form of commercially viable spacecraft and
new satellite systems, like SpaceX’s Starlink system. Innovations in launch mechanisms, such as
reusable boosters, are also under way.
Escalation in R&D has occurred in both the public and private sectors. Major players continue to
double down on investments for new product launches. Private-sector spending in space is reaching
all-time highs, with more than $70 billion invested in 2021 and 2022 combined.41

Arena-
Growth in the space arena may be driven by state-sponsored investments in the civil and defense
segments; consumer needs for space capabilities, such as communications and navigation; and
commercial infrastructure and support operations required to meet these state-sponsored and
consumer demands.
Cybersecurity. In 2020, the global direct financial damage from cybercrime was estimated at about
$950 billion, almost twice the $520 billion recorded in 2018. When considering direct, indirect, and
upstream systemic costs, the overall economic impact of cybercrime in 2020 was estimated to reach
approximately $4 trillion to $6 trillion, or more than 4 percent of global GDP and more than four times
the direct costs.42 As adversarial parties keep improving their capabilities, selling ransomware-
as-a-service, and creating marketplaces for hacking tools and data, players in this arena need
to continuously improve their offerings to protect computer systems from unauthorized access
by adversaries. The cybersecurity technological step change is the data protection and network
security that complements other fast-growing technologies with new vulnerabilities.
As a result, R&D investments escalate for the development of more effective products and services.
This often includes investing in proprietary data sets and processing power, which are frequently
needed for AI use cases. Innovation in the industry has been expanding, too, with the number of
patents growing at rates of 10 to 25 percent from 2017 to 2021. As attacks continue to increase and
the digital landscape grows and changes, cybersecurity spending is expected to represent a larger
share of global IT spending, from 6 percent in 2022 to between 7 and 14 percent by 2040.
Batteries. The global energy transition is fueling demand for batteries, mainly due to the continued
growth of EVs, battery energy storage systems, and consumer electronics. Manufacturers of
batteries have made technological step changes related to energy density, charging speed, longer
life cycles, and manufacturing sustainability.
R&D is likely to escalate as manufacturers innovate on battery technologies, such as enriching the
anode with silicon compounds. Manufacturers also are innovating in new designs such as solid-
state and new chemistries such as lithium-sulfur and vanadium redox, which could become direct
substitutes for lithium-ion and sodium-ion batteries in specific applications. Capital expenditures
could escalate as players try to capture the benefits of manufacturing at scale, as seen in the rise of
gigafactories. As a result, barriers to entry could rise for new entrants, improving the margins of large
incumbents.
Video games. The rise of mobile and cloud gaming, which allow games to be delivered faster through
new platforms, served as a technology step change in this arena.43 This coincides with a business
model step change, with free-to-play games like Fortnite aiming for wider distribution and to
generate revenue through in-game purchases. R&D investments, especially in game development,
have escalated in this arena. For example, several high-profile AAA games are released each year.44
These games have budgets of at least $200 million, more than the average production budget of the
150 most expensive movies from 2016 to 2023, $180 million.
The surge in mobile gaming and the success of the free-to-play in-game-purchase business model
on various platforms drive this large and growing market. Growing consumer spending on games,
as well as higher advertising revenue and console sales, are expected to increase revenues despite
recent setbacks in certain markets, such as below-expectations revenue growth and consumer
spending that returned to prepandemic levels.
Robotics. Advancements in the autonomous-robot segment represent a major technological reset
by helping to improve the mobility and dexterity of the machines and increase the range of tasks they
can perform easily. One way to understand this arena is through robot types, from single-purpose
robots typically used in industrial or manufacturing environments to general-purpose robots,
which are more recent and can execute a variety of tasks with limited human intervention. Boston

Arena-
Dynamics, Tesla, Figure, Google, and Sanctuary are among the players trying to develop general-
purpose robots to take on tasks currently performed by humans.
R&D investments may escalate, especially in the subsectors of general-purpose robots and
specialized robots for specific use cases, such as cooking or packing boxes. Players are intensifying
an already heavy R&D focus on deep technology as they vie for share. Acquisitions and investments
in robotics companies are also escalating. The robotics arena received a growing influx of capital,
with venture capital investment increasing from about $4.6 billion in 2018 to about $13 billion in 2022.
Industrial and consumer biotechnology. The technology step change in this arena is attributable
to breakthrough biological innovations, such as the decline in the cost of reading genetic codes and
the improvement in the capability to edit genes using CRISPR technology. As a result, application
of this technology may accelerate in four areas: agricultural biotechnology (such as bioengineered
crops), alternative proteins (such as cell-based meats), consumer products and services (such as
personalized wellness products), and biomaterials and biochemicals (such as bioplastics).
R&D and capital expenditures account for the majority of the escalatory investments for product
development, commercialization, and manufacturing needs. The large and growing market in
the industrial and commercial biotechnology arena may be driven by several factors, including
advancements in technology facilitating R&D, increasing consumer demand, the commercialization
of new products and services, and regulatory and consumer commitments to sustainability.
Modular construction. These are builders in the housing market that produce standardized,
prefabricated modules made at an off-site factory and then assembled into buildings. This method
can improve construction productivity amid lagging labor productivity growth in the gigantic
$13 trillion global construction industry. Improving construction productivity is urgent to address
an ongoing global housing shortage and affordability crisis, as well as a critical skilled labor gap in
the construction industry.
The modular, prefabricated process that enables the parallelization of construction phases is the
business model reset in this arena and could greatly improve construction productivity. While some
modular construction is widespread in a few high labor-cost countries like Japan, the Nordics, and
Singapore, global penetration is still relatively low.
However, there are signs that broader adoption—the rising part of the S-curve—could be near. There
is increasing investment and research into processes and digital tools that can help players solve the
complex value chain coordination challenges of modular construction. And growth in modular has
the potential for a flywheel effect. As more buildings are built using modular techniques, the supply
of modular increases, which in turn makes it more likely that other developers could choose modular
as well.
Nuclear fission power plants. Today, nuclear fission reactors are the second-largest source of
electricity generated with low emissions of greenhouse gases, after hydropower. In this arena, we
look at players that build nuclear power generation facilities. The next generation of small modular
reactors is an example of a technological step change because the reactors are designed to be both
safer than traditional large-scale reactors and cheaper to build by using prefabricated components
and standardized designs. As more countries build nuclear power plants and the global market
grows, learning curve efficiencies could push down this capital expenditure for both large-scale
reactors and SMRs.
R&D investments could escalate in the race to reduce building costs and develop new reactor
technologies, construction methods, and coolant types. Global venture investment funders seem
optimistic about this arena despite the high capital costs and the long time horizon before returns
are realized. Funding rose from about $60 million in 2018 to about $390 million in 2022. To reach

Arena-
emissions-reduction targets, countries would likely need to rely on nuclear power to complement
intermittent renewable energy sources, such as wind and solar. China, Russia, South Korea, and other
countries plan to significantly scale up nuclear power production. At the COP28 climate conference
in 2023, 25 countries pledged to triple their nuclear energy capacity by 2050 to reduce emissions.
Future air mobility. Innovations in materials, propulsion, batteries, and autonomous technology
for passenger electric vertical takeoff and landing vehicles (eVTOLs) and delivery drones are the
technological step changes in this arena. R&D investments in these components could continue
to increase as air mobility technologies begin to reach commercialization. The need for “vertiport”
boarding and landing facilities, as well as delivery distribution centers, means capital expenditures may
also increase. While many of these nascent technologies are yet to be commercially available, there are
early signs of future growth: disclosed annual funding for this arena grew from less than $600 million
in 2017 to $5 billion in 2023 as companies progressed on their regulatory and certification journeys
and aimed to tap the large addressable markets for new modes of transportation.
Drugs for obesity and related conditions. A report by the World Obesity Atlas estimates that
obesity prevalence could rise from 14 percent in 2020 to 24 percent by 2035, based on current
trajectories, with an attributable economic impact of $2.0 trillion in 2020 and $4.3 trillion in 2035.45
The technology step change in this arena was the introduction of GLP-1 agonists, a revolutionary
class of drugs that were initially approved to treat diabetes and, more recently, obesity as well.
An escalatory dynamic is evident in the R&D race, as major players develop their own GLP-1 weight-
loss therapies. With generics on the horizon, players are incentivized to further invest in R&D for more
efficacious drugs with fewer adverse effects or more convenient drug administration like an oral solid
pill, rather than the current injectables.
Other candidate arenas

In determining potential future arenas, we considered new industries that could eventually become
arenas but are less likely than some other industries to achieve the high growth and high dynamism
of our arenas of tomorrow. We discuss them briefly below, with an emphasis on swing factors that
could make them arenas.46 We consider two other groups of industries: the ceasing arenas—arenas
of today that are unlikely to persist as arenas of tomorrow—and the almost emergent arenas, which
have uncertain growth or dynamism prospects.
Ceasing arenas
Five arenas of today—biopharmaceuticals, consumer electronics, information-enabled business
services, industrial electronics, and payments—are likely to cease to be arenas. This does not mean
these industries will not see high growth or high dynamism; they remain critical to today’s economy.
However, growth or dynamism may not be at the same scale as in the past 15 years. These arenas
are hitting the plateauing phase of their S-curves. Many are likely to continue to benefit from a
mega-force like payments taking advantage of digitization opportunities in emerging markets, or
biopharmaceuticals serving a more prosperous world. However, the growth of these current arenas is
likely to be more stable than explosive.
Biopharmaceuticals. The distinction between biopharmaceuticals and their traditional
pharmaceutical counterparts is becoming irrelevant. As it is, players are already incorporating both
into their product lines as effective drugs for targeted treatments, combination therapies, or both. As
a result, biopharmaceuticals, which have been disruptive in the past, are already mainstream. There is
also a weak case for vulnerability or obsolescence of the traditional pharmaceuticals industry.

Arena-
Both biopharmaceuticals and traditional pharmaceuticals could see breakthroughs that spur the high
growth and dynamism required of an arena beyond those achieved in the treatment of obesity and
related conditions (already a potential future arena). These advancements could occur in oncology
research, neuroscience, immunology, or gene therapy, among other fields.
Consumer electronics. Established players, such as Apple and Samsung, are likely to drive growth
in this industry over the next few years as demand remains robust. However, growth may not be of
the same magnitude as before. Some estimates indicate that the consumer electronics industry
could grow at a 3 percent rate from 2024 to 2028, a lower pace than the 7 percent observed in this
arena from 2005 to 2020.47 This slower growth could result from plateauing of smartphone adoption
globally. It is also possible that market leaders could maintain intensifying and stable market share
competition over the next few years, making the dynamism prospects of this arena uncertain.
Information-enabled business services. Most of the growth from knowledge and business process
outsourcing as a result of globalization and digitization has likely already been captured, and greater
stability is expected as large incumbents buttress their market positions.
Industrial electronics. This arena has already captured large pockets of growth, including in China.
While growth may reach about 6 percent through 2030, it most likely will not be at the same pace as
before, a slowdown that could be similar to the likely gradual plateauing of the consumer electronics
arena.48
Payments. For established payment use cases, such as high-frequency and small-ticket
transactions, the expansion of payments brought about by digitization was mostly captured in the
past decade. That gave current players a strong hold on market shares but will likely result in this
arena’s reduced growth or dynamism prospects compared with the past two decades. However,
opportunity for some growth remains. Players could enter the market and innovate solutions for
low-frequency but large-size transactions, such as tuition and business services like home repair.
Regional players could emerge in developing markets, which could lead to new growth.
Almost-emergent arenas
These candidate arenas were not analyzed as deeply as our 18 arenas of tomorrow described
above. Some of these potential arenas, such as Web3, were not included primarily because of high
uncertainty about their growth or dynamism prospects. Others have a more certain rising trajectory
but a lower probability of evolving into arenas compared with the 18 we analyzed. Nonetheless,
we have chosen to discuss them below because each has exciting potential and we recognize the
possibility that these industries could still become arenas—we want to leave plenty of room for the
eventuality that things could turn out differently from expectations. These candidate arenas are as
follows, ordered alphabetically:
Clean hydrogen. This industry consists of companies that provide energy through low-carbon
hydrogen, including from natural gas, known as blue hydrogen, as well as renewable hydrogen
from electrolysis powered by renewables, known as green hydrogen. The current use cases, mostly
for industrial processes like oil refining, required approximately 90 million tons of hydrogen in
2022.49 The share of clean hydrogen in current hydrogen consumption is negligible, at less than
one million tons.
The new industry of producing low-emission hydrogen is already encountering regulatory tailwinds.
For example, in 2023, the EU Hydrogen Bank, backed by a $3.3 billion investment from the EU
Innovation Fund, held its first EU-wide auction for renewable hydrogen producers that wished to
receive support. The auction attracted 132 bids from 17 countries.50 In February 2024, Germany
allocated €3.5 billion in subsidies to support green hydrogen exporters through H2Global.51 The
US Department of Energy has announced a $7 billion investment to launch seven regional clean
hydrogen hubs.52

Arena-
Existing demand for industrial use cases is expected to be stable, increasing from about 86 metric
tons in 2020 to 111 metric tons in 2040, but new pockets of demand will likely come from new
industrial use cases such as green steel or grid-level storage. However, there is considerable
uncertainty about the uptake of hydrogen in industrial processes or emerging use cases. This
uncertainty stems from factors including an evolving regulatory climate, the development of enabling
technology and specialized infrastructure, competition dynamics with other decarbonization
technologies, and potential adoption barriers like distribution and delivery costs.
Lower-carbon materials. This industry consists of manufacturers of low-carbon materials, of which
steel, cement, and aluminum are most relevant for decarbonization (low-carbon bioplastics are
discussed in the industrial and consumer biotechnology compendium entry). These three materials
have varying starting points and face their own challenges as decarbonization continues. The
industry could undergo a big shift enabled by potential technological breakthroughs to decarbonize
the manufacturing process achieved by the use of electrification, circularity, carbon capture
storage, or low-emission fuels such as biomass. Today, about $6 billion of revenues are generated
from the sale of lower-carbon (“green”) materials. This figure could grow to hundreds of billions by
2040. Green steel will likely be the largest component of this growth, contributing about half of the
revenues. Projections are highly sensitive to future price fluctuations in the materials industry.
The lower-carbon materials industry has the potential for high dynamism. The acceleration of steel
decarbonization may entail a reconfiguration of the value chain as energy-intensive production
steps such as direct iron reduction locate closer to sources of clean energy, a requirement for the
decarbonization process. While this would imply large investments, it also opens up opportunities as
more geographies become attractive as locations for new plants. The massive energy requirements
of green steel manufacturing may require new capacity in regions with lower energy costs like Brazil,
the Middle East and North Africa, and countries like Spain.53 Many, if not all, incumbent steel players
have decarbonization projects under development or in implementation. Meanwhile, new players like
H2 Green Steel have started making moves aimed at building a large-scale green steel plant that
would start production as early as March 2026.54
However, green steel’s growth is uncertain for two reasons. First, global demand for steel may
be reduced by an unevenly distributed slowdown across regions and industries over the next
few years. The shifts that could cause this uncertainty include the “normalization” of demand in
China, which may be partially offset by growth in Southeast Asia and India, and the slowdown in
construction demand, which may be offset by growth in energy and transportation.55 Second, there
is uncertainty about how rapidly green steel could replace traditional steel. This is mostly due to the
competitiveness of green steel manufacturing costs compared with traditional steel. The cost could
be affected by several factors, including regulatory support, capital investments, manufacturing
technology advancements, raw energy costs, and value chain reconfigurations. Despite this
industry’s potential high dynamism and large shifts, the growth uncertainty diminishes the likelihood
that low-carbon materials could reach arena status.
Products and services for older adults. Companies in these industries offer services to meet the
needs of adults aged 65 and older. By 2050, there could be 1.6 billion people in this demographic.
In 2022, there were, on average, seven working people for every person over 65 (also known as
the old-age dependency ratio). By 2040, that ratio could shrink to four to one, creating a shortage
of caregivers.
Today, an estimated $13 trillion is being spent on and by older people, with housing, health, and
wellness the biggest drivers. Traditional players participate in select parts of elder care services,
including nursing homes that provide housing and meet care needs, and financial institutions that
provide retirement and pension management and other services. Only a few of the players we

Arena-
considered provide holistic, seamless services. At the same time, older adults’ needs like health,
connection, autonomy, and security are not always fully met in this rapidly changing environment.
Some early platform plays are surfacing today: JD.com is investing in a care ecosystem for
older people through partnerships, telehealth product launches, and healthcare management
products. Clariane Group, a nursing care provider, acquired 32 companies in 2019 and 2020 to
create a digital ecosystem for home care professionals and care facility residents.56 A truly digital
and comprehensive ecosystem by 2040 would require technological step changes, such as the
automation of caregiving, and coordinated innovation among tech players, property developers,
healthcare providers, and financial services players. Sufficient early signs of these breakthroughs
and innovations have yet to be observed at a level that could ignite an arena.
Aside from these technological uncertainties, it is also likely that the industry will remain fragmented
if global or even national plays are discouraged by limited economies of scale, market regulations,
cultural differences, and the way services are delivered. For these reasons, even though this industry
is likely to see major growth as populations age, there may be limited dynamism as players remain
small and local, making products and services for older adults less likely to be an arena of tomorrow.
Nuclear fusion. This industry consists of companies that build plants able to generate electricity
through nuclear fusion. More than 130 experimental fusion devices are operating, under
construction, or planned around the world today.57 While the technology is still in the experimental
phase, fusion has the potential to become the world’s zero-emissions power source, promising
cheap and abundant clean energy that could transform the world.
Yet even though the technology is promising, uncertainty about the rate of technological progression
is the primary swing factor determining whether nuclear fusion could become an arena. It is unlikely
that a commercially scalable and sizable industry could be formed by 2040, even in the best-case
scenario of a major scientific breakthrough as early as 2030. Several technical challenges must
be overcome before a fusion reactor can become commercially viable: magnet performance,
survivability of the first wall, plasma stability in confinement concepts other than tokamaks, and
reliable heating systems.58 Massive R&D investments, supply chain advancements, talent sourcing,
and partnerships to develop regulation would all be needed and would take a long time to scale up.
For these reasons, nuclear fusion has a relatively low probability of sufficient growth to become an
arena by 2040.
Renewables generation equipment and infrastructure. These are companies that manufacture
equipment for renewable energy generation (wind and solar), storage for renewable energy, and
transmission and distribution infrastructure. The potential growth in this industry is driven by the
continued push for decarbonization. Many governments have committed to net-zero targets, and
several technological advancements in renewable generation in the past decade drove production
and cost efficiencies. Solar and wind power have both become cost competitive with fossil-
fuel-based power sources in the past decade.59 Global investment in renewable energy rose to
$358 billion in the first half of 2023, an all-time high for a six-month period.
This industry is likely to see higher growth in the earlier phase of our time horizon, until 2030,
but more modest growth from 2030 to 2040. Three compounding swing factors may change
this estimate. The first is the pace of energy consumption growth, which could be influenced by
population growth, industrial needs, and EV adoption, among other factors. Second is the ability of
the renewables sector to capture a material share of this growth, which would require simultaneous
investments in steps from generation to distribution. Third, uncertainty about the pace of technology
advances and affordability translates to uncertainty about the total investment and timing of the peak
expenditure in this sector.

Arena-
While renewables generation is likely to take an increasing share of global electricity production, the
industry’s growth trajectory to 2040 is uncertain, especially if peak expenditure occurs before then
and growth from new installs slows. The industry could also see low dynamism as high barriers to
entry favor incumbents. These factors make renewables generation equipment and infrastructure an
almost emergent arena with a promising trajectory but also relatively lower likelihood of becoming an
arena of tomorrow.
Sustainable fuels. This industry consists of companies that use sources other than fossils to make
liquid fuels for transportation, primarily for cars, trucks, aviation, and marine. Over their lifetimes—
that is, from when they are produced to when they are burned in vehicles—sustainable fuels typically
emit fewer greenhouse gases than fossil fuels.
There are three kinds of sustainable fuel. First are conventional biofuels, made from raw materials
that are edible, such as sugars and starches derived from corn or sugarcane, or from oils derived
from sunflower or rapeseed. These mostly alcohol- and ester-based fuels need to be blended with
traditional fossil fuels to be used in a typical internal combustion engine, which is why they are called
“blend-in” fuels. Second are advanced biofuels made from raw materials that are not used for food,
including agricultural residue, dedicated crops, and waste. These fuels, called “drop-in” fuels, are
chemically similar to gasoline and can be used without blending with traditional fossil fuels. The third
kind is e-fuels, made by breaking down water into hydrogen and oxygen, typically by combining the
hydrogen with carbon.
There is potential for sustainable fuels to grow to arena status by 2040 if the industry captures a
large enough share of the fuels market. However, great uncertainty surrounds the industry’s growth
prospects because of the challenges that would need to be overcome to reach scale, such as
customers’ willingness to pay a green premium for sustainable fuels, and how much the trajectory
of electrified mobility will affect global fuel demand. Further, degrees of fragmentation vary among
players in the three segments of sustainable fuels. For example, drop-in fuels are likely to have a few
large producers, whereas the supplier base for blend-in fuels and e-fuels is highly fragmented. These
uncertainties make it hard to assert that sustainable fuels are a likely arena of tomorrow, though they
have that potential.
Virtual reality and augmented reality (VR and AR). This industry consists of players that
manufacture the hardware required for VR and AR, and players that provide the software operated
on this hardware. VR and AR companies enable customers to immerse themselves in virtual worlds
where users are “in” the experience rather than merely observing it. The VR and AR industry had
estimated revenues of $22 billion in 2022 from hardware and software sales.60 About $30 billion in
venture capital flows went to VR and AR from 2020 to 2023.
VR and AR can be applied for both consumer and commercial use cases. For consumers, the
metaverse is an example of the extended application of VR and AR (although the metaverse can also
be accessed through conventional interfaces, such as phones and laptops). Corporate use cases
range from healthcare applications, such as 3D body mapping, to retail, such as trying on clothing
and experimenting with furniture placement in a room, to education and training, such as VR-assisted
instruction for machinery-intensive jobs.
While the potential technological step changes in this industry are clear in both the innovative
hardware and the accompanying software that enable the VR and AR experience, it is uncertain
whether the escalatory dynamic in investments and addressable market are sufficient to develop the
growth and dynamism features required of an arena by 2040. Two mutually reinforcing swing factors
may determine the future of this industry. First, there will likely need to be substantial step changes
in the technology, particularly in lighter-weight headsets that do not have to be tethered for power
and do not cause nausea in a substantial set of users. Second, the industry would require widespread

Arena-
adoption of the technology and software platforms for new use cases. If the technology performs well
enough to encourage adoption, that could spur more investments to improve the technology. If these
two factors create a virtuous cycle, the VR and AR industry could be an arena by 2040.
Web3, including decentralized finance. Web3 is the idea of a new, decentralized internet built
on blockchains, which are distributed ledgers controlled communally by participants. Web3 has
the potential to change how information is managed, how the internet is monetized, and even how
web-based corporations function. Most of the potential growth could come from financial use cases
related to tokenization, trading, payments, decentralized finance, and custody. However, some areas
of Web3 have seen downturns as others continue to soar. For example, the crypto market’s market
capitalization fell more than 50 percent in 2022, while there was a 68 percent increase in nonfungible
token sales in the same year.61 Technology adoption also seems to have advanced: as of 2023, there
were thousands of decentralized applications running, compared with about 1,000 in 2018.62
This is a nascent industry that is shifting rapidly and faces several obstacles, mainly related to the
need for clarity in regulations regarding asset classification, tax reporting, and know your customer
(KYC) standards. However, there has already been significant regulatory activity. For instance, the
US Congress proposed more than 50 crypto regulations in 2022.63 Technical obstacles, including
privacy concerns, have arisen because blockchains are public by design, which limits potential
use cases. Other obstacles include the risk of fraud, high transaction costs, and services that are
sensitive to failure.64
Web3 is still in its infancy and could transform the way we execute transactions, run applications, and
own assets. However, there are major uncertainties on the path toward adoption that could influence
the development of the ecosystem and which players could be active in it. As a result of this great
uncertainty, it is hard to estimate whether Web3 would display the revenue growth and competitive
dynamism to qualify as an arena of 2040.
Geopolitical disruption, the

development of AI, and the pace
of the green transition could
be critical swing factors.

Arena-
The significance of potential future arenas

In chapters 1 and 2 we demonstrated that the 12 arenas of today are not only centers of value creation
but also industries that materially change the way we live. Our 18 arenas of tomorrow could be even
more transformative, as they could shape how we consume and process data, approach health and
wellness, and interact and communicate with one another. Innovations in areas such as robotics,
mobility, AI, and clean energy may play a growing part in expanding the choices available to us in our
daily lives.
We may soon be able to take an eVTOL from our local vertiport instead of ordering an Uber. We may
have AI virtual assistants to help us finish writing emails or plan our vacations. We may have the
option to commute to work in a shared ride in an autonomous shuttle instead of driving our own car.
Homeowners may be able to store renewable electricity in their EV batteries and use those batteries
to buy electricity when prices are low and to sell it back to the grid when prices are high. A wide
variety of robots may come to assist us with physical labor, speeding up task times in agriculture
or construction.
But also, looking back to 2005, we recognize cause for humility. No one could have predicted all of
the ways in which the arenas of today have evolved. Therefore, we also highlight some critical swing
factors that go beyond uncertainties in the modeling and are more fundamental to the evolution of
the arenas of tomorrow in technology, investment patterns, and sources of demand. Many of these
factors are specific to individual arenas and are described in the compendium at the end of this
report, but looking across arenas, we see the following three themes:65
Developments in geopolitics: Geopolitical events affect many arenas. Swing factors include the
possibility that technology stacks could diverge across regions, with different groups of countries
pursuing technologies in parallel, such as in AI if government actions create multiple geographically
distinct ecosystems. And technology players could be affected by geopolitical changes that create
turbulence in trade flows and markets. It also remains to be seen whether the internet could become
more regional, and the impact that could have on cybersecurity, cloud services, and other industries.
While geopolitical developments could affect the growth of many arenas, they will have an even
greater impact on dynamism by defining who the players are and where they are allowed to play.
Geopolitical splintering and disruption would be particularly relevant to the following arenas:
cybersecurity, robotics, AI, cloud services, space, EVs, batteries, semiconductors, e-commerce, and
digital advertisements.
Development and adoption of AI technology: AI software and services is an arena, and the evolution
of artificial intelligence, especially gen AI, could fundamentally change the shape of several other
arenas over the coming decades. New capabilities enabled by gen AI could lower the cost of
creative content used in digital ads, create storylines for immersive video games, give robots lifelike
personalities, or change the way consumers find products and services on search platforms. But
much will depend on the pace of AI adoption, how quickly use cases come online, and the value of the
productivity gains from using the technology.
In addition, the adoption of automation could be sharply accelerated. Without gen AI, automation
could take over tasks accounting for 22 percent of the hours worked in the US economy by 2030,
according to our estimates. With gen AI, that share could jump to 30 percent.66 The trajectory of AI
development could affect all 18 arenas but is especially relevant to video games, streaming video,
digital ads, cloud services, cybersecurity, e-commerce, semiconductors, robotics, industrial and
consumer biotechnology, and SAVs. The technology still faces technological hurdles before full
deployment, including gen AI’s factual accuracy and safety concerns over privacy and misuse.

Arena-
Pace of the green transition: Global efforts to alter the course of climate change by reducing CO2
emissions67 could drive demand in many of the 18 potential arenas of tomorrow (and determine
whether as many as five of the eight almost emergent arenas become arenas). Often, their potential
revenue size depends substantially on how much of a priority governments, consumers, and
businesses place on carbon emissions reductions and how much they’re willing to pay to abate them
in cost, time, or convenience. For example, if countries remain committed to clean energy targets,
they will likely require some nuclear fission power production to balance dips in solar and wind
generation. If emission reductions become a lower political priority, countries could have a much
broader range of options to meet their energy needs. The robustness of decarbonization targets
and the commitments people and governments make to achieve them will likely have a major impact
on the future market size of EVs, batteries, nuclear fission power plants, industrial and consumer
biotechnology, and modular construction.
As the direction of these swing factors becomes clear, so too will the trajectory of our growth and
dynamism scenarios. Our research provides an initial view of where to expect the most growth and
dynamism and how to update that view as the future takes shape.
Analyzing the trajectories of the potential 18 arenas of tomorrow surfaces critical questions, from the
ethics underpinning data and privacy to imperatives for businesses to be inclusive and sustainable.
While arenas’ precise trajectories are difficult to predict, economic and social disruptions are natural
consequences as these industries form, grow, and impact the world. Recognizing how and when
arenas originate, understanding how they evolve, and anticipating the way they impact the world can
offer a unique view of the arc of society’s progress.
To give more insight into the economics of arenas, we present a compendium of the 18 future arenas
and discuss in detail how high growth and high dynamism might play out in each. These descriptions
should not be read as comprehensive accounts. Instead, they explain a rationale for a range of
scenarios for growth and dynamism over the coming decades and describe some swing factors that
could alter the outcomes.

Arena-
Arenas of tomorrow
compendium
1. E-commerce 7. Shared autonomous vehicles 13. Video games
2. Artificial intelligence software and services 8. Space 14. Robotics
3. Cloud services 9. Cybersecurity 15. Industrial and consumer biotechnology
4. Electric vehicles 10. Batteries 16. Future air mobility
5. Digital advertisements 11. Modular construction 17. Drugs for obesity and related conditions
6. Semiconductors 12. Streaming video 18. Nuclear fission power plants

Arena-
Contributing author:
Roberto Longo 1. E-commerce
The e-commerce industry consists of companies that sell goods to consumers through digital
channels. This industry is already well established, its rise fueled by the spread of broadband
internet access, increasing ways to access the internet (especially smartphones), and innovations in
product delivery, including on-demand warehousing, last-mile delivery, increased automation, and
crowdsourced delivery methods.68
E-commerce is one of today’s arenas, as we describe in chapter 1, and there are many reasons to
believe that it could be one of tomorrow’s, too. In 2022, the retail e-commerce segment accounted
for 20 percent of a $17 trillion overall global retail market, leaving plenty of room for further expansion
in both geography and product categories. Online markets in China, Latin America, and other
developing economies are likely to continue to grow as digitization increases. And e-commerce is
expected to continue to expand into new major product categories, such as personal care and food.
Social commerce, or making purchases on social media apps, is also on the rise.69
Growth
We define e-commerce as completing purchases on computers and mobile devices, including selling
and buying items that consumers pick up at brick-and-mortar stores. This definition encompasses
revenue from manufacturers with direct-to-consumer (DTC) channels, platform providers such as
Amazon that have marketplaces for third-party retailers, third-party retailers (known as merchants)
when they sell on platforms such as Amazon, and traditional retailers with online channels. This
excludes several industry-adjacent markets, including peer-to-peer or consumer-to-consumer (C2C)
sales, such as those available on Facebook Marketplace, and ride-hailing and streaming services.
Our overall sizing of the industry has two main segments: retail e-commerce and food e-commerce
(shipping costs excluded). The retail e-commerce segment includes categories of consumer goods
such as general merchandise, apparel, and appliances.70 Together, the revenues of the two segments
reached roughly $4 trillion in 2022. That could grow to between $14 trillion and $20 trillion by 2040, a
CAGR of 7 to 9 percent.71
The retail e-commerce segment currently accounts for one-fifth of global retail revenues (both online
and offline), and that share could reach 27 to 38 percent by 2040. In our estimates, revenues could grow
from $3.4 trillion in 2022 to $11 trillion by 2040 in the lower range of scenarios and to $16 trillion in the
higher range of scenarios. Meanwhile, the food e-commerce segment brings in $630 billion in revenues
today, or 4 percent of global retail revenues. That could grow to $3 trillion by 2040 in the lower range of
scenarios and to $4 trillion in the higher range of scenarios, or 7 to 9 percent of global retail revenues.
Retail e-commerce could come to represent a larger overall share of total retail sales in developing
economies, where it is now less prevalent than in developed economies. For example, in 2022, its
share of retail sales was just 12 percent in Latin America, 4 percent in the Middle East, and 2 percent
Note: This section describes
the potential growth and in Africa. The lower range of scenarios assumes that retail e-commerce could account for 26 percent
dynamism of the e-commerce of retail revenues in Latin America by 2040, 15 percent in the Middle East, and 10 percent in Africa.
arena. It should not be read
as a comprehensive account In the higher range of scenarios, Latin America’s share increases to 29 percent, the Middle East’s to
of the industry. To learn more 20 percent, and Africa’s to 15 percent.
about e-commerce and
associated industries, please
E-commerce in those markets could grow with the expansion of the middle class, increased access
refer to content from the
McKinsey Retail; Technology, to broadband, and the development of infrastructure and services, such as last-mile delivery,
Media & Telecommunications;
and Growth, Marketing &
that enable sales. For example, smartphone penetration in India grew from 26 percent in 2018
Sales Practices. to 36 percent in 2022 and is expected to reach 56 percent by 2027. In sub-Saharan Africa, the

Arena-
smartphone penetration rate increased from 19 percent in 2018 to 36 percent in 2022 and is
expected to hit 48 percent by 2027. This growth added 400 million smartphone users in these two
regions from 2018 to 2022 and could account for an additional 550 million users globally by 2027.
Smartphone penetration was much higher in developed economies such as North America, reaching
86 percent in 2023. Some connectivity growth is likely to come from broadband: from 2022 to 2040,
penetration in India could rise from 10 percent to 21 percent, and from 5 to 9 percent in sub-Saharan
Africa.72 These trends suggest potential for improving connectivity in developing regions.
Developed markets could see more e-commerce, too. In North America, retail e-commerce held a
25 percent share of total retail spending in 2022. The Asia–Pacific region’s share was 22 percent, and
Europe’s was 16 percent. The lower range of scenarios estimates that retail e-commerce would increase
to a 30 percent share in North America, to 28 percent in the Asia–Pacific region, and to 27 percent in
Europe by 2040. The latter two markets lag behind North American penetration rates by six and eight
years, respectively. In the higher range of scenarios, retail e-commerce penetration could grow to
50 percent in North America by 2040, to 40 percent in Asia–Pacific, and to 36 percent in Europe.
Most of the growth in developed markets would result from e-commerce’s expansion into new
categories. Consumers around the world are already accessing twice as many industries online as
they did before the COVID-19 pandemic, and there are at least 25 million “high potential” customers in
the United States and Europe who tried e-commerce in new categories during the pandemic but have
not fully adopted it.73 The food segment had the fastest growth from 2017 to 2022, when spending
increased 34 percent a year. In some estimates, the segment’s share of all e-commerce could grow
from 16 percent in 2022 to about 20 percent in 2040.74 More emotional goods such as apparel,
home goods, skincare, jewelry, premium tea, and arts and crafts are also sold online through social
commerce and the storytelling this approach enables (see below). In the long run, that could be a
significant driver of growth for the industry, even in developed markets. The lower range of scenarios
assumes limited category expansion, while the higher range assumes significant expansion.
The Asia–Pacific region has a mix of developing and developed economies with diverse e-commerce
dynamics. The penetration rate of online shopping in developed economies varied. In 2023, it was
30 percent in South Korea and 15 percent in Japan.75 It also varied in developing economies, at 36
percent in China and 32 percent in Indonesia. The rate was just 8 percent in India in 2023, but that
was double the 4 percent rate in 2018, suggesting momentum.
New e-commerce formats are also driving growth. Social commerce platforms use online media to
boost user and influencer interaction, promoting online sales of products and services by merging
social content with shopping to enhance informed purchasing. In 2023, the top 35 leading social
sellers recorded more than $24 billion (170 billion reminbi) in sales in China via live sessions across
platforms. The sales from all live sessions represented more than 30 percent of total e-commerce
sales in China that year.76
Discount and low-price platforms are also showing success. Online marketplaces such as Pinduoduo
are growing fast in China through their value commerce platforms that offer affordable pricing and a
game-like shopping experience. This platform model also gained popularity globally as companies
such as Temu and SHEIN leveraged their strong supply chains in China, quickly scaling in the United
States and Europe.
Additionally, conversational commerce, which enables sales through messaging platforms such as
WhatsApp, is increasingly becoming integral to the shopping experience. The rise of generative AI
(gen AI), the widespread use of chat apps, and the surge in online shopping are propelling conversational
commerce, potentially boosting its market size to $41 billion by 2030. And quick commerce promises
delivery to customers in hours or even minutes, a model that is particularly suited to relatively small
and frequent purchases of merchandise such as convenience store items and groceries.77

Arena-
Dynamism
In 2023, the three largest e-commerce platform providers, Amazon, Alibaba, and JD.com, accounted
for 15 percent of the revenue share of all companies that sell online, including manufacturers with
DTC channels, platform providers, and traditional retailers with an online channel.78 However, the
retail value of the merchandise—the total value of goods sold by merchants on the platforms, rather
than the revenue of those platforms—flowing through these three platforms represented up to
42 percent of total e-commerce retail value in 2023.79 This large share is attributable to several
factors: the big investments that e-commerce companies need to make in delivery logistics, the
companies’ economies of scale, and the mutually reinforcing network effects of a two-sided
marketplace. Customers find more value in a platform with many merchants, and vice versa.
However, emerging contenders are disrupting the status quo, rapidly scaling through innovative
strategies. At least four key elements pose challenges to the incumbents, as follows:
First, DTC sales could capture a larger share of the industry’s revenues. Companies such as Casper
and Warby Parker sell to customers without a retail middleman. These sales are supported by third-
party logistics providers such as Shopify that help smaller companies establish and maintain online
retail systems (including those built into social media platforms).
Second, the growing popularity of online food purchases—the grocery segment was worth $8 trillion
in annual global revenues in 2023—could open the field to emerging contenders if they can succeed
in food e-commerce.80 However, the large e-commerce incumbents are also competing in food, as
Amazon did by buying Whole Foods in 2017 and offering delivery.
Third, because e-commerce could grow most quickly in emerging markets, companies that specialize
in those regions could take a significant share if they enhance services for local consumers. For
example, China’s Tencent expanded at a 14 percent CAGR from 2018 to 2023, buoyed by the
popularity in China of its social media platform WeChat, which allows users to chat, post, and
purchase in one “super app.” In Latin America, Mercado Libre’s marketplace ecosystem recorded a
59 percent CAGR from 2018 to 2023.81 Key features, such as an integrated payment option for Latin
American countries and localized services for sellers, have helped the platform maintain its market
position in a region with rapidly growing internet penetration.
Last, regulations in some markets that aim to limit the size and influence of e-commerce players
could lead to more fragmentation in the industry.
Swing factors
— How might social commerce change the industry structure?
— To what degree could e-commerce develop in categories that currently are mostly nondigital,
such as healthcare?
— Will adoption be faster or slower in regions with lower e-commerce penetration than in more
mature markets? The pace could be influenced by factors such as regulation; by supply-side
constraints, such as logistics infrastructure; or by governmental restrictions.
— How much could gen AI and other new technologies affect e-commerce? Could they kick off an
escalatory race that draws new entrants? Or will technological advances be incremental and
mostly used by the large incumbents to solidify their market position?
— How might innovations in physical retail influence the trajectory of e-commerce?

Arena-
Ben Ellencweig 2. Artificial intelligence
software and services
Artificial intelligence (AI) refers to a machine’s ability to perform cognitive functions that we usually
associate with humans.82 Some types of AI, such as machine learning, have been around for decades
(these primarily numerical techniques used for applications such as prediction, categorization, and
optimization can be described as “analytical AI”), while others are newer (although built on previous
technologies). The most prominent is generative AI (gen AI), which uses unstructured inputs, most
often prompts from a natural language such as English, to generate unstructured outputs. These are
outputs such as text, images, audio, video, and computer code that aren’t typically represented in the
rows and columns of a spreadsheet or database.
We define this potential arena as services and software that incorporate AI, such as those that use
natural language processing, predictive analytics, and other AI algorithms and functionalities. We
excluded the hardware required to operate the technology, including components to run AI systems
such as processing units and chips designed by companies such as Nvidia (which are covered in the
semiconductors entry in this compendium).
The promise of gen AI became apparent in late 2022, when OpenAI released ChatGPT, a service that
responds to users’ questions and instructions with remarkably humanlike written answers. Since
then, AI has made headlines and excited markets. Investors are flocking to companies developing
advanced AI and particularly to those focusing on gen AI: equity investments in that technology
jumped from $5 billion in 2022 to $36 billion in 2023. In an April 2023 McKinsey survey, nearly a third
of respondents reported that their organizations were using gen AI regularly in at least one business
function, and 40 percent reported that their organizations would increase their investment in AI
overall because of advancements in gen AI.83 By early 2024, ChatGPT had more web traffic than such
giants as Netflix, Pinterest, and Twitch, according to a survey by FlexOS on gen AI platform usage.84
Predicting the future of an industry that has so recently grabbed the world’s attention is difficult.
Yet despite this uncertainty, there are already identified opportunities along the gen AI value chain,
including building end-user applications and model hubs, and providing the tooling for machine
learning or large language model (LLM) operations. There are additional opportunities for new
applications of AI technologies, such as gen AI–enabled agents that act in the world, and the more
expansive potential for systems whose scope of capabilities is sufficiently broad to be called artificial
general intelligence.85 In the meantime, systems tailored for specific applications are being built with
customized AI models using techniques such as fine-tuning and retrieval-augmented generation.86
And there is still plenty of headroom for the adoption of analytical AI. As companies seize these
AI-enabled opportunities, the industry could become an arena. Given the rapid pace of development
Note: This section describes and breakthroughs in AI, many aspects of this field will likely undergo considerable change by 2040,
the potential growth and
dynamism of the AI software with significant potential for outsize growth and dynamism over the next two decades.
and services arena. It
should not be read as a
comprehensive account of the
industry. To learn more about
Growth
AI software and services Developments in analytical AI and gen AI are poised to drive the industry’s growth by improving
and associated industries,
please refer to content from business and worker productivity. In our modeled scenarios, the arena’s revenues grow from
the McKinsey Technology,
Media & Telecommunications
$85 billion in 2022 to $1.5 trillion in a lower range of scenarios in 2040 and to $4.6 trillion in a higher
Practice and McKinsey Digital. range of scenarios, a CAGR of 17 to 25 percent.87

Arena-
Past McKinsey research analyzed more than 500 uses for AI and estimated their potential economic
impact, which would have two components.88 The first is the economic value added by analytical
AI—such as machine learning and deep learning—which could amount to an estimated $9.4 trillion
to $15 trillion in 2040.89 For instance, analytical AI could revolutionize vehicle maintenance by better
predicting failures, cutting operational costs in the transportation industry up to 10 percent. Or it
could help clinicians detect rare diseases in patients more quickly by processing and analyzing large
data sets of images.90
The second component is the economic value added by gen AI. About $2.6 trillion to $4.4 trillion
of that annual value would come from gen AI enterprise use cases, and about three-quarters of
that value would be in four areas: customer operations, marketing and sales, software engineering,
and R&D. These use cases include those that reduce costs as well as those that increase revenue
for individual firms, overall increasing productivity in the economy. But these enterprise use cases
do not account for all the productivity gains of individual knowledge workers, automating aspects
of their occupations. Incorporating all these cases of individual worker productivity enabled by
gen AI in addition to the enterprise use cases could unlock a total of $6.1 trillion to $7.9 trillion of
value annually.91
Adding these components together yields our estimated range of total economic potential of
$15.5 trillion to $22.9 trillion annually by 2040. The actual amount will likely depend on two factors:
how quickly AI is adopted and the technology’s effectiveness. To reach the $22.9 trillion figure, the
vast majority of the 500-plus use cases would have to be adopted around the world, and enterprises
would have to achieve reduced costs and improved productivity. In practice, companies could most
effectively implement these new technologies through transformation of business domains rather
than individual use cases.92
To arrive at our estimates of the potential arena’s revenues, we multiplied the full economic value
by the 10 to 20 percent of it that technology providers typically capture. Here, if the economic
impact was $15.5 trillion in 2040, and if 10 percent of that amount was captured by AI companies,
the companies’ revenues would come to $1.5 trillion, the amount estimated in our lower range of
scenarios. If the economic impact was $22.9 trillion, and if AI companies captured 20 percent of it,
their revenues would come to $4.6 trillion, the amount estimated in our higher range of scenarios.
The actual capture rate will likely depend on the industry’s competitive dynamics. If a few AI software
and services companies can differentiate themselves from their competitors by means of advanced
technological capabilities, those companies may be able to set prices higher and produce a capture
rate closer to 20 percent. If many companies are battling fiercely for business, they may have less
pricing power and a lower capture rate.
Dynamism
The AI software and services industry can be described as having two segments, one nascent with
the advent of gen AI, and another that is a natural evolution of the software industry. Their market
dynamics and player composition may change over the next 15 years.
The first segment consists of companies that train “frontier” foundation models—that is, the biggest,
most advanced models, which can perform a variety of tasks by using expansive artificial neural
networks and enormous sets of varied data. (For example, the GPT models by OpenAI, Gemini by
Google, Claude by Anthropic, and Command by Cohere are frontier models with language and
increasingly multimodal capabilities, such as processing images.) This segment of the industry
currently has a few major players because building a foundation model requires access to massive
amounts of data, capital, and infrastructure. Those requirements can pose challenges even to
large, established technology companies. Another factor creating barriers to entry is this segment’s

Arena-
dependence on the semiconductor industry, because that technology can be expensive and scarce,
giving an edge to companies with access to large amounts of capital (and cutting-edge talent).
This advantage can become a virtuous cycle that keeps industry leaders ahead as a result of the
difficulty a competitor might have in creating industry-leading frontier models. The first escalating
barriers to entry are the high training costs and huge amount of computing power required to build
frontier models. For example, Meta plans to deploy 350,000 Nvidia H100 AI chips, with a value of
more than $9 billion.93 However, some experts believe that the advent of more open-source models
as well as smaller, often more customized models may lead to more fragmentation even among
foundation models, potentially altering this industry segment’s composition.
The industry’s second segment consists of companies that produce software and services that are
enabled by AI and fulfill a wide variety of specific needs in the economy. For example, companies
recently funded by the start-up accelerator Y Combinator plan to use AI to serve clean-energy
developers, behavioral health companies, logistics vendors, and fashion designers as clients.94
Specialized AI software and services may be based on the frontier foundation models or on models
that require less investment to train. Frameworks and techniques could be layered onto the models to
build solutions that cater to specific use cases. For example, retrieval-augmented generation could
be added to larger gen AI platforms to help users gather information from verified data repositories,
creating a tool designed to answer specific questions accurately.95 Also, many customers of AI
software and services will require full solutions, that is, integrated offerings that include building
applications, integrating software, access to foundation models (and compute power), and software
maintenance and servicing.
This segment largely resembles the overall software market: it is fragmented because of the huge
variety of needs and applications, and new competitors continually emerge and outcompete
established ones. AI could be applied in almost all industries, vastly increasing the number of specific
uses and creating significant space for new entrants. Barriers to entry would be low, and the large
companies training frontier foundation models could focus on just a few of the most profitable and
scalable use cases, rather than tailoring their products more narrowly for hundreds of them.
Swing factors
— Will AI continue to be constrained by the limits in the capacity of the semiconductor industry’s
ability to supply chips, or will another constraint such as limits on electrical power emerge as a key
bottleneck in AI development and deployment?
— How will governments act to shape the burgeoning AI arena? How will they approach regulation?
How might some governments invest in “sovereign AI,” that is to say, an interest in having AI
systems developed within their borders? How will AI regulation and capital flow differ in China,
Europe, the United States, and elsewhere? How will regulations on intellectual property, safety,
and the size of AI companies interact with geopolitics across different markets?
— How will the distribution of value evolve among players that train foundation models, those that
provide software based on those models, and the businesses and consumers they serve, for
example with the release of some open versus proprietary models? Are AI models a long-term
source of value? Will they commoditize or be captured through open source?
— How will pricing models evolve? How will training costs be managed? How will the various players
across the AI value chain capture value as costs and pricing models change over time?
— Is AI a supporting feature or capability of the software industry, or is it an entirely new modality of
how we use technology that will transform the software industry?

Arena-
Bhargs Srivathsan 3. Cloud services
Cloud services providers, the largest of which are often referred to as hyperscalers, offer IT services
over the internet, including on-demand IT infrastructure and platforms as a service. Cloud services
have been historically divided into three segments: infrastructure as a service (IaaS), platform as
a service (PaaS), and software as a service (SaaS). IaaS offers computing and storage resources
over the internet, PaaS offers platform services that allow users to develop apps that can put those
resources to use, and SaaS consists of software provided over the internet. In practice, IaaS and
PaaS are deeply integrated, so we consider them one offering in our analysis. In addition, the types of
infrastructure and platform services that cloud providers offer continues to expand, for example by
encompassing capabilities enabled by artificial intelligence (AI). Our definition of the cloud services
arena includes only the integrated IaaS/PaaS segment. With the growth in AI as a part of software,
an increasing amount of SaaS is covered as AI software and services (also identified as a potential
arena). That said, since all cloud providers are likely to see themselves as AI providers, and cloud is
necessary for the computation-at-scale requirements of AI, these two arenas are inextricably linked.
By introducing new technologies, optimizing operations, and scaling, the cloud services industry
can deliver services with more resiliency, enable more agility, and foster more innovation than
“on-premises” infrastructure and platform solutions run by corporate IT departments. Over the
past decade, more and more computing has shifted to the cloud from individual corporate systems,
enterprise-owned data centers, and local networks. A survey estimates that around 94 percent of
enterprises use cloud services.96 As a result, industry revenues have grown sharply—from $32 billion
in 2017 to $270 billion in 2023.
The cloud services industry is already an arena, as described in chapter 1, and is likely to remain one.
In an increasingly connected world, companies will need more and more computing and storage
capabilities. The Internet of Things (IoT), for instance, often requires the public cloud’s ability to store,
process, and analyze data on a massive scale. Companies have scaled up their cloud capabilities:
over the past decade, Walmart has built a hybrid cloud platform that runs more than 1 million CPU
cores and can execute 170,000 backend adjustments to the retailer’s website each month, 1,700
times more than could be achieved with earlier capabilities.97 Cloud migration at scale can also bring
benefits in other industries, such as the financial and automotive sectors. For instance, autonomous
vehicles are now generating and recording thousands of parameters in real time that would require
the cloud to store and process and feed into R&D. AI also needs this computing ability, and smaller
providers focusing on the AI market have recently entered the cloud services industry.
As the cloud opens up the potential to reallocate where computing gets done, some specific
workloads could be intentionally deployed “at the edge,” that is, on local computing systems.
These include applications with low latency requirements (including real-time applications such as
autonomous vehicles or manufacturing) or where security/privacy concerns could be mitigated by
Note: This section describes constraining data to specific devices. And many workloads will be spread across cloud and edge,
the potential growth and resulting in hybrid solutions.
dynamism of the cloud arena.
It should not be read as a The rate of cloud migration may be slower than some expected due to large time and investment
comprehensive account of
the industry. To learn more needs—companies with cloud programs profiled by McKinsey increased their cloud adoption by
about cloud and associated only 5 to 10 percent between 2022 and 2023.98 However, migration of workloads to the cloud
industries, please refer to
content from the McKinsey is also expected to continue. By 2030, cloud services could generate $3 trillion of EBITDA
Technology, Media &
Telecommunications Practice
increases for Forbes Global 2000 companies by helping them digitize core operations, accelerate
and McKinsey Digital. product development, and more.99 This means value remains on the table: cloud players still have
opportunities to harness returns on investments on cloud at scale.

Arena-
Growth
We anticipate that cloud adoption could grow further as a result of continued enterprise migration
to cloud services and increased demand for cloud services to meet the computational needs of new
technologies. We excluded SaaS in our estimates because it occupies a separate competitive space
from cloud infrastructure/platforms; pure software companies and infrastructure-and-platform
providers operate differently, compete in different markets, and have different financial models.
Global external IT spending was about $3 trillion in 2022 and is expected to grow to $8 trillion in
2040, a 6 percent CAGR. These outlays are distributed across key areas, such as hosting, network,
cybersecurity, and application development and support. Our estimates include spending on public
cloud service providers and exclude internal labor cost. The cloud services industry’s 2022 revenues
accounted for 8 percent of global IT external spending. In our estimates, the industry could represent
between 19 and 41 percent of that spending by 2040. In our analysis of today’s cloud arena, the
industry’s revenues grew at 17 percent CAGR from 2005 to 2020. They could continue to increase at
a similar rate, or growth could be slightly moderated to a 12 percent CAGR. Revenues could increase
from $220 billion to $1.6 trillion in the lower range of scenarios and to $3.4 trillion in the higher range
of scenarios in 2040.100
Our scenarios are based on two factors: first, companies continue to migrate their traditional
computing and storage processes to the cloud.101 Despite the strong growth of cloud services over
the past decade, a 2023 survey of enterprises with more than $1 billion in revenues found that only
about 20 percent of their workload was in the cloud, and most planned to more than double that
share by 2026.102
This continued migration would increase the share of cloud expenditures in overall hosting and
networking IT spending. In addition, the cloud could capture a larger share of cybersecurity outlays,
as certain on-premises cybersecurity software services migrate to the cloud.
In the higher range of scenarios, migration happens quickly, and the cloud industry captures more
spending from hosting, networking, and cybersecurity. In the lower range of scenarios, migration is
slower and, in some cases, reverses for certain workloads (such as cloud “repatriation,” the process
of moving applications from the public cloud to private infrastructure). Slower cloud migration and
increased repatriation could have several causes, including malfunctions or breaches at cloud
providers, data protection laws, and security and privacy concerns (for example, national security
concerns for defense-related industries).
The second factor underpinning our estimates is the computational requirements of newer
technologies—such as AI, IoT, and virtual reality—that could boost demand for cloud services. The
need for computational power of generative AI (gen AI) models is particularly high during the training
phase, which typically involves computationally intensive operations, but the volume demand
for inferencing (the use of these models) could also rise as AI adoption increases (though some
inferencing could also be deployed at the edge). Many in the AI industry have highlighted insufficient
computing power and limited availability of graphics processing units (GPUs) as the key bottlenecks
in the development of the technology. Nvidia, a leading player in the AI technology market, has
announced plans to meet this growing need by shipping 1.5 million AI server units that will consume
85 terawatt-hours (TWh) of electricity a year by 2027,103 representing 25 to 35 percent of the 240–
340 TWh of energy that cloud data centers used globally in 2022 (see the entry for semiconductors
in this compendium).104 The rapid growth in AI services may continue to drive significant demand
toward cloud services. In the higher range of scenarios, AI technologies grow more quickly and
businesses require large amounts of cloud service to enable that growth. In the lower range of
scenarios, businesses do not employ these technologies as widely, so the need for additional cloud
services is limited.

Arena-
Dynamism
The cloud services industry has few major players; its four largest companies currently account
for about 60 percent of total revenues. In 2023, Amazon Web Services (AWS) took in 31 percent,
Microsoft 20 percent, Google 7 percent, and Alibaba 3 percent.105 These companies have grown
thanks to significant network effects, scale and scope advantages, and high barriers to entry. To
be competitive, providers must make very large investments, particularly in data centers: the three
companies with the highest capital expenditures from 2016 to 2021 (AWS, Microsoft, and Google)
invested more than $380 billion over a five-year period, nearly 60 percent of the industry’s total.
With companies migrating more of their workloads to the cloud, competition may intensify as cost
and convenience become more important to customers. Achieving competitive pricing and migration
experiences also requires large amounts of R&D.
In the long term, the industry may continue to have just a few large players given its escalatory
dynamics and global radius of competition. Large companies are already investing heavily in
acquisitions to help them offer more specialized services.
However, data sovereignty regulations106 and new investments in different regions of the world could
cause some fragmentation by enabling the emergence of regional competitors. Yet even in that
scenario, investment and economics of scale would probably leave room for only a few surviving
competitors in each region. In addition, the global presence of major hyperscalers will likely allow
them to adjust to data sovereignty rules and continue to compete in regional markets.
Cloud services tailored for AI—a rapidly growing segment within the larger cloud-services industry—
could lead to shifts in market share among companies. Such services give clients the tools and
infrastructure they need to develop, train, and deploy AI and machine learning models at scale. A
few new players have cropped up; for example, specialized GPU cloud providers have grown quickly
due to their data centers and services tailored to meet the unique needs of AI model training and
deployment. The largest companies in the cloud services industry, including Microsoft and Google,
also offer “frontier” foundation AI models, which clients can use to create more specialized models.
It is easy to imagine that the fierce competition for performance among frontier models could shift
share among cloud providers.
Swing factors
— How costly and challenging will cloud migration be?
— Will the massive volumes of data in legacy systems make it less cost-effective for enterprises to
transfer workloads to the cloud?
— Could there be regulatory concerns particularly about concentration risk? Could this drive
companies to use private infrastructure rather than the public cloud?
— How will increasing data residency regulations and sovereign AI impact the adoption of public
cloud for enterprise workloads?
— How could gen AI adoption accelerate cloud adoption, and how would it shift corporate budget
allocations among IT, product, R&D, and other areas? For instance, could generative AI make it
cheaper to build applications? Will AI workload demand surge as anticipated, or can businesses
optimize training processes that will reduce this demand in the cloud?
— If a catastrophic failure or cybersecurity attack occurred at a major provider, what potential
scenarios could unfold?

Arena-
Contributing authors:
Moritz Rittstieg,
Patrick Schaufuss
4. Electric vehicles
The technology to build electric vehicles (EVs) has existed since the invention of the automobile in
the 19th century, and electric cars vied with steam-powered and gas-powered vehicles well into the
20th century. The introduction of the Ford Model T, the proliferation of cheap gasoline, and the use of
internal combustion engine (ICE) vehicles during World War I eventually led to mass-market adoption
of the gas-powered vehicle.
More recently, however, EVs have been mounting a global comeback as concerns about CO 2
emissions from traditional ICE vehicles made them an attractive alternative. With recent
investments in technology, they have become increasingly practical and mainstream—in 2023,
EVs accounted for 18 percent of all new passenger cars sold globally. Between 2010 and 2019,
breakthroughs in battery capacity and efficiency increased the range of the average battery
electric vehicle (BEV) from 130 kilometers to 340 kilometers. In the same period, the launch of
lower-cost EV models brought down the average purchase price from $55,000 to $37,000, a
decrease that put the total cost of ownership of many EVs over their lifetime roughly on a par with
that of a comparable ICE car.107
At the same time, EVs are getting a boost from mandates and timelines in several countries seeking
to eliminate the use of ICE vehicles. That includes a law approved by the European Union in 2023 that
effectively banned the sale of many new ICE vehicles by 2035, and mandates in Canada and the state
of New York requiring that 100 percent of new passenger vehicle sales be electric in the next 10 to
15 years.108
As a result, the EV industry became one of today’s arenas, as described in chapter 1. Plenty of signs
indicate that it could be one of tomorrow’s, too. EVs are growing significantly faster than the rest
of the automotive sector. Overall annual passenger vehicle sales are projected to increase at a
modest rate, from 74 million in 2023 to 79 million in 2030. By contrast, global passenger EV sales are
projected to increase fourfold, from 13 million in 2023 to between 31 million and 46 million in 2030, or
about half of the total annual sales of all cars forecast for that year.
Escalatory competition has been ramping up in the global EV market. Chinese EV companies have
begun exporting to global markets, with OEM BYD overtaking Tesla in global EV sales in the fourth
quarter of 2023 and Chinese brands accounting for about 50 percent of global BEV sales. Other
OEMs are looking to compete. BMW reached one million cumulative EVs delivered in the first quarter
of 2024, with sales up more than 25 percent year-on-year.109 Ford’s electric vehicle sales were up
85 percent year-on-year in the first quarter of 2024, even as the company has delayed production of
new electric pickups and SUVs.110
However, the market has seen some slowdowns, such as delayed electrification targets of some
OEMs, and in 2024 experts revised EV sales growth projections downward for the first time. There
the potential growth and are concerns that early adopters have already purchased their first electric vehicles, and the next
dynamism of the electric wave of potential adopters may have stronger concerns about price and range that would slow
vehicles arena. It should not
be read as a comprehensive growth. But even with these recent headwinds, EVs are still experiencing strong growth, though
account of the industry. To it is less accelerated than in the vehicles’ early years. R&D investment in EVs remains a priority,
learn more about electric
vehicles and associated and EV-focused companies continue to spend more than traditional carmakers. In 2023, Tesla
content from the McKinsey
spent $4 billion on R&D, or about $2,200 per car, 1.2 times more than what Ford spent, and almost
Center for Future Mobility. 3.0 times Toyota’s spending.

Arena-
Growth
The growth of EVs responds to the continued prioritization of sustainability by consumers and
governments, lower lifetime costs of ownership, escalatory competition to improve quality, and
enabling investments in charging infrastructure and electricity supply and distribution. Our estimate
of electric vehicles’ growth includes revenues attributable to producers of BEVs, plug-in hybrid
electric vehicles, and fuel-cell electric vehicles (FCEVs), which run on hydrogen. We include both
passenger vehicles and commercial vehicles such as trucks and vans.111 We exclude hybrid vehicles
that don’t have the ability to be plugged in to recharge, as well as vehicles powered by natural gas.
To obtain our estimate of revenues for the entire EV industry, we included the sales of electric
passenger vehicles and commercial vehicles, including heavy-, medium-, and light-duty trucks, as
well as light commercial vehicles. In our estimates, revenues would grow from $450 billion in 2022
to $2.5 trillion in the lower range of scenarios and to $3.2 trillion in the higher range of scenarios by
2040, implying a CAGR of 10 to 12 percent. In 2023, revenues were $590 billion.
In 2023, 18 percent of all new passenger vehicles sold globally were EVs. However, the sales numbers
vary by region: in China, almost 30 percent of new passenger vehicles were EVs; in Germany,
30 percent; in France, 20 percent; in the European Union, 16 percent; and in the United States,
8 percent. In July 2024, more than half of the vehicles sold in China were “new energy vehicles,”
including BEVs and plug-in hybrids.112 In our estimates, the global share increases to between 82
and 96 percent by 2040. Projections also vary greatly by region. In high-penetration regions like
Norway, the EV share of passenger car sales could grow from about 90 percent currently to 100
percent, while lower-penetration countries such as Nigeria could see that share growing to only 15 to
22 percent by 2040. As a result, industry revenues from passenger EVs could grow from $530 billion
in 2023 to between $2.0 trillion and $2.3 trillion in 2040, implying a CAGR of 8 to 9 percent. In 2023,
revenues were $530 billion. With EV sales estimated to overtake ICE sales sometime between 2025
and 2030, most of this growth would come from replacing traditional ICE sales, whose share would
decline by a corresponding amount (Exhibit 1).
Electric vehicles could account

for about half of global passenger
vehicles sales by 2040.

Arena-
Exhibit 1
Sales of electric passenger vehicles may surpass sales of internal

combustion engine passenger vehicles by 2030.
Global passenger vehicle sales, million Global passenger vehicles in operation, billion
Electric Internal combustion engine Electric Internal combustion engine
Projected Projected
1.6
80
1.4
70
82–96%
of sales 1.2
60 electric
by 2040
50 1.0
40 0.8
40–54%
30 0.6 of vehicles
40–59%
of sales electric
electric by 2040
20 0.4
by 2030
10 0.2
0 0
2025 2030 2035 2040 2025 2030 2035 2040
Note: Projections show middle scenario. Passenger vehicle totals exclude commercial vehicles, such as light commercial vehicles, light-duty trucks,
medium-duty trucks, and heavy-duty trucks. Electric category includes battery electric vehicles, plug-in hybrid electric vehicles, and fuel-cell electric
vehicles.
Source: McKinsey Center for Future Mobility (May 2024)
McKinsey & Company
Among commercial vehicles, by contrast, the share of BEVs and FCEVs sold was only about 4 percent
in 2023 (Exhibit 2). In our estimates for 2040, that share would grow for medium- and heavy-duty
trucks to between 60 and almost 90 percent in regions of high penetration, such as Europe, where
regulations aid EV adoption. In regions with medium penetration, such as the United States and
China, the estimated share is 40 to 80 percent. In the rest of the world, that share could grow to
between 10 and 50 percent. To reach these shares, however, battery electric trucks could require
either significant improvement in battery energy density or reduced payloads compared with
traditional diesel trucks, which could challenge the economics of the trucking industry. An alternative
to BEVs would be a higher share of FCEV trucks, which do not face the same range and payload
trade-offs but would need a major scale-up of hydrogen refueling stations and development of a
robust hydrogen supply chain. Because of this trade-off, some medium- and heavy-duty trucking use
cases such as long-haul freight could be suited to FCEV power trains. However, heavy-duty trucks
in use cases with more predictable downtimes and distances driven per day, such as those used for
garbage, sanitation, or local distribution, are more likely to use BEV power trains.

Arena-
Exhibit 2
Sales of electric commercial vehicles may surpass sales of internal

combustion engine commercial vehicles by 2030.
Global commercial vehicle sales, million Global commercial vehicles in operation, billion
Electric Internal combustion engine Electric Internal combustion engine
Projected Projected
20
0.3
15
41–70%
of sales
electric 0.2 19–35%
10 by 2035 of vehicles
electric
by 2040
0.1
5
0 0
2025 2030 2035 2040 2025 2030 2035 2040
Note: Projections show middle scenario. Commercial vehicles include light commercial vehicles, such as light-duty trucks, medium-duty trucks, and
heavy-duty trucks. Electric category includes battery electric vehicles, plug-in hybrid electric vehicles, and fuel-cell electric vehicles.
Source: McKinsey Center for Future Mobility (May 2024)
McKinsey & Company
By contrast, light-duty trucks and light commercial vehicles are likely to be primarily BEVs. This
is because these smaller trucks and commercial vehicles are much lighter and so will not face as
stark of a range and payload trade-off as heavier trucks. For example, Tesla’s Cybertruck weighs
roughly 3,000 kilograms (kg); the battery alone for a heavy-duty garbage truck can weigh more
than 7,000 kg. EVs’ share of vehicle sales for these lighter segments is estimated to be 85 to
almost 100 percent in regions of high penetration such as Europe and China; between 50 and
almost 100 percent in the United States, which has medium penetration; and between 30 and
80 percent in the rest of the world.
Taken together, revenues from EV light commercial vehicles and light-, medium-, and heavy-duty
trucks combined would grow from $42 billion in 2022 to between $540 billion and $900 billion in
2040, implying a CAGR of 15 to 19 percent. In 2023, revenues were $53 billion.
Four main factors are driving growth to the higher or lower end of that range: consumers’ willingness
to pay a premium for sustainable vehicles, EV cost and quality, charging infrastructure, and the supply
and distribution of electricity to the charging infrastructure.
Willingness to pay a purchase premium for sustainable vehicles and policy incentives. The
global effort to mitigate climate change and reduce carbon emissions is driving governments and
consumers toward lower-emissions options. In 2023, ground transportation globally accounted
for roughly seven gigatons of CO2 emissions, or about 20 percent of global emissions, making it a
priority sector for emissions reduction.113 These considerations help power consumer demand, with
57 percent of all global car buyers in the McKinsey Center for Future Mobility (MCFM) consumer
sentiment survey for 2024114 stating that sustainability was a primary factor when purchasing a new

Arena-
car, though this varies hugely across regions. In the United States, only 49 percent said sustainability
was a primary factor. In China, the figure was 59 percent, while in Germany it was 43 percent.
These concerns also drive policy. Governments around the world are offering incentives for EV
purchases, and the EU and 12 US states have instituted bans on new ICE vehicle sales that will take
effect in 2035.115 In the United States, buyers of many new EVs receive a tax credit of up to $7,500,
and buyers of heavy commercial EVs could receive a tax break of up to $40,000 per truck under the
2022 Inflation Reduction Act (IRA).116 Chinese EV buyers receive tax exemptions worth $2,000 to
$4,000 on EV purchases, and in major Chinese cities, EVs receive preferential vehicle registration
that can save car purchasers meaningful money or time. The strength of continued consumer
demand and political pressure for lower carbon emissions will determine whether this factor drives
growth toward the higher or lower end of our estimates.
EV cost and quality. Several technological factors drive EVs’ cost and quality, and they could
determine whether revenues tend toward the lower or higher end of our range. By 2023, most EVs
had reached parity with comparable ICE vehicles on total cost of ownership over the vehicle’s lifetime,
though this depends on the region, car make and model, usage profile, and available subsidies.
However, this parity has limited impact on consumers, who are largely more sensitive to purchase
price than to a complicated calculation of the total cost of vehicle ownership over many years. And
EVs remain meaningfully more expensive to purchase than comparable ICE vehicles.
The battery is a main driver of EV cost and accounts for 30 to 40 percent of the total cost of a vehicle.
Several forces affect battery prices, with more than half the cost of a lithium-ion battery coming from
its raw materials. Prices of lithium, a critical raw material, have been volatile, increasing about fivefold
from January 2021 to January 2023 before falling back to 2021 levels by January 2024. Supply
dynamics could also affect battery prices. Even though identified global lithium reserves are four times
larger than the amount that would be needed to supply batteries for projected EV sales between 2022
and 2040, meeting EV demand could still require a three- to fivefold surge in lithium production.
The rapid proliferation of EVs is also likely to add to the demand for other key materials such as
cobalt and nickel. Yet EV battery pack prices have fallen significantly over the past decade, thanks
to technological improvements and economies of scale, which caused costs to plummet from
thousands of dollars per kWh in the early 2000s to under $100 per kWh today. McKinsey’s Battery
Insights team forecasts that the cost could to continue to decline to $50 to $75 per kWh by 2040,
mostly driven by continued technological improvements in energy density. While the price of the raw
material inputs is volatile and harder to predict, designers have managed to adapt the mix of materials
in batteries (for example, by reducing the amount of cobalt used) and may be able to keep doing
so (for example, by using cheap, abundant sulfur). In addition, batteries can be recycled to reclaim
the raw materials within them, as some companies are already doing and some regulations already
require. If those trends continue, they could help keep down future prices in the event that the price
of raw materials spikes.
Battery lifespan can be another concern for EV buyers. Recent research has suggested that
batteries tend to last for the lifetime of a typical vehicle, about eight to 12 years, depending on
how far they are driven. A 2024 survey of 20,000 EV drivers found that less than 1 percent of cars
manufactured since 2016 required battery replacement, compared with 13 percent for cars made in
2015 or before.117 Most EVs do not use a full battery cycle—charging to 100 percent and depleting
to zero percent—but instead charge to about 80 percent and avoid falling lower than 20 percent,
which can significantly improve battery lifespan. And fast-charging a vehicle doesn’t seem to make a
meaningful difference in battery lifespan, as long as the fast charge does not fully cycle the battery.
See further discussion in our “Batteries” entry in this compendium. Range also influences EVs’ rate
of growth. About 70 percent of consumers in a February 2024 survey by MCFM agreed or strongly

Arena-
agreed with the statement, “I am anxious about buying an electric vehicle because of the limited
driving range.” This “range anxiety”—the worry that a single charge might not allow an EV to go as
far as an ICE car—is their biggest concern about owning an electric vehicle, and about 30 percent
of those surveyed said insufficient range is their primary reason for not purchasing an EV. According
to a previous MGI analysis,118 however, passenger BEVs are already able to meet most use cases in
driving range. In the United States, the median range of a BEV is about 400 kilometers, which would
be sufficient for about 80 to 90 percent of journeys made without having to recharge by average US
households. The average driver in the United States drives close to twice as far each year as drivers
in Australia, France, and the United Kingdom, implying that the share of use cases currently met by
BEVs would be even higher. Thus, the rapidity with which EV car design, battery technology, and
electric power trains improve will determine whether some consumers will be able to overcome range
anxiety and purchase a vehicle.
Charging infrastructure. The sheer scale of the infrastructure needs for EV charging introduces
factors that could limit the industry’s growth. While the infrastructure to distribute and sell gasoline
around the world was built up over decades, the public and private chargers for millions of EVs would
have to be built at scale in less than a decade to meet growing EV demand and to achieve climate
goals. Although public charging infrastructure grew at a rapid rate of about 50 percent annually from
2018 to 2023, this growth was heavily driven by China, which accounts for about 70 percent of global
public charging points. Other regions have struggled to keep pace with the growth of EV sales: there
are eight EVs per public charging point in China, 14 in Europe, and 30 in the United States. Charging
infrastructure would need to be built in homes, offices, shopping centers, and along highways, to
name just a few locations. We estimate that 135 million charging points may be needed globally by
2030 to meet EVs’ energy demand, nearly five times the 25 million charging points installed by 2023,
including about 15 million public chargers.
This infrastructure requires substantial investments. About $2 trillion in global capital expenditures
may be needed until 2050 to build the public and private charging infrastructure required to meet
EV demand.119 While EV charging infrastructure (EVCI) is likely to become a meaningful industry in its
own right, whether it can scale up sufficiently to support growing EV adoption is unclear. Profitability
for public EVCI operators is highly dependent on the utilization rate, which is a function of how
many EV drivers are nearby, and EV demand is capped by consumer concerns about the availability
of public charging. Timing will be critical: scaling up too slowly would limit consumer EV demand;
scaling up too rapidly could mean large losses while EV adoption catches up. Many public-charging
operators are operating at a loss, investing in a network in hopes that it delivers future profits. Some
legislative efforts (such as the US IRA) have aimed to encourage EVCI network building, but it remains
to be seen whether those measures will spur enough EVCI investment fast enough to allay concerns
about charging availability.
Electricity supply. The EV industry could also be limited by the supply and distribution of electricity
to the charging infrastructure. To power the more than 240 million EVs on the road just by 2030, we
estimate annual electricity needs would grow from around 140 million megawatt-hours (MWh) in
2022 to more than one trillion MWh by 2030, representing about 3 percent of future global electricity
generation.120 This electricity would also need improved transmission and distribution infrastructure:
many of the 26.5 million chargers projected to be built in private settings (such as those in homes,
offices, and retail locations) may need the local utility to install improved substations that can handle a
larger load than the current infrastructure. Timing is important as well. Many EVs are charged at night,
so there will be a need for better energy storage solutions as well as renewable power generation at
night (for example, excluding solar).
And to deliver on its promise of reducing greenhouse gas emissions, the electricity powering EVs
would need to come from low-emissions sources. Because the manufacturing process for EVs

Arena-
emits more carbon than the making of an ICE car does, it can take 30,000 to 80,000 kilometers
of driving for passenger cars and 35,000 to 65,000 kilometers for heavier sport utility vehicles to
break even on their lifetime carbon emissions on grids such as those in Europe and the United States.
Typical vehicles in the United States and Europe are driven distances of about 150,000 to 250,000
kilometers, so BEVs may produce emissions that are 20 to 60 percent lower than top-performing ICE
vehicles in their lifetimes. For heavier vehicles and in areas where the grid is more carbon intensive,
that break-even point gets further away (a heavy SUV on a carbon-intensive grid such as India’s may
not reach carbon breakeven during its expected lifetime). If EVs fail to deliver on their promise to
reduce carbon emissions, consumers and governments may not continue to support their growth.
Dynamism
The EV market, which currently features a few large players, has seen increased numbers of Chinese
players and EV-focused new entrants alongside traditional manufacturers that have existing
competencies in manufacturing and are competing at scale. In the early 2010s, Chevrolet and Nissan
started selling plug-in electric cars, followed closely by Tesla. These manufacturers were able to gain
large market shares in the early years because there were fewer players and limited models were
offered. They could move quickly and scale up in EVs before other carmakers entered the market.
Notably, Tesla successfully scaled up Model 3 production, gaining up to 40 percent share of the US
market in 2018.121
Fast-forward to today, and the market still has just a few large players. In 2023, the five largest
EV producers, led by BYD, controlled about 50 percent of total EV production, while the top five
players accounted for only 35 percent of the overall passenger car market, still led by ICE vehicles.122
Nonetheless, there have been many new entrants into the EV market, both start-ups, such as
Polestar and Lucid, and major automotive OEMs, such as Ford and Hyundai. And in China, an entirely
new ecosystem has developed around EVs, with vertically integrated players such as BYD. The
rapidly expanding Chinese domestic market for EVs features just a few larger players, with the five
biggest Chinese companies combined selling 60 percent of all EV units sold in 2023.123
Chinese companies such as BYD and Geely have also increased their EV sales massively outside of
China: in 2023, they produced about 50 percent of EV units sold globally. In Brazil, four of the top five
EV sellers were Chinese companies, selling a combined 80 percent of EV units in the country.
New entrants have found fertile ground in EVs: in 2023, about 50 percent of new EV units were
produced by companies that sold more EVs than ICE vehicles. Part of the reason new entrants
are finding success in the market is that the embedded capital and skills of the traditional ICE
incumbents—such as their expertise in building ICE power trains and managing an extremely
complex network of suppliers—did not confer a strong enough advantage to these incumbents to
prevent new entrants from successfully entering the market. This is an example of the escalatory
dynamic we explore in chapter 2, in which a technological step change can open up the bottom of a
new S-curve, enabling the entry of disruptive new players.
However, OEMs that sell more ICE vehicles than EVs still occupy a major role in the industry and are
increasingly investing in EVs. However, the dynamics are different in different markets. In Norway,
OEMs that sell more EVs than ICE vehicles sell about half of EVs, whereas in Brazil, new EV entrants
have a roughly 70 percent market share, with BYD leading at more than 36 percent of EV sales.124
The game could play out in many ways. The older automakers have some advantages, including,
above all, the expertise necessary to manufacture products at large scale. But EV-only companies
also have an advantage: a laser focus on EV power trains. EV-only companies can streamline their
research, design, and manufacturing operations to focus solely on electric vehicles, while traditional

Arena-
automotive OEMs must continue to optimize across a wider portfolio of EVs and ICE cars, and may not
be able to optimize for EV production in the same way. In addition, expensive battery packs are the
heart of every EV, and the R&D invested in enhancing battery performance could prove even more
valuable than the ability to build the rest of the car. EV-only companies could successfully entrench
themselves in the market by capitalizing on first-mover advantages, scaling up and bringing down
prices, and offering customers higher quality, such as better range. And players in either group
may continue to integrate vertically, as BYD has done with battery production and even battery raw
material mining; the carmaker is involved in lithium mining from China to Chile.
Swing factors
— Will more consumers be won over to EVs as electric vehicles get better and the infrastructure
expands, or are there deeper consumer preferences toward ICE cars beyond just solving for cost
and convenience?
— Will the industry be successful in ramping up mining to meet raw material demand at the right
speed to match the growing demand in EV batteries? If not, how will the swings in price and
availability along the EV supply chain change the trajectory of the industry?
— How will the pace of the energy transition, including government support for it, affect the EV
market? Depending on the development of the market, will bans on ICE vehicle sales remain in
place or be reconsidered?
— Will the buildup of public charging infrastructure be sufficient to meet EV demand?

Arena-
Craig Macdonald,
Lia Grigg
5. Digital advertising
Digital advertising funds an enormous swath of the internet by providing an increasing percentage of
its revenues.125 Growth in digital ads is expected to continue, indicating that the industry could be one
of the arenas of tomorrow.126
There are several reasons to believe digital ads will continue to expand their reach. First, more
people around the globe are coming online and joining the global middle class, resulting in more
consumers for companies to target with digital advertising. Second, consumers are spending more
time with media, and increasingly in digital forms, resulting in both an increased volume of novel ad
products and more demand from advertisers. Third, the digital shift is creating more innovation in,
and better performance from, ad products. The digital shift also lowers costs for advertisers to enter
the market (for example, generative AI, known as gen AI, reduces the cost of content creation). This
generates more demand and expands the total addressable market. Last, more places to display
advertising are emerging, and ads are becoming a more important revenue model for non-traditional
advertising sectors such as e-commerce, retail, last-mile delivery, banking, and telecom, especially
in developing economies.
Digital ads could also see high dynamism. Several factors contribute to this: the evolution of walled
gardens (closed platforms where providers have full control over content and consumer data), the
impact of gen AI on traditional search and on small and midsize business (SMB) access, regulatory
and data privacy actions, the rise of social media influencers, the convergence of commerce and
Note: This section describes advertising, and shifting consumer behavior.
dynamism of the digital
advertising arena. It
should not be read as a
Growth
comprehensive account of Digital advertising is expected to continue to grow faster than GDP. In our estimates, revenues
the industry. To learn more
about digital advertising from digital advertisements increase from $520 billion in 2022 to $2.1 trillion in the lower range of
scenarios and to $2.9 trillion in 2040 in the higher range of scenarios, a CAGR of 8 to 10 percent.127
please refer to content from
the McKinsey Technology, Much of this growth comes from the ongoing shift in advertising spend from traditional to digital
Practice and Growth,
media: in 2022, digital advertisements accounted for 65 percent of total advertising, a share that is
Marketing & Sales Practice. expected to grow to 80 to 90 percent by 2040 (Exhibit 1).

Arena-
Exhibit 1
A mix of shifting market share and overall market expansion could

power growth in digital ads.
Spending on advertising, $ billion

+7–8% per year
3,200
320
2,700 (10%)
540
(20%)
Traditional ads¹
2,900 Digital ads²
(90%)
2,100
800 (80%)
280
(35%)
520
(65%)
Lower and higher

2022 scenarios
2040

¹Traditional ads are those displayed on non-digital channels such as TV, radio, and print media.
²Digital ads are those consumers see on computers and mobile phones.
Source: McKinsey analysis incorporating data from EMARKETER, Statista, Oxford Economics, PwC Global Entertainment & Media Outlook 2023–2027:
Moffett Nathanson, and Magna
McKinsey & Company
The growth of digital ads is primarily driven by four factors: growth of the global middle class,
consumer attention moving to digital, the demonstrated ROI of digital ads, and new forms of
digital media.
First, digital advertising may grow as more of the world comes online and more of the global
population enters the middle class. This could increase the number of people that companies
around the world are trying to reach and enable digital ways to reach them. In particular, developing
economies and regions could see substantial increases in smartphone and broadband penetration
bringing more consumers onto the platforms that serve digital ads and improving the effectiveness
(and pricing) of those ads with more user scale and richer and broader data signals. For instance,
the Asia–Pacific region, Latin America, and the Middle East and Africa could achieve a smartphone
penetration level of almost 90 percent as soon as 2026 (Exhibit 2). In addition, the ten countries
with the largest projected increase in middle-class population from 2020 to 2030 are all developing
economies. Collectively, these ten countries could have about 900 million people making the
transition into the middle class in this period.128

Arena-
Exhibit 2
The world—especially emerging markets—will be increasingly online.
Smartphones per 100 people, by region

North America European Union Asia–Pacific Latin America Middle East and
North Africa
115
104
88 94 88
82 83 85
76 75 72
65
57
44
28
2016 2021 2026* 2016 2021 2026* 2016 2021 2026* 2016 2021 2026* 2016 2021 2026*
Fixed broadband connections per 100 households, by region
90 95
78 81
74
61 67 66
60 57
38 39
18 22
13
2016 2021 2026* 2016 2021 2026* 2016 2021 2026* 2016 2021 2026* 2016 2021 2026*
*Projected
Source: Analysys Mason
McKinsey & Company
Second, consumer attention will continue to shift to media, and increasingly to digital formats.
Overall time spent on media has mostly been increasing across economies, but more notably in
developing economies. For example, Indian consumers spent 225 minutes per day on average on
media in 2016, and this could increase to almost 400 minutes by 2026, or a 6 percent CAGR, while
Chinese consumers’ time spent on media could climb from 384 minutes per day in 2016 to almost
500 minutes by 2026, a 3 percent CAGR. The share of digital media—where consumers are served
digital advertising—in particular is increasing across economies. In the United States, digital media
accounted for 40 percent of consumers’ media consumption in 2016. That could grow to 66 percent
by 2026 (Exhibit 3).

Arena-
Exhibit 3
People are spending more time consuming digital media.
Distribution of time per day spent on different media channels, %

Digital: Traditional:
Mobile Connected TV Desktop/laptop Other digital TV Radio Print and other
US Germany UK
100 100 100
66% 65%
digital digital
50 in 2026 50 50% 50 in 2026
vs 40% digital vs 45%
in 2016 in 2026 in 2016
vs 36%
in 2016
0 0 0
2016 2026* 2016 2026* 2016 2026*
641 767 589 614 600 598

Total minutes per day
South Korea China India

100 100 100
71%
58% digital
50 digital 50 in 2026 50
in 2026 vs 51% 43%
vs 48% in 2016 digital
in 2016 in 2026
vs 27%
in 2016
0 0 0
2016 2026* 2016 2026* 2016 2026*
420 498 364 493 225 397

Total minutes per day
*Projected
Source: EMARKETER; McKinsey Global Institute analysis
McKinsey & Company

Arena-
Third, digital ads continue to demonstrate improving ROI. Digital ads offer advertisers more
measurable formats, such as search and “performance” advertising, in which companies pay for
clicks and measurable results instead of promoting general brand awareness. These formats also
offer the ability to deliver ads to very specific and precise target audience segments, allowing
advertisers to more efficiently deploy their advertising budgets and measure much more precisely
their effectiveness.
The ability to execute targeted and highly measurable campaigns has enabled SMBs, which could not
afford and did not benefit from broad-audience brand advertising on traditional media, to become a
significant force in advertising. It is also increasing the overall number of advertisers substantially:
in 2020, 200 advertisers supplied 88 percent of US TV ad revenue, while three million advertisers
were on Facebook,129 the vast majority of which were SMBs. These SMBs will continue to expand
the market as the shift to online video and connected TV opens up new digital advertising formats to
smaller advertisers who would never advertise on traditional TV.
New technologies such as gen AI are expected to make digital ads more effective by enabling
hyper-personalization of content and could increase demand for digital ads by making digital ads
cheaper and easier to produce. For example, the online used-car marketplace Carvana used gen AI
to create 1.3 million bespoke videos aimed at individual customers. Each ad celebrated a customer’s
car purchase in a personalized way.130 Virgin Voyages created a tool that uses the digital likeness
of entertainer Jennifer Lopez (the company called the tool “Jen AI”) and allows customers to share
personalized video invitations with friends.131
Fourth, new forms of digital media, such as retail media, video games, and digital out-of-home, are
emerging. Many of these forms did not exist or had little scale a decade ago but are growing rapidly
today. Innovative formats enable new players outside of traditional and digital media to sell ads,
further increasing the supply of ads. Companies including Amazon, Chase, United, and Uber have
added ads to their digital platforms. Video game platforms such as Roblox have increasingly used
digital ads to boost revenues; digital advertising almost doubled as a proportion of global mobile
gaming revenues, from 28 percent in 2017 to 51 percent in 2022.132
Advertising has also become an increasingly important revenue model for media companies. This
is especially true in developing countries where consumer willingness to pay for media is low,
contributing to the growth of advertising-enabled media models. For instance, around 60 percent
of the 135 million monthly active Spotify users in Latin America, one of the fastest-growing regions,
are on the free, ad-supported version of the application.133 In developed markets such as the United
States, ad-supported tiers of the leading streaming platforms are emerging as both higher revenue
and higher margin than subscription-only tiers. For example, in 2023 the ad-supported tiers of
Discovery, Hulu, Netflix, and Peacock each had higher average revenue per user than their premium
subscription tier.134
Dynamism
The industry benefits players with scale due to two-sided network effects of both users and
advertisers: a larger pool of users and data enables better matching and signaling, giving advertisers
incentive to continue placing advertisements on these platforms as performance improves. As of
2023, four companies accounted for about 64 percent of the industry’s global revenues—Alphabet,
with 30 percent, Meta, with 21 percent, Amazon, with 7 percent , and ByteDance, with 6 percent.135
The industry’s total addressable market may continue grow, but shifts in market shares remain
uncertain. While large incumbents may still see growth, two questions are relevant in understanding
how the industry’s structure may change: First, will the value from newly emerging revenue pools
be captured by new entrants or by incumbents? Second, to what degree is the core business of

Arena-
incumbents at risk of capture by new entrants? In a more disruptive scenario, new entrants may
emerge and evolve into a new set of large players if outsize shifts occur.
Several major dynamics will likely determine how these share shifts play out in the next 15 years,
across both the supply side, meaning the extent of advertising players can provide, and the demand
side, or the channels where consumers will interact with ads and the number of advertisers and the
share of revenue they spend on digital advertising.
From the supply side, several factors could change the industry’s structure, potentially shifting
shares of current players. First is the degree to which walled gardens stay global. As new mixes of
content and user data are emerging, the competitive position of existing walled gardens may be
challenged. Because SMBs may drive regional growth, a new wave of walled gardens whose sizes
and scopes would be more regional than global could emerge.
Second are new formats. New form factors, such as virtual conversational agents that help
consumers discover products, may provide new marketing experiences for consumers. The extent
to which players succeed at creating innovative formats could drive large share shifts even among
the existing players, and open up space for new entrants, though those new entrants would need to
invest heavily in technologies such as AI in order to compete. The rise of influencers is an example
of how an emerging format shifted shares in the industry; advertisers have reallocated spend to
influencers and away from traditional paid media to reach more consumers.
Third is the convergence of commerce and advertising. The extent by which closed loops win out
and how much commerce players move to invest in content could structurally change the industry.
Commerce players have large amounts of consumer data, which they could leverage when expanding
into advertising on their channels.
Fourth are regulatory and data privacy actions. Developments in local or regional policies, such
as laws related to data collection, management, and residency, could have significant operational
requirements for advertising companies that aspire to participate globally. These policies include the
General Data Protection Regulation in Europe and the American Data Privacy and Protection Act,
which is the subject of ongoing discussions.
From the demand side, the shifts in consumer behavior could change how and where advertisers can
find consumers. For instance, channels in which consumers discover various products or services,
and the ways consumers engage with brands such as through connected TV and app stores, can
influence where consumer attention—and hence revenue—shifts. Consumer preferences could
also move toward user-generated content and short-form videos such as those on TikTok. Ad
expenditures follow consumer attention and spending; if new platforms attract more consumer
attention, the relative ad spend on those platforms will also increase due to the supply factor of new
advertising formats explained above.
The impact of gen AI has crosscutting effects on both supply and demand, affecting how advertisers
can participate as well as changing customer behaviors. For one, gen AI could transform how
consumers navigate media and the internet. For example, it could disintermediate traditional search
and allow customers to discover information or products directly, thereby creating a new channel for
ad placements. Gen AI also democratizes content creation by making it cheaper and easier for SMBs,
which can now advertise in more formats.

Arena-
Swing factors
— As new walled gardens emerge, how will the interoperability between platforms evolve, and how
will ad delivery enablers including “open ad tech”—the ecosystems on the open web that connect
advertisers and publishers—and ad agencies respond?
— How will gen AI affect the customer journey, from brand to product and from discovery to
purchase, including potentially disintermediating search? How will it affect the degree of
consolidation or democratization of digital content?
— Which emerging media formats—for example, augmented reality and virtual reality, metaverse,
and wearables—will create new channels of customer attention, and how will advertisers adapt to
deliver ads using these new formats?
— How will brand advertising change as performance advertising becomes increasingly common,
and where will brand advertising spend be directed?
— How might additional ad inventory coming onto the market affect prices and margin? At what level
will inventory growth boost the overall market, and when will it just lead to lower prices?

Arena-
Nikolaus Lehmann 6. Semiconductors
Semiconductors, sometimes called microchips or integrated circuits, are the invisible building
blocks of the digital world. Their ubiquity and importance were highlighted during the COVID-19
pandemic, when a surge in demand led to supply shortages of not only computers, smartphones, and
other consumer electronics but also of cars and household appliances, all of which rely heavily on
microchips.136 In the early years of the 21st century, as semiconductor manufacturers chased smaller
and smaller distances between chips’ transistors, a disaggregated business model appeared, with
some “fabless” companies designing chips and other companies specializing in manufacturing them
in factories called “fabs.” Those shifts made the industry one of today’s arenas, as we describe in
chapter 1. From 2005 to 2020, its revenues grew by a CAGR of 7 percent.
The industry could be one of tomorrow’s arenas, too. Some recent technological trends, such as
AI, machine learning, robotics, and autonomous driving, are likely to boost demand for complex,
domain-specific chips. Generative AI (gen AI) is expected to generate 0.2 QFLOPs (quettaFLOPs)
of computational demand in 2024.137 By 2030, demand for gen AI computational power alone
could increase to 25 QFLOPs, an annual rate of growth of about 125 percent.138 And increasing
global prosperity is likely to continue to drive demand for electronics in developing markets. The
semiconductor industry is investing heavily in R&D and manufacturing: total capital expenditures
increased at a CAGR of 16 percent from 2015 to 2022, though there was a dip in 2023.139
Furthermore, many governments, including those of China, the European Union, and the United
States, are offering the industry significant subsidies to bolster domestic manufacturing.
Growth
Increased demand in several markets could drive the growth of semiconductors. About 80 percent
of this growth could come from four market segments: computing and data storage, automotive,
Note: This section describes wireless communications, and industrial electronics (exhibit). This industry could continue to escalate
the potential growth
and dynamism of the R&D and capital expenditures to efficiently meet rising demand, as seen by continuous node resets
semiconductors arena. and the emergence of fabs. In our modeled scenarios, the semiconductor industry’s revenues grow
It should not be read as a
comprehensive account of from $630 billion in 2022 to $1.7 trillion in a lower range of scenarios in 2040 and to $2.4 trillion in
a higher range of scenarios, a CAGR of 6 to 8 percent.140 In those scenarios, prices increase by an
about semiconductors
and associated industries, average of about 2 percent a year, and supply and demand are balanced in the long run (though there
the McKinsey Industrials &
is often some shorter-term volatility, as a result of boom-and-bust cyclicality and the impacts of other
Electronics Practice. factors such as the COVID-19 pandemic).

Arena-
Exhibit
The data storage, wireless, automotive, and industrial electronics

segments drive overall growth in the global semiconductor market.
Semiconductor industry revenue by segment, $ billion

Lower scenario, 2022–40 Higher scenario, 2022–40
$1.1 trillion increase $1.8 trillion increase
2,500 2,400
2,000
1,700
1,500
1,000
630 630
500
0
2022 2040 2022 2040
lower scenario higher scenario
Share of Revenue Share of Revenue

2022–40 CAGR, 2022–40 CAGR,
Segment increase, 2022–40,
increase, 2022–40,
% % % %
Wired
9 8 7 9
communications
Consumer
8 5 6 6
electronics¹
Industrial
14 6 13 8
electronics
Automotive 23 10 19 11
Wireless
22 5 19 6
communications
Computing and
24 4 37 8
data storage

¹Consumer electronics consists mainly of non-mobile devices (eg, TVs, smart speakers, smart bulbs, set-top boxes).
Source: Omdia; IDC; McKinsey Global Institute analysis
McKinsey & Company

Arena-
Computing and data storage. In our scenarios, revenues from semiconductors that are used in this
segment grow from $230 billion in 2022 to between $480 billion and $890 billion in 2040, a CAGR of
4 to 8 percent. Most of that growth would be fueled by demand for semiconductors used in servers,
though chips for PCs, storage, and peripherals would also play a part.
The higher end of these scenarios yields the $890 billion market size and is based on the following
underlying dynamics:
— The AI and cloud services industries grow very quickly and need servers to support them (see
“Cloud” and “AI and software services” in this compendium). For example, Nvidia’s data-center-
specific revenue grew 427 percent in the first quarter of 2024 compared with the first quarter of
2023, with 87 percent of the company’s overall revenues attributable to data-center solutions.141
Industry leaders have already signaled their intention to sharply increase AI investment: Google’s
DeepMind is expected to surpass $100 billion in total AI spending, and Microsoft and OpenAI are
reportedly considering building a supercomputing cluster worth $100 billion.142
— Demand continues to increase for higher-value chips for servers, such as graphic processing
units and AI accelerators, which are now employed mostly by cloud services companies and high-
performance computing users.
— Demand for computing in developing countries rises quickly as their economies grow.
In the lower range of these scenarios, which yields the $480 billion market size, the magnitude of
each of these factors is lower: the AI and cloud services industries grow less quickly, the higher-value
chips for servers do not reach spending parity with standard ones, and demand for computing in
developing countries grows less quickly.
Automotive. In our scenarios, revenues from semiconductors used in vehicles grow from $60 billion
in 2022 to between $300 billion and $390 billion in 2040, a CAGR of 10 to 11 percent. While overall
automotive demand rises modestly over this period, averaging about 1 percent a year, growth is
primarily driven by the increasing value of semiconductors in the average car, which was $600 in
2022 and by some estimates could reach $1,400 by 2028. That increase, in turn, is driven by three
trends. The first is a growing share of electric vehicles, which can need roughly ten times as many
semiconductors as traditional vehicles, on average. In addition, traditional vehicles are increasing
their use of semiconductors. Second, driving assistance features, which require sensors and
computing power, become more common. Third, cars’ “infotainment” systems grow increasingly
sophisticated, including more head-up displays, digital instrument clusters, and rear-seat
entertainment.
Wireless communications. In our scenarios, revenue from chips used in wireless devices grows from
$180 billion in 2022 to between $400 billion and $500 billion in 2040, a CAGR of 5 to 6 percent.
About half of that growth comes from increasing demand for smartphones, particularly as they
become more common in developing economies. The rest of the growth comes from chips used in
other wireless devices and infrastructure. Actual revenues in 2040 will depend on how rapidly people
shift to mid-tier, high-tier, and 5G-enabled smartphones, which use more expensive semiconductors.
Industrial electronics. In our scenarios, revenues from semiconductors used in industrial
electronics grow from $75 billion in 2022 to between $200 billion and $300 billion in 2040, a
CAGR of 6 to 8 percent. These semiconductors are used in a variety of industries that are expected
to grow quickly, such as medical electronics, industrial automation, and electricity generation
from renewable sources.

Arena-
Dynamism
The semiconductor industry has four major types of players: fabless designers such as Nvidia, which
design chips; foundries such as TSMC, which manufacture them; integrated device manufacturers
(IDMs), such as Intel and Samsung, which do both; and companies such as ASML that sell specialized
manufacturing tools to the foundries and IDMs. The semiconductor industry features a few large
players, and companies that offer better or cheaper products typically hold a majority of the global
market. This is especially clear in specific technology segments along the value chain. For example, in
2022, ASML captured 87 percent of the $18 billion worldwide lithography equipment market;143 Intel,
78 percent of the $62 billion market for central processing units; Nvidia, 84 percent of the $20 billion
market for graphics processing units; TSMC, 63 percent of the $107 billion foundry market; and
Samsung, 37 percent of the $144 billion memory segment.144 These five segments make up more than
half of total semiconductor market revenues.
Although the industry seems likely to continue to feature a few large players, companies’ market
share could be changed by a move, already under way, from general-purpose to domain-specific
chips. The shift derives from the growing popularity of fast-growing applications for semiconductors,
such as machine learning, 5G, and the Internet of Things. For those uses, domain-specific chips can
help reduce energy consumption. The adoption of these chips could increase the share of the market
held by fabless designers, which could provide a larger portfolio of solutions.
The move by tech companies outside the industry to design their own semiconductors instead
of buying them could also shift market shares. Apple has designed processors for its mobile
devices and, more recently, for its computers. Meta and Alphabet have designed chips to suit
their requirements. And Tesla is designing chips for autonomous driving. If this trend persists,
these tech companies could capture market share from the fabless chip designers such as Nvidia.
Although huge share shifts in chip manufacturing will likely be harder, given the large amounts of
capital required to compete with TSMC in the global chip foundry market, it is reasonable to expect
increased competition in this space as geopolitical considerations drive countries to diversify their
chip-manufacturing supply chains.145
Swing factors
— How will evolving national policies and geopolitical shifts affect regional and global markets?
— What will happen to the market for AI-specific chips as incumbents, cloud services companies,
and start-ups try to gain market share?
— How long can the current pace of technological advancements and step changes in
semiconductor capabilities be expected to continue? What impact could a potential slowdown of
innovation (for example, a disruption to Moore’s law) have on chip prices, computing performance,
and research advancements?

Arena-
Ani Kelkar 7. Shared autonomous
vehicles
Shared autonomous vehicles (SAVs) are road vehicles that have no human driver and can be hired
to transport passengers for a fare. The industry is in its infancy. However, many companies are
designing purpose-built SAVs, and several are already road testing regular vehicles enabled with
autonomous technology.
The SAV industry’s rapid technological development is a sign that it could become an arena. Although
autonomous vehicles were first tested in academia in the 1960s, development took off with the
DARPA Grand Challenge contests starting in 2004. SAV development has been led by companies
such as Apollo (owned by Baidu), AutoX, Cruise (owned by GM), Pony AI, and Waymo (owned by
Alphabet). Robo-taxis are already in service in some US cities, such as Phoenix and San Francisco,
and in China in the cities of Beijing, Chongqing, and Wuhan.146
Furthermore, from 2010 to 2021, more than 400 companies—mainly venture capital and private
equity firms, followed by big tech and automotive OEMs—reported a total of more than $100 billion
in investment in robo-taxis and robo-shuttles.147 Although venture capital and R&D investments in
SAVs in the United States fell from a peak of $25 billion in 2019 to $3 billion in 2022, the mobility
industry more broadly has experienced an uptick in investments from late 2023 into 2024.148 Recent
investments in the SAV industry have included Alphabet’s plan, announced in July 2024, to invest
an additional $5 billion in its autonomous technology provider, Waymo.149 This variable investment
sentiment suggests that commercialization in the industry may go more slowly than some previous
forecasts had indicated.
Growth
The potential for outsize growth in the SAV arena is based on sustained technological progress,
a compatible regulatory environment, the financial feasibility of broader adoption, and consumer
acceptance. The broader shared mobility industry changed dramatically in the first decade of the 21st
century as regional and multinational tech companies, beginning with Uber, began competing with
traditional taxi companies by offering ride-hailing services. As of 2022, taxis still garnered 58 percent
the potential growth and of the industry’s global revenues of $1.2 trillion, while ride-hailing services took in 14 percent. The
dynamism of the shared remainder was earned by car-sharing and car-rental services, shuttle services, and micromobility
autonomous vehicles arena.
It should not be read as a services, such as those that let people rent bicycles from curbside racks. We exclude the revenues of
comprehensive account of mass transit operators from our definition of the industry.
about SAVs and associated In our scenarios, SAVs could capture 25 to 51 percent of the shared mobility industry’s revenues by
content from the McKinsey 2040, taking market share from human-driven taxis and ride-hailing services. Because the pace
Automotive & Assembly
Practice, and the McKinsey
of technological development and adoption of SAVs is uncertain, our range of estimates for the
Center for Future Mobility. industry’s revenues in 2040 is wide: between $610 billion and $2.3 trillion (Exhibit 1).

Arena-
Exhibit 1
The shared autonomous vehicle industry’s revenue could grow to

$610 billion–$2.3 trillion in 2040.
Global market for shared and micromobility ground transportation,¹ $ billion

5,000
4,500
Car rentals
4,000 Other shared transportation

(car-sharing, pooled shuttles)
Taxis
Ride-hailing
3,000 Micromobility
(bicycles, shared and private)
2,400 Shared autonomous vehicles
(robo-shuttles, robo-taxis)
2,000
1,200 2,300
1,000 (51%)
0.06
(<0.01%) 610
(25%)
0
Lower and higher
2022 scenarios
2040
¹Market data excludes private cars and public transit.

Source: McKinsey Center for Future Mobility (May 2024); McKinsey Global Institute analysis
McKinsey & Company
To estimate the industry’s potential size, we considered the revenues that it could earn by
transporting passengers in fully autonomous vehicles, which have what the industry calls level four
or level five automation.150 Such vehicles might include robo-taxis or robo-shuttles. We excluded
autonomous vehicles that would not run on roads, such as trains, and those that would fly (which we
discuss in the entry “Future air mobility”). We also excluded personal vehicles, even though many of
them might have automated driving systems by 2040.151
Several factors will determine the industry’s revenues in 2040.
The first factor is technological progress. Significant obstacles stand in the way of operating self-
driving cars safely and reliably. Teaching cars to identify objects, people, and animals on roads is a
challenge, and so is making sure that the vehicles can continue navigating in bad weather. Although
companies are making impressive headway, some of the challenges remain, because the array
of circumstances that will need to be taken into account will grow over time as SAVs spread to
increasingly complex and unpredictable environments.152 As a result, SAVs are likely to first emerge
in the center of cities, which offer the highest density of potential passengers, and find cities where
weather conditions are less likely to cause issues for sensors.153 Unforeseen problems can arise, even

Arena-
in the areas that have already been selected. SAVs need to anticipate accidents and be equipped to
respond to them.
Second is the regulatory environment. There are myriad unanswered social and political questions:
how would liability for accidents be distributed between SAV manufacturers, operators, and
riders? Do we measure the safety of an autonomous vehicle versus a human driver, or should a
different standard be applied? What vehicle registration process would be mandated for SAVs, and
what powers should law enforcement or government have over the vehicles? Some countries and
jurisdictions have started issuing guidelines for autonomous driving.154 However, globally, regulatory
risks continue to be one of the biggest impediments to autonomous-vehicle adoption. In a 2021
survey of executives from automotive, transportation, and software companies, 60 percent said
that they viewed insufficient regulatory support, rather than technological obstacles or low demand,
as the factor that is most likely to limit the adoption of autonomous driving.155 In a 2023 McKinsey
survey of autonomous-vehicle leaders, the percentage of respondents who said that regulation is the
biggest bottleneck for the industry remained at 60 percent.156
A third factor that will determine the industry’s revenues is financial feasibility. Deploying SAVs
at scale requires significant capital investment, and considerable outlays are needed to attain
profitability. The cumulative capital outlay requirement is projected at about $50 billion to $67 billion
through 2030.157 In a capital-constrained environment, commercialization might occur more slowly
because only a limited group of profitable markets will emerge. Although the initial price tag is high,
the capital required to manufacture SAVs is expected to shrink over time. Some manufacturing
costs associated with traditional vehicles could be reduced by eliminating parts that are redundant
in an SAV.
In addition to capital requirements, the unit economics of operating SAVs would also dictate the
industry’s course. In our scenarios, operational costs that are lower than those of alternative modes of
transportation would encourage adoption. Once the industry is mature, the cost of a trip by SAV could
range from 60 percent less than the cost of using a personal car to 30 percent more, depending on
whether a person is traveling alone or pooled.158 The cost could be 30 to 80 percent lower than using
a ride-hailing service (Exhibit 2).159 The savings result from lower labor costs (because SAVs have
no human drivers) and lower fuel and maintenance costs (because computer-directed driving could
optimize fuel consumption and reduce maintenance requirements). SAVs do have some operations
and management costs that human-driven cars do not, such as remote operations support centers
that assist vehicles in emergencies. However, these additional costs are expected to be offset by the
lower per-trip cost of SAVs.

Arena-
Exhibit 2
Travel by shared autonomous vehicles could become cost competitive

relative to travel by private car by 2035.
Cost per passenger-mile traveled, forecast for 2035 and beyond,¹ index (1.0 = private vehicle cost)
Shared autonomous vehicle (SAV) scenarios:
Accelerated ramp-up of usage Expected ramp-up Other modes of transportation
Public transit² 0.2–0.3
SAV robo-shuttles 0.35

(pooled riders) 0.4
SAV robo-taxis 0.7

(pooled riders) 0.9
SAV robo-taxis 1.0

(solo riders) 1.3
Car-sharing (eg,
1.3
Zipcar, Turo, Miles)
Car rental 1.4
Ride-hailing (solo) 1.8
Taxis (solo) 2.1
0.83 1.0
Per trip price parity of an owned private car, Per trip price parity of an owned private car,
not including the cost of vehicle depreciation including the cost of vehicle depreciation
¹For shared modes end-customer price per passenger-mile traveled and for private vehicle MSRP share per year and operational costs (eg, maintenance,
insurance, charging, cleaning, parking, tolling, financing). Private vehicle cost assumes MSRP of $34,360 and lifetime mileage of 197,106 miles.
²Depending on city-specific pricing and trip length.
Source: McKinsey Center for Future Mobility
McKinsey & Company
Fourth, the degree to which consumers are willing to choose shared mobility of any kind will affect
the size of the SAV industry. This is influenced by both consumer sentiment and regulations. We
estimate that roughly 40 percent of all passenger miles were traveled in shared vehicles in 2022. In
our scenarios, that share in 2040 would range from 40 percent in the low case to 60 percent in the
high case.160 The share could be pushed up by regulatory changes in large cities, such as license
plate rationing and the imposition of congestion charges.161 It could also rise if people became less
willing to own their own cars for reasons related to convenience, cost, safety, or the environment,
for example.
Dynamism
The SAV market is nascent and is likely to show high dynamism as the arena grows. An SAV product
needs to bring together three things: the autonomous driving technology, the physical cars, and the
shared mobility service that plans trips, logistics, and interfaces with the customer. This makes the
industry particularly dynamic as competition emerges from companies that currently excel at each

Arena-
one, with technology providers, ride-hailing companies, and carmakers approaching the industry
from different angles.
The autonomous technology providers, such as Baidu’s Apollo, GM’s Cruise, and Alphabet’s Waymo,
are the furthest along today. They plan not only to operate fleets but also to build the technology
necessary for autonomous driving. Some of those companies will probably procure vehicles from
established carmakers and then integrate software into them; Waymo uses Jaguars, for example.
Others, like Cruise, might be owned by carmakers and use vehicles built by their parent company.
Carmakers, too, might play a role by equipping their vehicles with autonomous driving capabilities
and then selling them to other companies.162 Tesla has stated ambitions to enter the SAV market, and
Volkswagen aims to supply self-driving vans to ride-hailing companies.163
By contrast, ride-hailing companies might end up procuring the technology and cars they need from
others. Ride-hailing companies will have various competitive advantages from their experience
operating mobility services, including existing reputations, existing customer bases, and developed
platforms. For example, Uber has partnered with both Waymo (2023) and Cruise (2024) to make their
shared autonomous vehicles available on Uber’s platform.164
There are significant escalatory factors that create tendencies toward eventual consolidation. The
extensive amount of research and development needed to develop autonomous vehicles creates a
high barrier to entry for technology providers, encouraging consolidation within the industry and with
carmakers. There are also high barriers to entry for operating a shared autonomous vehicle service;
as with today’s ride-hailing operators, there are strong network effects and cost advantages to scale
that encourage consolidation. These factors could create a market led by a few key players along with
some smaller regional companies. Similar dynamics have emerged in today’s ride-hailing industry.
Of course, the emergence of SAVs would also shift market share in the broader shared mobility
industry. We could see a disruption similar to that of the early 2000s, when the shared mobility
industry’s composition began tilting away from local taxi companies and toward the regional and
global ride-hailing companies that offered customers convenience and lower costs. In the same way,
SAVs could disrupt the industry by offering customers a cost-effective alternative to ride-hailing.
Swing factors
— How quickly can SAV manufacturers overcome key technological hurdles, such as increasing
range and mobility beyond geofencing, reducing reliance on remote interventions, and creating
purpose-built platforms that enable positive unit economics? 165
— How will regulations governing autonomous driving affect the SAV industry? And how will policy
makers determine operators’ and owners’ liability for SAV-related accidents?
— How will the evolution of SAVs affect consumer sentiment and future competition in the market
(for example, product differentiation by vehicle amenities)?
— How will the business model differ across the industry’s three kinds of companies—autonomous
technology providers, ride-hailing companies, and carmakers (for example, different offerings,
customer base, and impact of regulations)?
— Will players experience regionalization of their tech stack due to regulations, data limitations,
privacy mandates, and restricted access to high-end semiconductors?

Arena-
Andrew Sierra 8. Space
Since the first moon landings over 50 years ago, most of the demand for space endeavors has
been driven by government interests rather than commercial opportunities. Over the past decade,
however, the industry’s commercial sector has seen robust growth. Because of this surge of
interest, commercial applications like remote satellite connectivity could help make space an
arena of tomorrow.
Several space companies have accelerated growth in the past decade. SpaceX, one of the most
active commercial service and infrastructure players, provides internet service via its Starlink
system—a constellation of more than 5,000 small satellites—and offers launch services (more than
300 launches as of June 2024).166 In addition, the company is testing Starship, a fully reusable launch
system that could carry humans for interplanetary and orbital spaceflight. Other companies have
also been exploring space tourism: Virgin Galactic carried passengers to the edge of space and
back twice in the first half of 2024 and plans to enter commercial service by 2026.167 Planet Labs,
another space services company, provides imagery of the Earth’s landmass several times a day (for
example, to enable data-driven agriculture). And Origin Space, a Chinese space mining company, has
launched satellites into low Earth orbit to collect debris and assess space resources; it aims to begin
commercial asteroid mining by 2045.168
Alongside private companies, governments around the world continue to shape the space arena by
increasing their investments and deepening their collaborations with private players. For example,
NASA awarded multibillion-dollar contracts to three private companies to develop competing
designs for a lunar terrain vehicle.169 At the same time, some government-backed missions have
succeeded at a relatively low cost: India’s Chandrayaan-3, which landed on the moon, had a
$74 million budget—less than was spent on some Hollywood space movies, such as Gravity and
Interstellar, that had budgets of more than $100 million.170 More nations are entering the space race,
too. For example, Saudi Arabia is seeking to diversify its economy by promoting scientific research
and boosting national security via space, while Peru has launched the “Internet para Todos” program
to enhance internet connectivity in rural regions using satellites.171
Four key trends have transformed the space industry over the past decade and are helping to pave
the way for space products and services to become ubiquitous.
First, launch costs have decreased exponentially. For example, when adjusted to 2021 dollars, low-
Earth orbit heavy launch costs have declined by more than 95 percent, from $65,000 per kilogram in
1980 to $1,500 per kilogram in 2021.172 Second, commercial innovation makes it possible to do even
more in space. Commercially available applications now allow for identification of objects on Earth
at a resolution of 15 cm, a 20-fold increase compared with an initial resolution of three meters. Third,
investment is flowing from a wider group of investors to a larger and more diverse set of applications.
Private-sector investment in space is reaching all-time highs, with more than $70 billion invested
in 2021 and 2022 combined.173 Fourth, there is excitement over, and interest in, the latest space
developments, with government and business leaders considering what space could enable for
Note: This section describes the future.
dynamism of the space arena. Growing investment and interest in space could dramatically change our lives by helping to address
It should not be read as a
comprehensive account of a wide range of challenges. For example, space-based Earth observation could help identify
infrastructure risks by monitoring structures, such as dams and nuclear power plants. Developments
about space and associated
industries, please refer to in satellite data and analytics could transform disaster relief efforts by providing real-time information
Aerospace & Defense
to shorten response time during humanitarian crises, quickly identify affected populations, and
Practice. optimize coordination of first responders and evacuations.

Arena-
Space-based capabilities could also help bridge current digital divides separating developed and
developing nations by providing access to high-speed internet, education, and healthcare services.
And space technologies could provide precise monitoring of agriculture, natural resources, and
environmental changes. The advancements not only would bring financial gains but could also help
alleviate global challenges like climate change and hunger.
Growth
The space industry consists of two segments, commercial and government. Over the past decade,
the space industry has been tightly linked to government policies. With growing commercialization
and expansion of downstream services, the space industry could produce strong growth in
both segments.
These two segments each divide into two subsegments. The commercial segment consists of the
services and end-user equipment subsegment, and the infrastructure and support subsegment; and
the state-sponsored segment is divided into the civil subsegment and the defense and intelligence
subsegment (exhibit).
Exhibit
The space arena is divided between the commercial and

state-sponsored segments.
Segments and subsegments of the space industry
Commercial State-sponsored
Services and Infrastructure Defense
Civil
end-user equipment and support and intelligence
Communications tools for Satellites, ground Military-run missile Projects by civil space
mobility in remote areas, stations, rockets, defense systems, agencies (eg, NASA, ESA)
internet connectivity, and support services positioning and such as Global
commercial positioning to maintain space reconnaissance, Positioning System
and navigation, satellite operations satellite-based and risk response
networks, and rocket surveillance, and
launch services intelligence gathering
39–51%
of market 11–18% 22–32% 16–22%
in 2040¹
Example companies
SpaceX, Amazon, Eutelsat, SpaceX, Airbus, Lockheed Martin, Boeing, Lockheed Martin,
Planet Labs, NEC Space, Lockheed Martin, Raytheon, Boeing, General Northrop Grumman,
Mitsubishi Heavy Industries Blue Origin, Axiom Space, Dynamics, ThyssenKrupp, SpaceX, Airbus,
Mitsubishi Heavy Industries Northrop Grumman, China Arianespace,
Aerospace Science and BAE Systems,² China
Technology Corporation Aerospace Science and
Technology Corporation
¹Middle range scenario.

²Includes Ball Aerospace (acquired 2024).
McKinsey & Company

Arena-
For all four subsegments, we considered the industry’s “backbone,” which includes satellites (such
as those used for space-based communications), launchers, services such as satellite TV and
GPS, and applications with revenues that are directly attributable to space hardware and service
providers. A report by McKinsey and the World Economic Forum published in April 2024 describes
this backbone in greater detail.174 The report also looks at a “reach” component, defined as the wide
range of industries that are or would be enabled by the space backbone’s solutions and services but
are not a direct product or service of a space provider. Reach services include delivery providers like
Uber Eats that leverage satellite signals, and logistics companies like Amazon that use space data
to monitor and optimize fleet traffic. In this article, we focus on the backbone solutions and services
that are core to the space arena, but we note that the reach component is a critical driver of demand
for the space economy.
In our scenarios, the space industry’s revenues grow from $300 billion in 2022 to $960 billion by
2040 in the lower range of scenarios and to $1.6 trillion in a higher range of scenarios, a CAGR of 7 to
10 percent.175 The commercial segments of the industry may grow between 6 and 9 percent a year
from 2022 to 2040, and the state-sponsored segments may grow at a slightly faster pace, between
7 and 10 percent a year, as countries accelerate investment in the space race.176 In 2023, revenues
were $330 billion (62 percent from the commercial segments and 38 percent from the state-
sponsored segments).
The commercial services and end-user equipment subsegment, including communications and
commercial positioning, navigation, and timing (PNT) solutions, is projected to grow from $180 billion
in 2023 to between $420 billion and $690 billion in 2040 (a CAGR of 5 to 8 percent). Communication
services will likely remain the largest source of commercial revenue, with new constellations
extending the reach of satellites to meet demand for connectivity in remote areas that aren’t covered
by existing technology. Aided by large constellations, such as SpaceX’s Starlink and Amazon’s Project
Kuiper, the data price for commercial satellite communications has decreased rapidly; a Euroconsult
report estimated a 77 percent decline in data prices in the five years through 2024.177 As these
constellations progressively reach full deployment, the decrease in data prices could continue, albeit
at a slower rate. At the same time, lower cost, reduced latency, and greater coverage could increase
the demand for data by 60 percent (in gigabits per second) between 2023 and 2035 as consumers
seek affordable options for remote connectivity.178
PNT solutions are likely to continue to build on free signals from state-sponsored satellites to
provide commercial positioning and navigation. This service could continue to grow as more people
adopt smartphones and sport watches (three billion devices with GPS receiver chips could be
manufactured every year by 2035, up from two billion today).179
The commercial infrastructure and support subsegment is projected to grow from $22 billion in
2023 to between $120 billion and $220 billion in 2040 (a CAGR of 10 to 14 percent) as growing space
investments help accelerate the deployment of commercial infrastructure to support the demand for
space services and as companies continue seeking greater space presence and scale.
This subsegment will likely have two major tailwinds for growth.
First, satellites are becoming cheaper, as technological advancements increasingly allow
constellations to be built from small units at scale. This, in turn, promotes more manufacturing to
enhance connectivity and observation capabilities.
Second, infrastructure players are developing reusable heavy rockets, which can launch satellites
at a lower cost per kilo of payload. Launch costs have already decreased significantly and could
decrease 40 percent more between 2023 and 2035.180 Growth in this subsegment could be driven by
the reusability and increase in size of the vehicles, which improve launch economics.

Arena-
State-sponsored investments, including in the defense and intelligence and civil subsegments,
are likely to remain the cornerstone of the space arena’s expansion over the next 20 years, with
continued growth spurred by incumbents and new entrants. Government agencies are expected to
remain key customers, especially for high-end products and services.
The state-sponsored civil subsegment is projected to grow from $59 billion in 2023 to between
$180 billion and $270 billion by 2040 (a CAGR of 7 to 9 percent) as it benefits from increased efforts
to develop new use cases of space assets. Many organizations have only explored the surface of
space applications and are creating strategies to increase impact. For example, the European Space
Agency is seeking to engage potential partners in three areas: sustainability, economic growth,
and resilience on Earth and in space. New, tailored space applications could enhance research,
risk anticipation, and disaster response efforts. For example, when Tropical Cyclone Idai landed
in Mozambique, the British Geological Society was able to use satellite data to produce a hazard
classification of rainfall-triggered landslides that was then used by civil protection teams to bolster
disaster relief efforts.181
In addition, many countries are seeking to improve their space capabilities. The number of space
agencies has grown from 40 in 2000 to more than 75 today.182 For example, the United Arab Emirates
created a space agency, launched a mission to the International Space Station, and sent a probe to
Mars within a decade. India has also expanded its space presence and became the fourth country
to land a spacecraft on the moon in 2023 after a failed attempt in 2019.183 The country is increasing
its space efforts with the goal of launching a solar observatory and supporting entrepreneurs in the
private space and satellite sector.
The state-sponsored defense and intelligence subsegment is projected to grow from $66 billion in
2023 to between $240 billion and $410 billion in 2040 (a CAGR of 8 to 11 percent) as nations increase
spending to strengthen their national security and autonomy. While the United States spends about
twice as much on space programs as the next ten countries combined, other nations are beginning to
increase their expenditures.184 China, in particular, has achieved an approximately 20 percent CAGR
in estimated space outlays, compared with about a 12 percent CAGR in the United States from 2018
to 2023.185 According to a report by the US Space Force, Defense Innovation Unit, US Air Force, and
US Air Force Research Laboratory, China “continues to compete toward a strategic goal of displacing
the US as the dominant global space power economically, diplomatically and militarily by 2025.”186
These trends will likely add weight to requests to allocate more funding to space programs.
In all four subsegments, varying trajectories of technological evolution and the ability of players in
the space arena to achieve scale and wide adoption could determine whether the market size will be
closer to the lower range of scenarios or the higher one.
In our higher estimate of the space industry’s revenues, we assume that improved access to
data collected from space and the standardization of that data may create new revenue streams,
particularly for software application and analytics players. For example, aggregating and classifying
space data with user-friendly and easily distributable structures could improve access. Data of this
kind could enable use cases for non–space players, such as feeding the data into AI to better manage
a supply chain network. At the same time, cost reductions across the value chain from space-vehicle
manufacturing to ongoing space data reporting may accelerate adoption and the accessibility of the
space economy.
In the lower range of scenarios, we assume that cost curves stay steady as economies of scale
take longer to achieve, hindering expansion of space-related activities. In addition, while examples
of terrestrial alternatives are limited, future technological advances could be viable substitutes
for traditional space-based solutions, affecting continued interest in sectors such as satellite
communications and navigation.

Arena-
The development of new space technologies over the next two decades could unlock revenue
potential in a number of reach industries beyond the backbone segments in the space arena, such
as transportation and logistics.187 In many cases, services that improve product quality and access
as well as drive efficiencies will likely be created or enhanced by leveraging space-based data. For
example, data from satellites that powered the global rise of the ride-hailing service Uber could
enable faster and more efficient last-mile delivery for perishable goods, food, and beverages. We
estimate the range of the potential revenues of these space-enabled applications to be $1.4 trillion to
$2.6 trillion by 2040.
Dynamism
The space arena has drawn increasing interest from the private sector. More than 600 start-ups are
operating in the arena, up from 250 in 2010.188 In the past decade, about 1,800 companies worldwide
have received private equity investments of about $290 billion, according to Space Capital.189 The
top three companies by revenue represented roughly 10 percent of the entire market in 2023.190
Escalatory patterns of investment, however, may lead to barriers to entry that favor early movers:
for example, the first companies to send large-scale constellations to space for communications
would be able to provide broader network coverage more quickly, harness network effects, and
garner greater customer adoption. This race to the top for scale and a presence in space, coupled
with expensive launch and infrastructure costs, could eventually limit the number of companies that
can sustainably thrive on the space value chain (with some exceptions for specific space-based
services). Barriers to entry include the need for substantial investment in R&D capabilities to innovate
and demonstrate capability, compliance with stringent regulations, and the high risk of capital loss if
space projects fail. Besides these dynamics that could limit the number of players, there are a limited
number of buyers in certain segments of this industry.
Market dynamics, however, vary by subsegment.
The commercial services and end-user equipment subsegment presents the lowest barriers to
entry and will likely have the most players. A large number of specialized use cases allows for entry
of new players that cater to specific sectors. For example, data analytics companies could leverage
the use of space monitoring capabilities in a wide range of industries that have distinct needs such
as agriculture and logistics. Data analytics companies with access to satellite data could help a
transportation player enhance its mobility services by improving tracking and management of
vehicles, reducing costs, minimizing downtime, and improving operational efficiency. Spire Global,
for instance, uses a constellation of orbital nanosatellites to track ships at sea, allowing maritime
companies to directly monitor their locations.191 Data analytics companies focused on weather
services can help anticipate and mitigate risks along transportation routes, minimizing delays and
ensuring safety of delivery personnel. Players can leverage a broad array of data and capabilities and
can tailor a unique value proposition to compete, leading to a more fragmented market that continues
to see new entrants.
The commercial infrastructure and support subsegment has fewer players than commercial services
and includes companies such as SpaceX, Airbus, Lockheed Martin, and Blue Origin. Projects in this
subsegment require significant R&D activity to succeed, raising barriers to entry. Companies that can
reduce manufacturing costs and create new capabilities—with technology such as reusable launch
vehicles—will likely gain an edge over competitors by managing launch processes more efficiently.
For example, within the next two decades, the emergence of super-heavy rockets led by extensive
R&D efforts at SpaceX could disrupt the subsegment by significantly reducing launch costs and
lowering the number of launches needed. Super-heavy launches could capture about 70 percent of
commercial launch market revenue, excluding the Chinese market, by 2035.192 The number of players
could decline as small launch providers face competitors that offer lower launch costs using super-

Arena-
heavy rockets. As a result, rocket manufacturing and launch services may become industry segments
where a few large players have a majority of market share.
The state-sponsored civil subsegment also has high barriers to entry. Boeing, Lockheed Martin,
Northrop Grumman, BAE Systems, China Aerospace Science and Technology Corporation, and
SpaceX are examples of companies operating in this space.193 Compared with the defense and
intelligence subsegment, the civil subsegment offers more opportunities for players to enter because
it covers a broader range of applications beyond just defense and intelligence, including scientific
research and disaster management. This creates more diverse spaces for players to occupy, though
civil services would still be government-funded and face strict mission requirements. For example,
private companies other than defense and intelligence contractors can enter the industry via
partnerships with government entities for use cases such as disaster response.
The state-sponsored defense and intelligence subsegment of space has relatively fewer players.
They include Northrop Grumman, Lockheed Martin, Boeing, and Raytheon. Traditionally, larger
players have benefited from a deep understanding of the contracting process and the resources to
navigate it. This advantage could persist for established companies, especially those with cutting-
edge defense technology, strong engineering, and extensive contracting capabilities. However, as
certain market segments become more fragmented, opportunities are emerging for new entrants,
particularly smaller players in specialized niches. Procurement strategies are also becoming modular
and agile, fostering increased competition and new avenues for growth. Moreover, the industry is
shifting from constructing a small number of complex satellites toward larger satellite constellations
with smaller units, which could create opportunities for new entrants.
Swing factors
— How might commercial interest in space continue to evolve and shape future innovation and
commercially viable business cases?
— How could new space data services increase the number of use cases and have an impact on
various industries?
— How would geopolitics affect the structure and distribution of the global market?
— How might the role of government or government-backed organizations evolve? How could
procurement and safety policies on both a national and an international level affect the
subsegments differently?

Arena-
Jeffrey Caso 9. Cybersecurity
The cybersecurity industry protects computer systems—computers, networks, cloud and web
applications, network-connected manufacturing equipment, programs, and data—from unintended
and unauthorized access, modification, or destruction. Unlike other software industries, this one
owes its existence to a set of determined adversaries, including private hackers seeking ransom
and nation-states advancing geopolitical goals. Because those adversaries keep improving their
capabilities, the cybersecurity industry has no choice but to do the same.
Indeed, one indication that the cybersecurity industry could become an arena is that cyberattacks
are becoming more frequent, complex, and costly. Marketplaces for hacking tools and data are
growing. Generative AI (gen AI) has reduced the cost of mounting ransomware and phishing attacks.
In addition, ransomware-as-a-service represents a growing threat. In these systems, developers
provide tools to affiliates who lease them to malicious actors in exchange for a cut of the ransom.
As the technical barrier to entry for cybercriminals has become lower, smaller firms have been
increasingly exposed to attacks. In 2020, the full economic cost of cybercrime, including direct,
indirect, and upstream systemic costs, was estimated to be in the range of $4 trillion to $6 trillion.194
The recent high level of innovation is another indication that the cybersecurity industry could
become an arena. From 2017 to 2021, the number of patents issued increased at a CAGR between
15 and 71 percent in three younger segments of the industry (cloud security, security operations and
management, and cybersecurity for the Internet of Things and operational technology) and between
15 and 20 percent for two more mature segments (data protection and network security). The rate of
patent issuance in technology overall was only about 8 percent.195
Growth
In our modeled scenarios, the cybersecurity industry’s revenues could grow from $160 billion in
2022 to $590 billion in the lower range of scenarios in 2040 and to $1.2 trillion in the higher range of
scenarios, implying a CAGR of 8 to 12 percent.196 The share of all IT spending devoted to cybersecurity
could grow from 6 percent in 2022 to between 7 and 14 percent by 2040.
In our low estimate of the industry’s revenues, cybersecurity spending increases only slightly more
quickly than all IT spending.197 In our high estimate, cybersecurity spending as a share of total IT
spending doubles between 2022 and 2040. Our range of estimates is based on the extent of the
following factors.
First, attacks continue to increase, encouraging companies to spend more on protection.
Second, the digital landscape continues to grow and change, broadening the attack landscape.
Clients use more and more devices and apps, all of which need protection. They also generate more
data, a great deal of which needs to be analyzed by cybersecurity companies to protect sensitive
information. As a result, the responsibilities of the typical chief information security officer become
Note: This section
describes the potential far more expansive.
growth and dynamism of
the cybersecurity arena. Third, as the need for cybersecurity increases, new business models make the technology more
It should not be read as a appealing and accessible to small- and medium-size clients. Traditional consumption-based pricing
comprehensive account of
the industry. To learn more models (which, for example, charge by the number of gigabytes of logs analyzed) work well for
about cybersecurity and companies that have a sophisticated understanding of their needs and that are large enough to
refer to content from the secure volume discounts. Smaller clients could be more likely to use newer pricing models. These
McKinsey Digital Practice
and the Technology, Media &
include outcome-based models, in which the price of a cybersecurity product is based on meeting
Telecommunications Practice. predefined levels of protection, and attack surface-based models, in which the price is based on the

Arena-
number of servers or workloads protected. For these models to be profitable, cybersecurity providers
might have to rely on AI to do much of the work.198
Fourth, cybersecurity-related regulations continue to increase. These often impose requirements on
certain companies that create incentives for them to spend more on cybersecurity.
The extent of the impact of those four factors will help determine the industry’s actual revenues
in 2040. For example, if geopolitical conflicts were to intensify, cyberattacks by nation-states
could increase, boosting the need for governments and companies to spend on cybersecurity. The
achievement of technological breakthroughs in such areas as AI and quantum computing could
broaden the attack landscape more quickly. So could faster-than-expected adoption of the internet
in developing countries.
The cybersecurity industry also faces a major challenge: a global shortage of talent. In 2021, for
example, 3.12 million jobs in the industry went unfilled.199 That shortage could create pressure to use
automation and AI to compensate.
Dynamism
The cybersecurity industry is highly fragmented; its ten largest companies generate less than 20
percent of its revenues. Indeed, the cybersecurity industry is the most fragmented of all the possible
arenas of the future we have identified in this report. However, cybersecurity could begin to see a
few large players develop in individual industry segments over time as businesses scale to remain
competitive.
The fragmentation has many causes. For one thing, the industry is highly diverse. It can be divided
into 13 segments that address a variety of threats in different ways (see sidebar “The cybersecurity
market has more than a dozen niche categories of business”), and those segments can be divided
further into about 70 subsegments. Many of the segments, such as endpoint security and cloud
security, focus on the protection of a particular technology; others, such as consulting, have a
broader purview.
Fragmentation is also the result of differing regulations from country to country.200 In addition,
the role of nation-states in mounting cyberattacks encourages the proliferation of cybersecurity
companies in different countries. Threats are constantly evolving and need new technologies to
counter them. The barriers to entry for new companies are low. However, solutions that work for one
client often need to be suitable for another. For many clients, cybersecurity is a “layered” solution, in
which a variety of different products and services leads to better security. The growing shortage of
workers with cybersecurity skills makes organizations particularly dependent on layered services to
compensate for the lack of talent available.
Nevertheless, a few larger players seem to be emerging as the industry matures, following the well-
established pattern of industries built on technical innovation. From 2012 to 2017, the number of
new cybersecurity companies increased at a CAGR of 10 percent, and the number of M&A deals
increased at a CAGR of 12 percent. From 2017 to 2022, the number of new cybersecurity companies
declined at a compound annual rate of 34 percent, while the number of M&A deals kept growing
(though more slowly than it had during the previous five years). The apparent decrease in the number
of large players could reflect an increase in the scale necessary to compete. For instance, broader
use of AI may require big proprietary data sets and higher up-front R&D spending. Also, vendors
offering holistic solutions that are integrated among multiple cybersecurity applications may have an
advantage, furthering the benefits of scale.

Arena-
Sidebar: The cybersecurity — IoT/OT security. There is an emerging — Network security. This tooling operates
market has more than a dozen market for the security of nontraditional at the network layer to manage the
IT devices, including smart appliances, security and authorization of network
niche categories of business
shop-floor robotics, turbines, and more. resources. It includes technologies
— Identity and access management and This category encompasses the security such as firewalls, which analyze
fraud detection. Organizations use this of programmable logic controllers and inbound packets, as well as virtual
process, known as IAM, to manage, supervisory control and data acquisition private networking (VPN), which allows
provision, deprovision, and adjust user devices. These devices typically prioritize for secure communications within an
access to critical data, applications, and availability for constant runtime and, in enterprise network.
services. It involves tools across the many cases, are legacy technologies
— Web security. This category
value chain and allows both employees (both hardware and software) that were
encompasses tools that protect the
(a process known as workforce IAM) not originally designed with cybersecurity
consumption of web resources as more
and customers (customer IAM, or CIAM) attack vectors in mind.
information and applications become
to securely onboard and verify their — Data protection. This tooling enables the web-facing. This includes content
identity. In the fraud detection use case, data security and privacy of structured filtering in an employee’s browser as well
for example, standard IAM processes and unstructured data, including data- as the network traffic management and
are supplemented with additional data at-rest, data-in-use, and data-in-motion. security of web applications.
such as know-your-customer or KYC It encompasses encryption technology
— Email security and security awareness
programs. and technology that facilitates encrypted
training. This tooling works with a
— Application security. Many organizations communications, as well as tooling
corporate email system to filter out spam,
make their own software for internal and that builds and leverages contextual
identify phishing emails, and build user
external use, and application security is understanding of data criticality in
awareness of threats such as spear
the process by which they manage the adopting dynamic security controls.
phishing that are often email-based.
security of applications. This includes — Endpoint security. This process
— MSSP and consulting. This category
tooling across the DevOps process, provides day-to-day security for user
encompasses managed security
such as application-security testing hardware, such as laptops, mobile
service providers (MSSPs), which are
technology that enables on-demand and devices, and servers. This category
third-party operations that manage
continuous scanning of applications for includes antivirus software as well as
security devices and processes. These
vulnerabilities. more advanced technology that push
services can include providing 24/7
— Security operations. Many companies enterprise security policies to endpoints
security monitoring, specialized skill
have created security operations centers and allow for real-time response to
sets such as advanced forensics for
that serve as the organization’s security threats across the enterprise.
incident response, and outsourcing
nerve center. Tooling in this category — GRC and IRM solutions. Governance, of commoditized services like identity
allows for log collection, ingestion, risk management, and compliance (GRC) provisioning and deprovisioning. In
normalization, and analysis, as well as and integrated risk management (IRM) addition, these providers offer consulting
automation and orchestration tooling that are typically functions of security risk services such as risk advising and
enables incident response. organizations (also known as the “second incident response.
— Cloud security. As organizations pivot to line of defense”) that exist within more
— Implementation. This category covers
the cloud, additional tooling is required regulated industries. These tools enable
the technical setup, configuration, and
to ensure that cloud environments (such the security risk and compliance process
maintenance of the security stack.
as instances, buckets, and containers) by managing and automating workflows
adhere to enterprise security policies. A of controls attestation, tracking
combination of policy management and enterprise risks for centralized reporting,
technical tooling helps enable security and recording risk appetite and related
controls across cloud, hybrid, and decisions.
multicloud environments.

Arena-
In the future, the trend toward fewer large players could continue within each of the industry’s
segments, fueled in part by established technology companies. For example, Microsoft has made
notable acquisitions (including Two Hat, RiskIQ, and CloudKnox Security) and has promised to invest
an additional $20 billion in cybersecurity through 2026.201 Google has pledged $10 billion to advance
cybersecurity solutions over the same period.202 However, there are important factors that could
keep the industry fragmented. The segments perform such different functions that combining them
may not make practical sense. Instead, clients could continue to employ the layered approach and
diversify risk by using multiple providers, calling on the cybersecurity companies whose offerings are
best suited to their particular needs and risk profiles. Furthermore, geopolitical divisions could result
in the shattering of a single global internet into parts, sometimes called the “splinternet,” which could
also discourage industry-wide consolidation.
Swing factors
— As cybersecurity technologies continue to mature, how might new segments emerge, and which
segments could become obsolete?
— How might AI affect the cybersecurity industry? Would it chiefly benefit the attackers or the
companies that combat them?
— How would the emergence of widespread quantum computing, if it becomes a reality, affect
cybersecurity? How quickly could the cybersecurity industry address security challenges
brought on by quantum computing, such as the obsolescence of traditional encryption and the
development of new pathways for attacks?
— How many geographically distinct markets might the cybersecurity industry encompass, and how
might geopolitical developments affect the structure and distribution of the global market?
— How could evolving consumer awareness of digital safety and data security affect the growth
of the cybersecurity industry, for example, if cybersecurity becomes more of a competitive
advantage for consumer internet companies or if public pressure affects policies or regulations
related to cybersecurity?

Arena-
Alexandre van de Rijt 10. Batteries
Batteries are key to the energy transition. The battery industry’s value chain extends from mining
and processing raw materials such as lithium and nickel, to cell manufacturing in gigafactories, to
combining cells into rechargeable battery packs used for electric vehicles, stationary storage, and
other applications. Batteries enable the electrification of transportation and storage of renewable
energy by facilitating the distribution of electricity across space and time. Our analysis focuses
primarily on batteries with lithium-ion chemistries, the most versatile and widely used today due to
their high energy density and overall decline in cost. We also include sodium-ion batteries, which
could scale as alternatives to lithium-ion batteries if lithium prices increase too sharply.
Batteries have come far technologically over the past 30 years. Costs have fallen from thousands
of dollars per kilowatt-hour (kWh) in the early 2000s to less than $100 per kWh today, and energy
density—a measure of the amount of energy a battery can store in proportion to its weight—nearly
tripled in that period.203
Batteries are expected to improve further as they achieve higher energy density, accelerated
charging, and improved life cycles, and reduce their environmental impact (for example, by using
renewable electricity to power the manufacturing process or by sourcing more sustainable raw
materials such as green nickel and cobalt). Some of this continuous improvement results in advances
in battery technology such as the ability to enrich the anode with silicon compounds. Other advances
come from new designs such as solid-state batteries or from new chemistries such as lithium-sulfur
and vanadium redox. These could become direct substitutes for lithium-ion and sodium-ion batteries
in specific applications (for example, to enable longer-duration storage). Older battery types with
lower energy density, such as lead-acid, are excluded from consideration because the focus of
battery innovation and market entry is increasingly shifting to novel and fast-growing segments.204
The following are the three main end-use segments in this market (Exhibit 1):
the potential growth and Electric vehicles (EVs): This segment includes passenger cars, commercial vehicles, two- and three-
dynamism of the batteries
wheelers, and off-highway, railway, marine, and aviation uses.205
as a comprehensive account
of the industry. To learn Battery energy storage systems (BESS): Stationary energy storage systems, including those for
more about batteries and utility-scale, commercial, industrial, and residential storage, make up this segment.206
McKinsey Battery Accelerator
Consumer electronics: This segment includes smartphones, laptops, wireless headphones, cameras,
Team. and other devices that rely on built-in batteries for power.207

Arena-
Exhibit 1
Electric vehicles could account for 84 to 88 percent of the total battery

market by 2040.
Segments of the battery industry
Electric Battery energy storage Consumer

vehicles systems (BESS) electronics
Passenger cars, commercial Front-of-the-meter (FTM) utility- Electronic equipment for everyday
vehicles, 2- and 3-wheelers, scale systems, behind-the-meter use, such as laptops, smartphones,
off-highway, railway, aviation, (BTM) residential, commercial, TVs, tablets, game consoles,
maritime or industrial systems speakers, and headphones
(eg, EV charging infrastructure)
84–88%
of market 11–14% 1–2%
in 2040
Example companies
BYD, LG, Panasonic, Tesla BYD, Tesla, Hithium, Sungrow, Samsung, LG, Panasonic, BYD
Fluence

McKinsey & Company
Because batteries are the enabling technology for the global energy transition, the industry is a
promising potential arena of tomorrow. Decreasing costs, increasing lifetimes, improvements in
energy density, faster charging, and advances in sustainability could expand the range of potential
applications and solidify two large demand pools. The first is the electrification of the world’s fleet
of one billion passenger vehicles. The second is the array of energy storage and power backup
solutions that complement renewable energy generation by smoothing and time-shifting supply
to more closely match the demand profile. Investors have already started to show confidence in
these use cases: venture capital and private equity funds have poured about $42 billion into battery
technology through more than 1,700 deals in the past ten years, with 75 percent of the investment in
2020 and 2021.
However, the industry is facing several headwinds, despite strong growth prospects. Global supply
exceeds capacity, placing downward pressure on prices. This, combined with the high fixed costs
required to build and ramp up a gigafactory, is leading to long-term profitability concerns, which also
make financing more difficult. Strong secular demand trends toward electrification still point toward
long-term growth, but growth could be tempered as margins shrink and financing becomes scarcer.
Growth
In our estimates, the battery industry produced 760 GWh of battery capacity in 2022, increasing
to 12,800 by 2040 in the lower range of scenarios and to 13,700 by 2040 in the higher range of
scenarios, a CAGR of just under 20 percent. The battery industry’s revenues increase from about
$98 billion in 2022 to $810 billion by 2040 in the lower range of scenarios and to $1.1 trillion by 2040
in the higher range of scenarios, a CAGR of 12 to 14 percent.

Arena-
In all three segments—EVs, BESS, and consumer electronics—growth may be powered by both
declining prices (and higher volumes) and increased future innovation. Batteries are still relatively
expensive and make up 35 to 45 percent of the total cost of an EV, presenting a hurdle to the mass-
market deployment of the vehicles and storage systems. Yet prices are decreasing as production
capabilities and capacity improve. Gigafactories are increasing their output and becoming more
efficient.208 The global average battery plant size is expected to increase from 16 GWh in 2023 to
28 GWh in 2030, which is likely to further improve production efficiency. Economies of scale could
enable lower yield loss, faster production-line speed, and more efficient use of labor and energy.
Recent decreases in raw material prices and overcapacity in the cell supply have put more pressure
on cell prices. And developments in new battery chemistries, for example moving from nickel and
cobalt to iron and phosphate, could reduce the use of expensive raw materials. Growth could also
be accelerated by a projected 2 to 3 percent annual increase in battery energy density, which would
decrease the weight of higher-efficiency batteries, their cost, or both (Exhibit 2). Our sizing of the
battery market is based on the projection that average cell cost could decrease from $80–$130 per
kWh in 2023 to $50–$75 per kWh in 2040.209 Prices could fall even lower if nascent technologies
such as solid state batteries accelerate.
Exhibit 2
Technological innovation is driving an increase in battery energy density

even as production efficiencies reduce costs.
Battery energy density and price outlook through 2035

Energy density per kg of lithium-ion batteries, Wh Cost per kWh of battery raw materials, $
600 225
Projected Projected
200
550
175
150
500
125
450
100
75
400
50
350 25
2020 2025 2030 2035 2020 2025 2030 2035
Source: BloombergNEF 2020 Lithium-Ion Battery Price Survey (for 2016 and 2017 data); McKinsey Battery Insights; McKinsey Global Institute analysis
McKinsey & Company

Arena-
For now, the three segments of the battery market are positioned for varying levels of growth,
depending on tailwinds.
EV batteries: This market is projected to propel future growth of batteries and to make up 84 to
88 percent of demand in 2040. Annual production capacity in the EV batteries segment is expected
to increase from 650 GWh in 2022 to between 11,000 and 12,000 GWh in 2040, a CAGR of about
18 percent. Revenues in the EV battery segment are expected to rise from $85 billion in 2022 to between
$700 billion and $950 billion in 2040, a CAGR of about 14 percent. Growth would stem from the projected
evolution of the market for EVs as passenger and commercial vehicles are steadily electrified.
In 2023, EVs made up 18 percent of global new passenger vehicle sales, with the numbers varying
by geography: in China, more than 30 percent of new passenger vehicles were EVs, in the European
Union 16 percent, and in the United States 8 percent.210 According to scenarios modeled by
McKinsey’s Center for Future Mobility, EVs’ share of global passenger vehicle sales grows to between
82 and 96 percent by 2040. In 2023, only 4 percent of commercial vehicles sold were battery electric
vehicles (BEVs) and fuel cell electric vehicles (FCEVs). In the scenarios for 2040, the share of BEVs
and FCEVs among light-duty trucks and light commercial vehicles would increase to between 30 and
80 percent in regions of low penetration and to between 85 and 100 percent in regions of high
penetration, with most of the vehicles likely to be BEVs.
Four main factors are driving battery growth to the higher or lower end of that range: EV cost and
quality, consumers’ willingness to pay a premium for sustainable vehicles and policy incentives,
charging infrastructure, and the supply and distribution of electricity to the charging infrastructure.
The EV market has seen some recent slowdowns, such as delayed electrification targets of some
OEMs, and in 2024 experts revised EV sales growth projections downward for the first time. The
market still displays strong long-term potential, though it is less accelerated than in the vehicles’ early
years. We discuss adoption of EVs in greater detail in the compendium entry on electric vehicles.
BESS: BESS is the fastest-growing battery segment and may account for more than 10 percent of
battery-cell demand in 2040, with global capacity deployment increasing from 50 GWh in 2022 to
almost 100 GWh in 2023.211 That could increase to between 1,600 and 1,800 GWh in 2040, a CAGR
of about 22 percent, according to analyses from McKinsey’s Battery Insights and Energy Storage
Insights solutions. Revenues to manufacturers for sales of BESS could increase from $7 billion in
2022 to between $100 billion and $140 billion in 2040, a CAGR of 16 to 18 percent.
This segment has the potential to enable the renewable energy transition, with growth in both front-
of-the-meter installations and behind-the-meter installations.212 Front-of-the-meter installations—
the utility-scale BESS systems installed on the electric grid to store and dispense energy—account
for the majority, about 65 percent, of the BESS segment.213 Growth in utility-scale BESS is driven
by both the expansion of renewable energy generation from wind and solar and the development of
other energy storage technologies. Greater reliance on renewable energy requires energy storage
that can balance out the natural intraday variability in wind and solar energy generation.214 Utility-
scale BESS is growing rapidly as utilities race to find a storage solution for the growing share of
variable energy production. In 2023, variable renewables like solar and wind accounted for 13 percent
of global electricity generation, nearly double the 7 percent share in 2018. By 2028, that share could
rise to 25 percent.215 Other than storage capacity, BESS also provides ancillary services to support
the grid such as frequency regulation, further driving growth.216
In addition to intraday storage, renewable grids also may require large amounts of long-duration
energy storage (LDES) to meet interday or even seasonal fluctuations in energy supply and demand.
Several technologies are being explored, from thermal storage and compressed air storage
to hydrogen and power-to-gas.217 If battery prices continue to drop as LDES develop toward

Arena-
technological maturity and commercialization, BESS could become a cost-competitive solution for
LDES use cases as well.
Behind-the-meter installations—systems installed at residential, commercial, or industrial locations
for generation, consumption, and storage—account for the remaining 35 percent of the BESS
segment. Although they are a smaller part of the segment, behind-the-meter installations have
the potential for growth, alongside distributed power generation resources such as solar panels in
homes or office buildings. Behind-the-meter systems can store excess electricity generation during
the daytime and help moderate energy costs by reducing peak demand spikes. They can also buy
electricity from the grid when prices are low and sell it when prices are high. Residential energy
storage (namely, smart homes to achieve energy self-sufficiency and optimize self-consumption) also
show growth potential. Technological advances that enhance price, energy density, safety, and ease
of installation will likely help push further adoption by businesses and consumers.
Consumer electronics: This market may make up less than 2 percent of battery demand in 2040,
rising from 50 GWh in 2022 to about 170 GWh in 2040, a CAGR of about 7 percent. This segment’s
revenues could increase from $6 billion in 2022 to about $14 billion in 2040, a CAGR of 5 percent,
driven by continued growth in portable consumer devices. The consumer electronics industry has
been maturing, and incremental advances have slowed. As a result, consumer demand has begun
to shift from new to refurbished products that are more affordable and comparable in capabilities.
Without further innovation and change, it is likely that demand for consumer electronics would grow
at a modest and steady pace, affecting the demand for built-in batteries. That growth in consumer
electronics—and the associated batteries—could be driven by the expanding middle class in
developing countries, such as Brazil, Egypt, India, and Indonesia.218
Dynamism
Today’s battery-manufacturing landscape has a few large players and is largely based in Asia.
This stems from a first-mover advantage that allowed top players to quickly develop economies of
scale and efficient supply chains. As a result, in 2021, the top four cell manufacturers controlled
upward of 60 percent of the battery production market for EVs.219 Japanese and South Korean cell
manufacturers tend to be established conglomerates with decades of experience, while Chinese
manufacturers started producing consumer batteries and expanding into EVs in the 1990s. Battery
producers are still ramping up in Europe and North America.220
The battery industry has high barriers to entry given the large capital requirements to build
gigafactories and the large operational challenges to ramp up gigafactories. These constraints,
together with increasing material cost and tightening global competition, mean thinning margins
that could continue to favor the large incumbent global cell manufacturers. For example, lithium-ion
batteries currently face pricing pressures as EV sales growth has not kept up with the rising global
production and temporary overcapacities exist. In times of tightening margins, smaller manufacturers
may have difficulties competing against larger competitors’ economies of scale, operational
advantages, and ability to take new contracts (and maximize capacity utilization) at a lower price
without sacrificing overall financial performance. While prices naturally peak and trough depending
on the market’s supply and demand, these swings will likely continue to favor industry leaders that are
able to weather market lulls.
Moreover, the market may continue to favor scale because the need for technology know-how
and supply-chain infrastructure stability could lead to a high-cost ticket to play that discourages
would-be entrants and existing players that are unable to achieve sufficient scale. Higher barriers to
entry—such as the need for aggressive cost reductions in the manufacturing and procurement of raw
materials—could benefit the players with the most scale. This could set the stage for escalation of
capital expenditures as players seek to generate greater scale to compete under tight margins. Large

Arena-
players have significant cash flow to deploy robust research teams to develop battery technology,
making it harder for smaller players to compete. In addition, supply-chain challenges give incumbents
a head start. Players would likely need to have the know-how to design batteries and optimize
manufacturing. To achieve scale and win customers, producers could need to secure machinery
capacity, develop long-term supply contracts, and improve product sustainability.
Supply-side factors could also shift global market share from the large Asian players to scaled
competitors outside Asia. China has faced overcapacity issues; for example, gigafactory average
utilization rates dropped below 45 percent in the first quarter of 2023. A utilization rate of 75 to 85
percent is the typical benchmark for profitability.221 This is the result of escalatory competition from
top players that overinvested in production capacity to try to capture future demand. Trade policy is
another important factor shifting market shares in the battery arena, as governments outside Asia
increasingly prioritize local battery manufacturers with policy incentives both for cell makers and EV
customers, as well as through import tariffs on foreign batteries. For example, the 2021 US Bipartisan
Infrastructure Law provided a $3.5 billion boost to domestic battery production.222
And there are opportunities for new entrants. Advances in technology and specialization in new
use cases could allow new entrants to compete against the incumbent battery manufacturers,
which must continually invest in R&D to maintain an edge. These new players could evolve existing
technologies, such as semi-solid-state lithium-ion batteries, or they could develop batteries with
longer durations than traditional ones. It’s not possible to predict whether the new technologies
currently under development—such as longer-shelf-duration, higher-density, lower-cost, and fit-
for-purpose designs in EV power trains—will be able to reach scale or disrupt the sector. Moreover,
new entrants would need to commercialize new innovations faster than existing players to capture
the benefits of this higher-risk technology development. But if these new breakthroughs come
from players other than incumbents, or if incumbents fall behind on the pace of research, space to
compete would open for new entrants with more efficient or lower-cost batteries. These market shifts
could tip the scale in favor of certain chemistries and benefit players that meet market needs.
Swing factors
— How will consumer demand and changes in government regulation, including climate-related
policies, affect the pace of growth in the battery industry?
— Will batteries be the technology of choice for utility-level energy storage, and how much will that
use case grow as countries decarbonize?
— How will a preference for local battery production stemming from geopolitical tensions affect
future supply-chain resilience?
— How might newer battery technologies, such as sodium-ion and lithium-sulfur, improve energy
density, reduce production costs, and ultimately affect industry dynamics, such as the growth rate
by battery segment and the number of new entrants by battery technology?
— How will the sourcing and processing of raw materials change as the industry shifts toward
batteries with smaller environmental footprints?
— What is the relationship between prices and the rate of adoption of battery technologies? How
would events along the supply chain, for example a spike in the price of a raw material such as
lithium or nickel, affect pricing or the adoption rate?
— How will the shift from a linear to a circular industry (use and dispose versus repair, reuse, and
recycle) affect the battery market?
— What are the potential constraints for battery growth (for example, the ability to sustainably
increase supply of critical minerals for lithium-ion batteries)?

Arena-
Co-authors:
Jan Mischke,
Dave Dauphinais
11. Modular construction
With a value of $13 trillion, construction is one of the largest industries in the global economy.
However, despite technological improvements by individual companies, labor productivity growth
in the industry has lagged for decades. Global construction productivity increased only 10 percent
(less than 1 percent annually) between 2000 and 2022, while productivity for the overall economy
increased 50 percent (2 percent annually) and the manufacturing sector’s productivity rose
90 percent (3 percent annually).223 Construction productivity actually declined 8 percent from
2020 to 2022.
The need for transformation is urgent for two reasons. First, there is a global housing shortage and
affordability crisis, especially in populous cities such as Mexico City, Mumbai, New York, and São
Paulo that account for more than 20 percent of the world’s GDP.224 The UN Special Rapporteur on
the Right to Adequate Housing estimated that about 1.6 billion people lacked adequate housing
and 3 billion people could be affected by 2030 due to declining mortality rates, increasing global
population, and limited affordable housing options.225 Second, there is a critical skilled labor gap in
the construction industry, especially in the United States. According to one estimate, for every new
hire in the American construction industry, there may be 20 new job openings for critical roles such
as laborer, electrician, and welder.226 An additional factor, though not as widely acknowledged, is
that the buildings and construction sector accounts for 37 percent of global emissions and faces
increasing pressure to reduce its carbon footprint.227
Industrialized modular construction has the potential to change the game in the massive global
construction industry—which had a market for new builds worth roughly $8 trillion in 2022—by
improving construction productivity. In contrast to the long-held practice of in situ assembly of
raw materials such as cement and timber, modular construction entails building standardized,
prefabricated 2D or 3D modules at an off-site factory. The modules are then assembled into
buildings or infrastructure on the construction site. Two-dimensional modules are typically single-
discipline units such as precast frames and panels, while 3D modules are volumetric, comprehensive
Note: This section describes structures. In this report, we focus on 3D modules, which are better suited to achieve the technique’s
dynamism of the modular productivity potential.
construction arena. It
should not be read as a Modular techniques can reduce the time needed for the planning, design, and on-site installation
comprehensive account of the phases of construction. These phases can be run in parallel, while traditional construction requires
modular construction and construction phases to proceed sequentially (exhibit).228 Modular construction requires fewer
workers than traditional building projects, an attractive prospect in high-labor-cost countries such
McKinsey Operations Practice. as Japan, the Nordics, and Singapore.

Arena-
Exhibit
off-site
McKinsey analysis
McKinsey & Company
However, modular construction has yet to reach scale globally. We estimate that roughly 2 percent
of the $8 trillion market for new builds in 2022 used the technique. In North America, the growth
of modular construction has been slower than projected, at a 4 percent CAGR from 2018 to 2022,
compared with forecasts of 6 to 12 percent. The slower growth can be attributed to inadequate
ecosystem partnerships, including with installation contractors, developers, and financers. Common
challenges encountered by modular firms include overdesign, inefficiencies in installation, unsteady
manufacturing pipelines, and insufficient capital, all of which prevent companies from realizing their
promised time and cost savings.229 By overcoming these challenges, modular construction could
completely transform the global construction industry with a step change in productivity. Because of
this potential, modular construction could become an arena of the future.
But there are trade-offs. First, modular construction factories need stable demand over long periods
to keep busy and running. Second, while module repeatability and ease of assembly reduce on-site

Arena-
complexity and the need for manual labor, the technique also creates greater logistical complications
and requires planning for standardization across various project types, from affordable housing to
single-family homes.
Growth
Experts forecast that the modular construction industry could expand substantially over the next
ten to 15 years, estimating an annual growth rate of between 6 and 10 percent. Growth could be
driven by the increasing global penetration of the process, from about 2 percent of new build in
2023 to a projected 5 to 9 percent in 2040.230 In our modeled scenarios, the modular construction
market’s revenues could grow from $180 billion in 2022 to $540 billion by 2040 in the lower range
of scenarios, a CAGR of 6 percent. In the higher range, market size could reach $1.1 trillion by 2040,
a CAGR of 10 percent, assuming growth continues in regions with higher penetration and less
advanced modular markets catch up.231
To estimate the growth of the modular construction market, we considered the end-to-end value
of an entire project, not just the value added during off-site construction. That is because modular
firms typically act as primary designers. However, there are other business models, such as modular
manufacturing firms that supply premade units but do not install them. This report disregards single-
discipline units, such as precast frames and panels, and concentrates instead on volumetric and
comprehensive structures.
Three critical factors contribute to the high and low ranges of estimates of where the modular
construction industry will be in 2040.
Partnering across the value chain: In the higher range of scenarios, modular companies coordinate
effectively in three major ways. First, they are integrated into project design with developers
and architects. This ensures that project specifications are ideal for a modular system. Next, the
companies collaborate effectively with suppliers and manufacturers to ensure efficient production.
Last, modular players integrate with and lead on-site contractors and logistics providers to ensure
that each component is assembled in the correct sequence. Modular’s potential for parallel execution
of design as well as off-site and on-site manufacturing can be realized only when those processes
are carefully sequenced. Delays can cascade if manufacturing rework, late deliveries, or improper
assembly diminishes the cost and time savings achieved by a modular project compared with
traditional methods. In the lower range of scenarios, these coordination challenges remain an issue in
many countries.
Project selection: In the higher range of scenarios, developers source modules for projects that have
repeatable units, limited customization, and consistent demand. This does not mean that all modular
structures need to look the same. There can be great variation in the aesthetics of modular buildings
and in the benefits that can be achieved if there is standardization in the technical interfaces and
connections. However, bespoke structures such as corporate campuses that need customized
spaces are generally less well suited to capturing the benefits. By contrast, multifamily housing
(especially affordable housing) and dormitories are better able to gain from modular construction.
Growth in the sector could also depend on whether players can establish partnerships with
developers and contractors that ensure a consistent flow of projects, because modular firms rely
on constant demand to scale up off-site production factories, given the high fixed capital expenses
involved. In the lower range of estimates, modular companies have difficulty finding projects that can
achieve the scale to maximize the productivity benefits.
On-site execution: The higher range of scenarios assumes that modular companies successfully
leverage experienced construction talent to manage on-site execution. Modular firms with this
kind of talent on their teams have an advantage. By investing heavily in tech talent to design and by

Arena-
innovating on the right product, modular firms could prioritize project execution talent that can train
on-site crews and manage deadlines. The labor force on construction sites is accustomed to working
with wood and cement. Successful modular firms are those that integrate traditional construction
expertise and experience with technical product development. In the lower range of scenarios, these
execution challenges remain widespread.
Dynamism
The construction industry is both fragmented and localized, reflecting the many stakeholders and
subcontractors involved in any project and the specific construction codes that can vary by location.
In the United States, public records list more than 440,000 homebuilder businesses. The largest
of these, D.R. Horton, employs 14,000 people and accounts for roughly 10 percent of the market
for new homes; the average construction company employs fewer than ten people, and the average
construction project involves more than 100 different suppliers and subcontractors.232
By contrast, as noted above, modular companies can only realize their full potential at scale, given
the high fixed costs of product development and a manufacturing base. Achieving this scale will
likely require forming mutually beneficial partnerships with contractors, developers, and other
stakeholders in the value chain. Better execution and efficient collaboration could create a cycle of
continuous improvement that ultimately could transform the industry.
Even though the industry’s growth potential favors players that achieve scale, the forces that
encourage fragmentation in the overall construction market limit the number of large players.
Different projects have different engineering requirements, equipment and building material needs,
and production techniques. As a result, different players specialize in different types of projects or
parts of the construction value chain. To get the right materials to the right place at the right time,
developers require knowledge of the specific site, and they must be familiar with local regulations and
codes as well as local suppliers and logistics. If modular construction becomes a larger portion of the
global construction industry, these forces will shape the industry’s transformation.
Swing factors
— Despite the continued global housing shortage and skilled construction labor gaps, have high-
profile failures of modular construction firms dampened financiers’ willingness to invest in the
industry? Are national governments likely to intervene to spur the market?

Arena-
Clayton O’Toole 12. Streaming video
The streaming video industry delivers long-form video entertainment over the internet. The global
number of households that used these services rose from 320 million in 2017 to 670 million in
2022.233 By 2040, the number could easily exceed one billion. Over the past decade, escalating
spending on content and technology gave rise to a phenomenon known as the streaming wars, as
companies vied for consumers’ attention and businesses’ advertising. More recently, companies have
faced increased pressure to achieve profitability, leading to reduced spending and a shift toward
the more collaborative business model of content bundling. Although volume growth has moderated
in the past few years, the streaming video market still has growth potential. The global number of
households around the world with broadband internet reached 1.5 billion in 2023, after expanding at
an 8 percent CAGR starting in 2013.234
Media and tech players are now trying to determine which content and platforms will attract
consumers—whether on smartphones, smart TVs, or dongles that enable streaming on TVs—and
how to split revenues. Escalation in this large and growing market explains why streaming video is
poised to become an arena of tomorrow.
Growth
The potential drivers of outsize growth in streaming video include increasing adoption of streaming
technology worldwide, higher spending on streaming services, and more advertising placements
on and revenues from streaming platforms. These growth drivers manifest across the industry’s
three revenue sources: subscriptions; transactions, such as pay-per-view purchases; and ads, such
as YouTube’s spots in its free content. As competition and the cost of content creation increase,
streaming companies are moving toward generating revenues from more than one of those sources.
Currently, there are five main types of companies. The first type is those constructed entirely around
streaming video, such as Netflix. Second are traditional entertainment studios and media and
distribution companies such as Disney and Warner Bros. Discovery that have moved into the streaming
market by expanding their businesses through forward integration. These players realize most of
their profits by offering content on pay TV. Third are big tech companies that have built streaming
businesses mainly to enhance their horizontal franchises, as Amazon did with Amazon Prime Video
and Apple did with Apple TV. Fourth are video-sharing platforms such as YouTube that offer both
short and longer-form content, paid for by subscriptions and ads. While social media companies have
video-sharing features, they have been excluded from this analysis, because short-form content
tends to serve a different purpose and audience than longer-form videos. YouTube and its $29 billion
of ad revenues in 2022 are included, even though the company has a social media component. Lastly,
there are hardware aggregators, such as Samsung, LG, and Roku which make streaming devices or
smart TVs and aggregate content from multiple providers. For example, Roku, whose user base is
larger than the combined audiences of the six largest pay-TV providers, features a set of exclusive
Note: This section describes free ad-supported streaming TV (FAST) channels.235 These aggregators also generate revenues from
the potential growth and advertising, a factor we included in our calculations. (Hardware sales are excluded from our estimates.)
dynamism of the streaming
video arena. It should not In our estimates, the industry’s revenues grow from $160 billion in 2022 to $510 billion in 2040 in the
be read as a comprehensive
account of the industry. To lower range of scenarios and to $1 trillion in 2040 in the higher range, a CAGR of 6 to 11 percent.236 A
learn more about streaming significant portion of those revenues would result from an increase in streaming video’s share of the
video and associated
industries, please refer to overall video entertainment market. In 2022, streaming video captured only 24 percent of the overall
Technology, Media &
revenues in video entertainment (exhibit). In the middle range of scenarios, that share increases to
Telecommunications Practice. 43 percent by 2040.

Arena-
Exhibit
Streaming video could grow to 43 percent of consumer entertainment

spending by 2040, up from 24 percent in 2022.
Revenue share among video entertainment industry segments, middle estimate scenario, %
2022 2030 2040
13
25
41
44 Cable and broadcast TV

and movies in theaters
39
Video games
35 Streaming video
43
35
24
160 350 720 Streaming video revenue,

$ billion
Source: Magna; Omdia; Oxford Economics; McKinsey Global Institute analysis
McKinsey & Company
Three factors underlie the industry growth scenarios:

First, the number of global households that stream video grows from 670 million in 2022 to between
1.0 billion and 1.4 billion by 2040, increasing household penetration from 29 percent to between
34 and 48 percent. The upper range of scenarios assumes that developing markets quickly adopt the
technology. The lower range of scenarios assumes that developing markets shift slowly to streaming.
Second, spending on subscriptions and transactions by each household with at least one paid
streaming service grows from a global average of $10 a month in 2022 to $20 to $40 a month in
2040.237 The upper range of our scenarios assumes that developed markets see rapid growth in
both the number of households with streaming subscriptions and spending on subscriptions. This
might imply not new outlays on entertainment but rather a shift from cable to streaming services. The
higher spending by consumers could also include bundled subscription services, such as Disney’s
bundle of Hulu, Disney+, and ESPN+.
Streaming companies are starting to invest heavily in sports. For instance, Disney+, NBC/Peacock,
and Amazon/ESPN signed a $76 billion, 11-year contract with the NBA for exclusive broadcasting
rights, and Amazon has acquired exclusive rights to live streaming of the United Kingdom’s
Champions League soccer matches.238

Arena-
Subscription growth is not unlimited, however. In a 2023 survey, up to 62 percent of US consumers

reported that they were overwhelmed by the number of video services.239 The lower range of
our scenarios assumes that the number of new streaming households and their spending on
subscriptions are relatively more constrained.
Third, advertising revenues grow at a CAGR of 8 to 9 percent from 2022 to 2040, representing
up to half of the streaming video industry’s overall revenues. Ad-supported streaming video has
grown rapidly: in the United States, it went from 60 million viewers in 2018 to 164 million viewers
in 2023, nearly half of the population.240 Ad-supported streaming video includes both FAST
and ad-supported subscription services such as Netflix or Amazon Prime Video’s ad-supported
tiers. Advertising revenues might continue to increase as ad-supported streaming video
viewership increases. The supply of advertising spots in streaming video would also increase
as more companies turn to ads for revenue growth, especially in developed markets. However,
the ad-supported streaming offerings, which are often free, could pull some viewers away from
subscription- and transaction-based streaming video. Our scenarios assume a varying pace of
growth for streaming video’s share of digital advertising.
Dynamism
The five main types of companies in the streaming video industry—streaming video–first players,
entertainment studios, big tech companies, video-sharing platforms, and hardware aggregators—
currently have just a few large players. In 2023, Netflix took in 27 percent of global subscription
revenues, and the next four largest services (Disney+, Hulu, Amazon Prime Video, and YouTube)
accounted for 27 percent combined. Increased verticalization and regulatory decisions could further
limit the number of key players. However, the industry may continue to see new entrants, including
well-capitalized major media institutions, regional companies, and smaller fragmented creators.
Streaming players have overlapping channels where they deliver their content. For instance, they
compete against one another with long-form content streamed on both TV and mobile phones, a
growing market that accounts for as much as 16 percent of streaming in the United States.241 The
number of hours global consumers spend taking in media is projected to increase at a 1 percent
CAGR from 2022 to 2027, driven primarily by a 7 percent CAGR in digital media usage, led by mobile
video and games.
Some of the lines separating products have blurred. Key streaming video players, such as Apple,
Amazon, and YouTube, are not traditionally associated with media entertainment and participate in
other parts of the attention economy. Also, video—particularly short-form video—is widely delivered
by social media companies such as TikTok and Instagram (part of Meta). For example, in 2024, the
number of subscribers to the user-generated content (UGC) streamed by the online personality
MrBeast surpassed Netflix’s global subscriber base.242 The current market structure was a result
of streaming companies’ efforts to pursue scale, which provides benefits: while fixed costs (mainly
content production and acquisition budgets) are high and rising, the marginal cost of serving
additional customers is relatively low. For example, Netflix achieved positive cash flow only when it
reached 220 million global subscribers.243
The virtuous cycle of benefits that scale brings was observable in the late 2010s, when the
companies with the biggest budgets invested in the most content to attract customers, and the
companies with the most customers invested more in content. This cycle led to a market structure
with a few large players. Nine of the major entertainment studios and media companies in the
industry in 2017 had merged by 2022, creating six bigger companies, while all pursuing scale.
Companies have moved toward both creating content and distributing it, as Netflix, Amazon, and
Apple are doing. Meanwhile, the large studios and media companies have built platforms to distribute
their content.

Arena-
The number of new players has decreased for most of the past decade, while mergers have increased
as companies that couldn’t match the velocity of investment or build better products were unable to
compete and exited the market (an illustration of the escalatory competition described in chapter 2).
More recently, the favorable economics of collaboration has lowered barriers, and companies now
see working together as an option for scaling. In addition, companies face stronger pressures for
higher profitability (Netflix introduced advertising in 2022).244 This has moved the market toward
greater collaboration, which allows players to license content, bundle streaming services, or rebundle
streaming with cable services to save on content creation.
On the other hand, several developments could cause more fragmentation. Regulators could proceed
with antitrust challenges to prevent M&A among existing players, or more companies from other
industries might enter the fray. Additionally, the advent of generative AI radically lowers the barriers
to creating content. It has the potential to reshape the market, including making it easier to produce
higher-quality UGC. Companies including Netflix are spending heavily on AI and machine learning to
help optimize their production, advertising, and mix of channels and shows.245
In addition, some viewers are already watching video-game competitions (commonly referred to as
eSports), and video-game companies could use their capabilities to produce video entertainment
beyond games. Disney has invested $1.5 billion in an equity stake in Epic Games to form a large-scale
entertainment and gaming partnership.246
Swing factors
— How will the industry cover content production costs, and which business models (subscription
based or advertising revenue based) will succeed?
— How will UGC evolve? To what extent will it replace traditional premium video? Will this evolution
vary between short- and long-form content, and how will streaming platforms adapt? How will
the lines separating these various forms of media blur? How will ecosystem players with an array
of monetization options, such as Amazon and Apple, evolve and become different from pure-play
companies such as Netflix?
— How will streaming advertising evolve? What level of adoption will advertiser-supported
offerings receive?
— What will the future bundle look like? Could players bundle with other stand-alone platforms,
which could enhance users’ value proposition? Additionally, how will the shift from cable bundles
to unbundled or bundled streaming affect industry dynamics? Who will be in control: cable
companies, third-party device makers, tech giants, or entertainment partnerships?

Arena-
Andrew Chang 13. Video games
The number of video-game players worldwide increased from 1.9 billion in 2018 to 2.6 billion in
2023. Video games compete for consumers’ time in a broader attention economy that offers leisure
activities ranging from shopping to social media to streaming video (see the entries for digital
advertising and streaming video in this compendium). People born between 1995 and 2010 spend
more time playing games than they do watching TV.247 In 2017, consumers spent only 40 minutes a
week on mobile games. By 2022, they were playing almost 100 minutes a week. That could increase
to 160 minutes a week by 2027. By 2030, roughly 40 percent of the global population could be video-
game players.
Two major trends suggest that the industry could become an arena. First, there has been a surge in
game playing on mobile phones. The launch of the iPhone in 2007 and of Apple’s App Store in 2008
fundamentally changed the landscape and business model for mobile gaming. In 2017, there were
an estimated 1.2 billion mobile gamers. By 2022, there were 1.6 billion. In addition, the mobile gaming
market is still relatively young and has considerable potential for expansion.
Second, free-to-play games have been extraordinarily successful on consoles, PCs, and mobile
phones, generating revenue from large bases of users who make small but frequent in-game
purchases. Spending on games grew from $52 billion in 2017 to $75 billion in 2022, a CAGR of
8 percent. Fortnite, one of the most popular free-to-play games, has 500 million players and
generates billions of dollars in revenues for its publisher, Epic Games.248
Spending on developing console and PC games has also increased sharply. In 2018, the offerings
with the highest budgets and production values, known as AAA games, cost between $50 million and
$150 million to produce. By 2023, AAA games with planned releases in 2024 and 2025 had budgets
of $200 million or more.249 By contrast, the average production budget for the 150 most expensive
movies from 2016 to 2023 was $180 million.250
The video-game industry has recently faced challenges. During the pandemic, limited entertainment
options led to increased consumer spending on games, but projections that revenue growth would be
sustained after the pandemic proved overoptimistic. The global market has been stagnating due to a
reduction in the number of games, challenges in monetization per gamer, recent regulatory changes,
and cooling investor interest. Furthermore, cross-platform play, growth in emerging markets, and
growth in user-generated content (UGC) and indie titles have challenged the model that worked so
well in the high-growth pandemic years. Industry players also face rising competition from vertically
integrated competitors, well-capitalized entrants, and global competitors. Consumer spending fell to
$57 billion in 2022, near prepandemic levels.251
Despite these recent setbacks, the long-term trends surrounding video games are likely to persist.
These tailwinds do not imply uninterrupted growth but indicate that if the industry can overcome its
current challenges, it could capitalize on these trends to become an arena.
dynamism of the video
games arena. It should not Growth
account of the industry. To Outsize growth in video games could result from increases in the number of game players—
learn more about video games particularly as the number of connected middle-class consumers in emerging markets increases—
please refer to content from along with growing advertising revenues and greater spending per gamer as time spent playing
the McKinsey Technology,
increases. Top-performing industry players increase their numbers of users by expanding cross-
Practice and McKinsey Digital. platform capabilities, investing in high-growth markets, enhancing their platforms, and maintaining

Arena-
product quality. To sustain profitability, they also take advantage of alternative monetization methods
and ad strategies, and they use generative AI (gen AI) technology to improve productivity and lower
costs, particularly in game production.
Video games’ revenues of $230 billion in 2022 accounted for 35 percent of the total for the overall
video entertainment industry, behind traditional video (cable and broadcast TV and movies in
theaters) at 41 percent. Video games’ share could increase to 44 percent by 2040, the largest in
the industry, outstripping streaming video as well as traditional video (Exhibit 1).
Exhibit 1
Video games could grow to 44 percent of consumer entertainment

spending by 2040, up from 35 percent in 2022.
Revenue share among video entertainment industry segments, middle estimate scenario, %
2022 2030 2040
13
25
41
43 Cable and broadcast TV

35 and movies in theaters
Streaming video
24
Video games
39 44
35
230 390 740 Video game revenue,

$ billion
Source: McKinsey analysis incorporating data from Magna, Omdia, and Oxford Economics
McKinsey & Company
In our estimates, the video-game industry’s revenues grow to $550 billion in 2040 in a lower-range
scenario and to $910 billion in 2040 in an upper-range scenario, implying a CAGR of 5 to 8 percent.252
The calculation of revenues excludes consumer spending on peripheral devices, hardware with
different uses (such as phones and laptops), community platforms and streaming, and e-sports
(organized video-game competitions).
The 2040 estimates reflect the video-game industry’s three revenue segments: consumer spending
on games, advertising revenue, and sale of consoles.

Arena-
Consumer spending on games. This segment includes in-store and digital sales, in-game purchases
(also known as microtransactions), and subscriptions to game services, such as Xbox Game Pass and
EA Play. It generated about 60 percent of the industry’s revenues in 2022.
Developments affecting the size of the gaming population and the amount spent per player will
determine future spending. To begin with, gamers are becoming more diverse across platforms,
a trend that increases their overall number. This effect is strongest in mobile games, which are
attracting more women and older players: in 2022, an estimated 55 percent of mobile gamers in
the United States were women, and almost 60 percent of adults aged 35 to 54 played at least
once a week.253 Even those who spent the most time playing PC or console games increased
their time spent on mobile gaming. This segment is also growing rapidly in developing economies,
especially in Southeast Asia, the Middle East, and North Africa, as internet and smartphone
penetration increases.
Expansion of the game-playing population also correlates with the growth of the middle class in
developing countries, a group that will have greater access to broadband internet, electronic devices
capable of running games, and disposable income to spend on both. The number of players has
grown especially quickly in Asia, and other developing regions could follow suit. Asia was home to
an estimated 1.5 billion gamers in 2022, and that could grow to 2.7 billion by 2040, a CAGR of about
4 percent, with Southeast Asia showing the most rapid growth. The number of gamers in the Middle
East and North Africa is also rising rapidly. Developed economies may experience slower but stable
growth, though a new wave of expansion in the number of gamers is still possible if innovations, such
as cloud gaming or augmented reality and virtual reality (AR/VR) gaming, become more widespread.
The introduction of UGC, which effectively transforms game players into developers and transforms
developers into publishing channels, is also contributing to the increase in players.254 For example,
Roblox users can create games and experiences on its platform and can play games made by other
users. By the end of 2023, Roblox had nearly 70 million daily active users, including more than half of
Americans under 16.255 And Fortnite’s UGC mode, Fortnite Creative, now has more players than the
game’s more traditional Battle Royale modes.256
Cloud gaming could also increase the amount of time and money that players spend on gaming.
Processing of these games takes place in the cloud, and the resulting video can be streamed to any
device with hardware and an internet connection that is strong enough to render the images. That
lowers the barrier to entry for consumers, gives users access to a wider library of games through
subscriptions, sidesteps the need to wait for large downloads, lowers hardware costs, and allows
users to play PC-quality games on their phones or other devices.
Gen AI has also become increasingly important in game development. It can be used to generate
game content, including voices, dialogue, and personalities for characters. Game developers also
have increasingly partnered with the owners of the intellectual property (IP) of popular franchises,
from movies to manga, which attract large fan bases. In 2023, 43 percent of the top 200 games in
the United States were IP-based. The share was 66 percent in Japan.257 IP-based collaborations
have boosted the number of daily active users of games by 11 percent within the first seven days
of launch.258
The upper range of revenue scenarios assumes stronger effects on the total number of players based
on those factors, and the lower range assumes weaker ones.
These scenarios are also based on assumptions about spending per player. Free-to-play mobile
games with in-app purchases can generate more than twice the revenues of mobile games that only
require a single initial purchase. This higher scenario assumes that free-to-play formats become even
more popular and increase spending per player. It also assumes that subscription models continue

Arena-
to spread, which would spur consumer spending. Free-to-play and subscription games are growing
quickly, but the industry’s revenues would be closer to the lower-range scenario if that growth slows.
In addition, the advent of AR/VR could make games even more interactive. If they are widely adopted,
they could expand the industry, but their high prices could be a barrier to consumers. Indeed, the
technology has already gone through several cycles of hype and bust.
While we do not include e-sports (video games played competitively for spectators) in our sizing of
this industry, they are an important factor boosting the popularity of gaming. The market’s revenues
are projected to increase from $1.2 billion in 2017 to $5.7 billion by 2028, a CAGR of 15 percent.
E-sports betting would contribute more than half of projected overall revenues in that period.259
In 2024, there will be an estimated 286 million e-sports enthusiasts, with an additional 292 million
occasional viewers, up from 194 million and 200 million, respectively, in 2019.260
Advertising revenue. Advertising in games generated about 30 percent of the industry’s revenues in
2022. Estimates of future growth assumed conservatively that these revenues would increase at the
same rate as advertising revenues in other digital media.
But revenues could grow more quickly than estimated for several reasons. For one, developers are
finding more opportunities for revenues from placing ads in games. For another, games played on
mobile phones are the fastest-growing segment, and they are the most likely to use ads: advertising
revenues from mobile games climbed at a 30 percent CAGR from 2017 to 2022. By 2027, advertising
revenues from mobile games could reach $140 billion, exceeding revenues from paid applications
($110 billion) and in-app purchases ($2 billion).
Games remain a far smaller and newer market for ads than, for example, internet search or TV, so
digital advertising has ample room to grow. Ad placements could start to emerge in PC or console
games, not just in mobile games.
Sales of consoles. Sales of consoles such as Nintendo’s Switch and Microsoft’s Xbox generated
just 7 percent of the video-game industry’s revenues in 2022, a share that could shrink over time.
The next generation of consoles is expected to come out in 2029 and could produce a spike in sales
followed by a steep drop. Cloud gaming may even make consoles obsolete for the next generation
by removing the need for a dedicated local gaming machine. The upper range of scenarios assumes
that consoles will maintain their relevance. The lower range of scenarios assumes that players shift to
cloud gaming and games on mobile phones, lowering demand for next-generation consoles.
Dynamism
The video-game industry has a few large players; its five biggest companies accounted for about 40
percent of industry revenues in 2022.261 A sign that the industry may be heading toward even fewer
players is that the number of M&A deals surpassed the number of new companies in 2022 for the
first time in at least a decade (Exhibit 2).262 The following year, Microsoft made the largest deal in
its history with the purchase of Activision Blizzard.263 But games for dedicated consoles, PCs, and
mobile phones, a fast-growing and currently highly fragmented market, are likely to show different
kinds of dynamism.

Arena-
Exhibit 2
The video-game industry could have fewer players if M&A

and new company trends continue.
Number of new companies and M&A deals in the video-game industry

New companies M&A deals
581 576
503
406 398
363 364
325
309
257
182 175
158 158 163 164
147 142
125
104
2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Source: McKinsey Global Institute analysis incorporating data from PitchBook Inc. Analysis not reviewed by PitchBook analysts.
McKinsey & Company
Recently, a few large publishers in the console-games industry used acquisitions of game developers
to scale. Vertical integration of the industry also increased as console manufacturers (particularly
Sony, Nintendo, and Microsoft) acquired game developers and publishers to produce exclusive
content and expand the value of their ecosystems. Vertical integration also allowed console makers
to expand margins, from 5 to 6 percent in the console business and from 25 to 35 percent in game
publishing. The trend toward a few large players could continue if new technologies such as virtual
reality increase costs further. However, gen AI could pose a threat to incumbent game developers by
making it easier for newcomers to create sophisticated games.
A similar trend toward narrowing the field could affect AAA PC games, whose increasing budgets and
complexity have also benefited big developers. Indeed, the biggest developers, such as Electronic
Arts, produce AAA games for both consoles and PCs. Higher costs may prompt companies to
reassess the growth potential of those games, including by examining average user engagement
and purchases. This could reinforce movement toward a few players in this high-end segment of the
industry, potentially benefiting big developers equipped to handle the cost and complexity. But the
market for other PC games is likely to remain more fragmented because developing them is much
easier, cheaper, and more open than developing console games, a process that is tightly controlled by
the console company.264
Mobile gaming may grow faster than console and PC games, and thus exhibit more dynamism. Mobile
games already earn more revenues than console games, increasing dynamism across the entire
arena, and the enormous mobile-phone market gives this segment ample room to expand, with even
faster growth that could raise market share to more than 80 percent by 2040.

Arena-
The market for games played on mobile phones is currently highly fragmented, but it is showing
some indications of consolidation. Small developers can build games for platforms such as iOS and
Android, grow rapidly, and try to take market share from incumbents. Movement among the large
players has already begun; examples include the acquisition of mobile-first developer Zynga by
Take-Two and the purchase of King by Activision Blizzard.265 In addition, there has been a substantial
increase in M&A by mobile native publishers (examples include Playtika’s $1.2 billion M&A budget
and Stillfront’s acquisition of studios to generate economies of scale). While the free-to-play,
microtransaction-based business model has been growing—and with it the quality and quantity
of free-to-play games—the cost of acquiring paying users is also increasing.266 That could lead
developers that are unable to acquire paying customers to be absorbed by more successful
gaming companies.
Swing factors
— Could regional differences lead to supply- and demand-side fragmentation? On the supply side,
could there be regional game developers? On the demand side, as internet connectivity improves
across the world, will newly connected consumers have differing preferences and follow different
usage patterns than their counterparts in developed economies?
— Casual gaming, especially on mobile phones, is driving significant growth in video games. Will
casual games continue to take share from other sources of entertainment?
— Will AR/VR technologies that promise next-generation gaming experiences attain widespread
commercial success? Although the Oculus Rift was first released in 2016, VR games remain
a small part of the market. Will Apple’s Vision Pro and other new hardware bring VR to a
wider audience?

Arena-
Ani Kelkar 14. Robotics
We define robots as programmable machines that automate physical tasks.267 Industrial robots
have been performing a growing number of functions since the early 1960s, when Unimation
installed a robot on an assembly line at a General Motors plant in Trenton, New Jersey, where
it helped make door and window handles, gearshift knobs, light fixtures, and other automotive
hardware.268 Today, technological advancements in robotics are accumulating, lowering barriers
to adoption. Improvements in robot dexterity and mobility have enabled the machines to handle
a greater variety of tasks. Robot-assisted surgery helps doctors perform delicate procedures
with greater precision, flexibility, and control. In agriculture, robots are taking on environmental
problems, such as reducing pollution.269 Restaurant chains have begun to deploy robots to
automate some kitchen tasks. And labor shortages in retail and service jobs create further
incentives for businesses to consider their use.
Technological advances in manufacturing processes and scale have driven down the cost of robotics
hardware, while advances in software technology have improved the ability of robots to work with
humans and enhanced quality through consistency. Advances in generative AI (gen AI) are speeding
the development and training process and enhancing the flexibility of general-purpose and humanoid
robots. For example, Figure has partnered with Open AI to develop AI-powered humanoid robots,
using large language and behavioral models.270 There has been a surge in interest in these machines.
While their resemblance to humans evokes science fiction (and the “uncanny valley” phenomenon),
the practical potential benefit of humanoid robots is that anything designed for people—such
as machinery, work areas, and other spaces or objects–would not need to be retrofitted to
accommodate them.271 That could ease integration of the machines once the technology capabilities
are fully developed.
The robotics industry is benefiting from a large influx of investment capital. From 2018 to 2022,
venture capital investment in the field increased from $4.6 billion to $13.1 billion (a 23.3 percent
CAGR), with much of the funds going to start-ups such as Covariant, an AI robotics company that
raised a $75 million Series C round in April 2024.272 The Chinese start-up Unitree Robotics raised a
$139 million Series B round in February 2024. And established players in other markets are entering
the robotics sector via M&A and investments (for example, Hyundai Motor Group’s acquisition of
Boston Dynamics in 2021, Amazon’s stake in Agility Robotics in 2022, and Figure’s $675 million
Series B round that included investors such as Microsoft, OpenAI Startup Fund, Nvidia, Amazon
Industrial Innovation Fund, and Jeff Bezos).273
As a result of these investments, robots could play an increasing role in our lives. Some notable
recent developments include Amazon’s deployment of humanoid robots in a warehouse in Seattle,
robot butlers providing room service deliveries in hotels, Google’s use of large language models
to develop robots that can help with a broader range of everyday tasks, and the announcement
by Nvidia in March 2024 that it planned to launch a general-purpose foundation model for
humanoid robots.274
dynamism of the robotics Growth
as a comprehensive account The outsize growth potential of robotics would be supported by technological advances and
of the industry. To learn more robotic solutions’ ability to automate physical work. Attaining the financial, social, and regulatory
about robotics and associated
industries, please refer to conditions that would allow mass adoption will also be key factors. Robotics is a broad and complex
Industrials & Electronics
industry that can be understood through robot types, use cases, and the end markets they serve.
Practice. The wide spectrum of robot capability includes agricultural automation, industrial robots in

Arena-
factories, and even autonomous consumer vacuums. The growth of single-purpose robots over
the past 20 to 30 years was largely driven by industrial automation, particularly in manufacturing
environments and increasingly in warehouses. This segment has experienced stable growth, with
an established value chain of original equipment manufacturers, system integrators, and software
players adding capability.
Unimation’s first installed robotics system attempted to be general purpose but required a highly
structured environment where everything had to be in the right place before the robot could be
programmed. More recently, the industry has made substantive progress in the development of more
autonomous general-purpose robots, which can now execute a variety of tasks with limited human
intervention or physical modification of hardware. Through AI and adaptive control, they can even
sense their surroundings, reducing the need for environmental control. Additionally, advancements in
haptic technology facilitate enhanced picking and manipulation.275 Boston Dynamics, Tesla, Figure,
Google, and Sanctuary are all working on autonomous general-purpose robots that can take on tasks
currently performed by people, such as unloading boxes in a warehouse.
To model the growth of robotics as a potential arena, we analyzed the possibilities for automating
physical work activities currently done by people, based on McKinsey Global Institute research on
scenarios of workforce automation and the report The economic potential of generative AI.276 This
is a conservative scope, given that some robots are being used and could increasingly be used to
perform activities that are beyond the physical capabilities of people, such as lifting much heavier
objects than a person can.
The entire robotics industry, which is largely made up of industrial robots, automated guided vehicles,
and autonomous mobile robots, had $21 billion in revenues in 2022.277 In our scenarios, in 2040,
revenues for the entire industry could grow to $190 billion in a lower range of scenarios and to
$910 billion in a higher range of scenarios, a CAGR of 13 to 23 percent.
Our analysis estimated the revenue opportunity for robotics solutions by assessing the technical
automation potential of 658 detailed work activities (DWAs) and three physical performance
capabilities (fine motor skills, gross motor skills, and mobility) in 776 occupations.278 We then modeled
adoption over time, including by comparing the declining cost of technology with hourly wage levels in
each occupation in 47 countries that make up more than 80 percent of the world’s workforce.279
We started with the total potential revenue baseline as the cost of human labor attributable to each
DWA, which is calculated by multiplying volume (number of hours spent on an activity) by the cost
(per hour) of a task. We then estimated the adoption potential of a technologically automated solution
for that activity, based on S-shaped adoption curves for similar solutions, assuming that adoption
of robotics does not begin until its cost is at parity with the cost of human labor. For each activity,
slower and faster adoption scenarios were modeled to range the potential market for automating
that activity. This analysis was aggregated for different points of adoption and different wage rates
across countries.
Our scenarios incorporated the pace and extent of the adoption of robotic solutions according to four
key factors: first, the technical automation potential of physical capabilities; second, the time needed
to develop solutions for activities requiring physical capabilities; third, the economic feasibility of
automation; and fourth, the pace of adoption, which includes differing levels of business investment,
regulation, and social acceptance (exhibit).280
For the first factor, the model assumed that the automation potential of physical capabilities hinges
on whether robotic technology has achieved the proficiency required to successfully execute a given
physical task. Technical automation potential must necessarily precede adoption because robots can
be deployed only after they match or outperform humans in the physical activities being automated.
To gauge the opportunity presented by workforce automation, robotics researchers were surveyed to

Arena-
Exhibit
Depending on the pace of adoption of robotic solutions, 40 percent of

physical work done by people could be automated in less than 20 years.
Share of time currently spent on physical work activities that could become automated, %
80
Early-adoption scenario¹
70
60
50
Midpoint
40
30
20
10
Late-adoption scenario²
0
2025 2030 2035 2040
¹Aggressive scenario for all key model parameters (technical automation potential, integration timeline, economic feasibility, and technology diffusion rates.
²Parameters are set for later adoption potential.
Source: O*NET; McKinsey Global Institute analysis
McKinsey & Company
derive a range of the time it would take robots’ gross motor skills, fine motor skills, and mobility skills
to reach the equivalent of median and top-quartile human performance.281 Our higher-range scenario
assumes these levels of capability are reached at the earlier end of these estimates, while our lower
range of scenarios assumes that they are reached at the late end of these estimates, leading to
delayed automation potential of relevant work activities.
Perhaps surprisingly, minor tasks that are straightforward for people can be an impediment to robots’
ability to work without supervision, lengthening the timeline for full automation of these tasks. For
example, specific actions such as tucking in the corner of a bedsheet present obstacles to developing
fully independent cleaning robots, even those that are capable of performing the majority of skills
needed to clean a hotel room. A robot that can pick an apple from a tree may be unable to harvest
strawberries, because the consistency and shape of the stem of each berry are different. Our higher
range of scenarios assumed that the technological capabilities allowing robots to tackle these
bottleneck skills would be in place by 2040, resulting in higher levels of automation for each task.
The lower range of scenarios accounted for the possibility that the bottleneck skills would delay full
automation, with a correspondingly lower task automation adoption rate.
The second factor was the speed at which robotic products that solve specific problems might be
developed. Our higher range of scenarios assumes that integration of discrete capabilities into
solutions occurs more quickly, decreasing the time to market. Our lower range of scenarios assumes

Arena-
longer timelines to integrate the technologies that transform specific capabilities into solutions that
automate work activities. For example, a robot that can automate harvesting activities would need to
have a wide range of skill sets to both work with a variety of produce and identify the type of produce.
Third, we assume that innovative automation solutions would have to be economically feasible for
users to adopt them. To make the substitution of physical automation for human labor economical,
robots need to improve productivity compared with humans, for example by matching human speed
on a single shift. Namely, we posit that adoption of robots for a given activity could begin only when
the cost of automating that activity is equal to or less than the human labor cost for the same activity
at the same level of skill.
Robotics can provide other benefits besides labor cost savings, such as addressing labor shortages
(Europe, for example, is facing a declining labor supply, with the workforce projected to shrink
by 13.5 million people, or 4 percent, by 2030), increasing throughput, decreasing downtime, and
creating a safer workplace.282
Finally, mass adoption of robotics solutions will likely depend on several financial, regulatory, and
social factors that would influence the rate of deployment. Our higher range of scenarios envisions
accelerated customer deployment of robots as people become more comfortable with working
alongside the machines, investment to create a workforce that embraces automation for productivity
gains, and government regulation in support of robots in the workplace. In addition, our higher
range of scenarios assumes that consumers would fully embrace mass adoption of robots and a
high degree of integration between humans and robots in daily interactions. In our lower range of
scenarios, robots could be deployed at slower rates for several reasons, including social hurdles such
as limited support for retraining workers, financial limitations such as delays in capital investment,
and a lagging and limiting regulatory response that contributes to uncertainty or headwinds for the
industry’s long-term outlook.
Dynamism
The robotics arena is likely to display high dynamism as the use of robots expands from traditional
industrial applications, such as manufacturing and materials handling, to service applications, such
as cleaning, cooking, and maintenance. The dynamism could also be driven by the availability of even
more applications through autonomous general-purpose or humanoid robots. Humanoid robots
could complete difficult or dangerous warehouse tasks; monitor remotely distributed field assets
such as power and telecommunication lines; assist healthcare professionals in physical tasks, such
as moving patients and delivering equipment; and perform risky construction work.283
In the traditional single-purpose industrial robotics segment, long-established players, such as
ABB, Fanuc, and Kuka, are leading a relatively stable, mature market. There is some potential
for competitive shuffling as companies such as Denmark’s Universal Robotics try to create
collaborative “co-bots” that work side by side with people, but generally, the industrial robotics
segment is likely to continue with just a few large players.284 Conversely, newer segments, e.g.,
autonomous general-purpose robots, could see high levels of dynamism. Tech-forward players are
creating several types of machines, including mobile, quadrupedal, and humanoid robots, as well
as high-density storage automation.
Moreover, the convergence of gen AI and robotics could dramatically improve robots’ abilities to
interact with the physical world and perform complex real-world tasks, as a report by McKinsey’s
Operations Practice points out.285 By incorporating AI-based large language models and large
behavior models, robot control systems can help the machines understand and respond to verbal
input as well as mimic human movement. This nascent market features many up-and-coming
disruptors vying to be first to launch a workable product, which requires heavy investment in R&D

Arena-
and a long runway reliant on venture capital funding. Boston Dynamics, Figure, Sanctuary, Tesla, and
ANYbotics are early players in this segment, while several global players are emerging in Asia, such
as Unitree, SinoRobotics, and DEEP Robotics.
The autonomous general-purpose robot market is likely to show high dynamism, but it also is likely to
have fewer large players over time. The massive amounts of research and data needed to develop and
manufacture the machines will likely benefit the players with the deepest pockets and largest scale.
For example, in August 2022, Hyundai Motor Group—which acquired a controlling stake in Boston
Dynamics in 2021 in a deal valued at $1.1 billion286—announced an initial $400 million investment to
launch the Boston Dynamics AI Institute,287 dedicated to developing the next generation of machines
at the intersection of AI and robotics. And, as with other AI-related products, the additional data that
a company with robots in the field can use to make even better robots will likely build barriers to entry
for new players.
Yet even though the wider autonomous general-purpose robotics market has fewer large players,
there may be a burgeoning and much more fragmented market for robots with specific end markets
and use cases. For example, Tortuga uses commercial harvesting robots to help farms automate
harvesting, forecasting, and pest treatment, and Blue Ocean Robotics develops, produces, and sells
service robots for tasks such as cleaning, disinfection procedures, and patient rehabilitation. This
segment could show high dynamism alongside fragmentation, given the huge variety of use cases—
there are as many potential use cases as there are specific physical tasks that humans are paid to
perform. As long as companies can find a viable business model, there can be more new entrants.
Use-case-specialized robotics players and autonomous general-purpose robotics players are
operating in nascent markets that will likely show high dynamism, while the traditional industrial
players operate in a more mature market environment. However, the boundaries between these
segments are not sharply defined. The traditional industrial players may expand their offerings into
some of the use-case-specialized end markets as these markets mature. Over a longer horizon,
autonomous general-purpose robots might replace use-case-specific machines if the technological
development of autonomous general-purpose robots is sufficiently advanced for them to take on a
variety of use cases that are served today by specialized players at lower cost. And there are other
applications of robots beyond our scope of automating human labor, for example deep sea and
space exploration. These aren’t captured in this growth analysis, but to the extent that they are new
applications and new technology solutions, they could develop high dynamism, too.
Swing factors
— What other technologies might accelerate the development of robotics? For example, gen AI is
enabling faster training of robots, better semiconductors are accelerating computing power, and
better batteries are enabling longer activity periods.
— What new tasks will robots be able to perform that are beyond the physical limitations of humans
today? How much productivity might this unlock?
— Apart from physical labor, how quickly will robots be capable of providing emotional labor and
caretaking? Will consumers accept robotic caregivers?
— How might policies related to labor issues affect the adoption of robotics?
— How might potential market fragmentation caused by geopolitical developments lead to a
regionalization of the robotics tech stack?

Arena-
Tom Brennan 15. Industrial and consumer
biotechnology
Industrial and consumer biotechnology is the use of genetic engineering to analyze and manipulate
biological processes at a molecular level to create bio-based products for industrial and consumer
purposes. This field could have transformative effects on society as companies develop revolutionary
products such as drought-resistant crops, at-home genetic testing, and specialty chemicals
manufactured in sustainable and cheaper ways by genetically engineered microbes.
The field has made huge strides over the past 20 years. Several advances have enabled the
commercialization of gene editing, including automated genetic sequencing (to “read” genetic code)
and CRISPR, a technology that can be used to selectively modify DNA (to “write” genetic code).288
Research on CRISPR started in 1987, and the full CRISPR-Cas9 gene-editing system was published
in 2012. The team that developed the technology was awarded the Nobel Prize in 2020. The precision
and applicability of CRISPR-Cas9 was a step change that enabled the wide-scale commercialization
of gene editing today. Using CRISPR and similar tools, scientists can edit genes with extraordinary
precision. This capability could upend established markets or create entirely new demand in many
areas, including cultivated meat and personalized beauty products, among many others.
Biotech has recently encountered investment headwinds: in 2023, industrial and consumer biotech
start-ups raised $2.1 billion, down from $4.9 billion in 2022.289 However, start-up funding in most
industries outside of AI also decreased, with global investments declining 38 percent between 2022
and 2023 (while funding to AI increased 9 percent).290 Despite this recent volatility, there are many
reasons to believe that the industrial and consumer biotech sector has the fundamentals to exhibit
promising growth and dynamism in the long term.
Growth
The drivers of potential outsized growth for industrial and consumer biotechnology are specific to
different segments, but they include demand for more resilient, personalized, and environmentally
friendly products, consumer comfort and regulatory support for genetically engineered products
and associated data collection, and technology advances to boost resource efficiency and bring
costs down. Additional growth enablers include technology advances to accelerate R&D, efficient
commercialization, governmental investments, and demand for sustainable products.
We analyzed four significant biotechnology segments in our sizing of the industry: agricultural
biotechnology, alternative proteins, biomaterials and biochemicals, and consumer products and
services (exhibit). We excluded biopharmaceuticals and sustainable biofuels, both of which we
Note: This section describes discuss in chapter 3. In our estimates, the industry’s revenues grow from $140 billion in 2022 to
dynamism of the industrial
$340 billion in 2040 in the lower range of scenarios and to $900 billion by 2040 in the higher range
and consumer biotechnology of scenarios, a CAGR of 5 to 11 percent.291
as a comprehensive account Agricultural biotechnology. This segment’s revenues are projected to grow from $56 billion in 2022
of the industry. To learn
more about biotechnology to $170 billion in 2040 in the lower range of scenarios and to $230 billion in 2040 in the higher range
and associated industries, of scenarios, a CAGR of 6 to 8 percent.
the McKinsey Industrials
& Electronics, Agriculture,
The pace of agricultural biotechnology’s adoption could depend on three factors. First, its use
Chemicals, Consumer could accelerate if it offers a resource-efficient way of addressing the growing global population’s
Packaged Goods, Life
Sciences, and Sustainability
food needs. Global calorie demand could increase more than 30 percent between 2020 and
Practices. 2050, and production, especially in developing economies, would need to almost double. At the

Arena-
Exhibit
Continued innovation could accelerate growth in four segments of

industrial and consumer biotechnology.
Segments of the nonmedical biotechnology industry
Alternative Agricultural Consumer products Biomaterials

proteins biotechnology and services and biochemicals
Cultured meat; Genetically modified Personalized genetic Bioplastics (eg, PHAs or
microbial protein animals and crops that insights; personalized polyhydroxyalkanoates);
(eg, mycoproteins)¹ help disease resistance, beauty, health, and fermentation-based
yield, etc; selective wellness products chemicals (eg, ethanol;
breeding of crops (eg, DNA-based diets, 1,4-butanediol; lactic
and animals through personalized skincare); acid); bio-based additives
genetic markers microbiome-based to food, feed (eg,
products (eg, probiotics) enzymes, probiotics);
microbial crop protection
10–50%
of market 30–42% 10–29% 10–19%
in 2040
Potential 2040 revenue, $ billion
24–390 170–230 71–170 70–100
Example companies
Upside Foods, Vow Bayer Crop Science, 23andMe, Danimer, Genomatica,

Syngenta PROVEN Skincare Corbion, Novozymes
¹Cultured meat is produced by cultivating animal cells directly. Mycoproteins are plant-based proteins derived from a natural microfungus, Fusarium
venenatum.
McKinsey & Company
same time, there is mounting pressure to reduce agriculture’s land use intensity. This includes
environmental efforts to protect rainforests and other nonagricultural lands, as well as concerns
about the increasing use of agricultural land for nonfood purposes such as biofuels.292 Agricultural
biotechnology could alleviate these pressures by increasing yields. For example, field tests
conducted from 2014 to 2017 in corn-growing areas of the United States showed that genetically
modified hybrids typically yielded 3 to 5 percent more grain than conventional plants, and in some
cases, up to 10 percent more.293
Second, biotechnology can help develop crops that withstand adverse factors such as pests,
droughts, and floods, which could become more frequent in an era of climate uncertainty. For
example, Bacillus thuringiensis (Bt) crops are genetically modified to contain endospore toxins of the
bacterium, making them resistant to certain insect pests. By 2023, adoption of Bt corn in the United
States reached 85 percent, more than double its 40 percent adoption rate in 2006.
Third, consumer resistance to genetically modified foods could slow adoption of biotechnology.
Significant movements in North America, Europe, and around the world are lobbying to restrict or ban
the use of the technology in the food system. Concerns revolve around the potential for allergenicity,
antibiotic resistance, and a spillover of genetically modified plants into conventional crops that could
reduce biodiversity. The European Union has long had strict regulations on genetically modified

Arena-
crops, though there has been recent movement toward permitting techniques such as gene
editing.294 Activists have protested India’s approval of genetically modified mustard, and an attempt
to introduce genetically modified maize seeds to Kenya in 2022 was blocked by legal challenges.295
The policies and regulatory frameworks that result from these growing movements may affect the
adoption of the technology.296
Alternative proteins. This market’s revenues are projected to grow from $2 billion in 2020 to
$24 billion in 2040 in the lower range of scenarios and to $390 billion in 2040 in the higher range of
scenarios, a CAGR of 15 to 35 percent. The middle estimate for alternative proteins puts the market at
$130 billion in 2040. This wide range reflects the market’s nascency and the uncertainty that stems
from its dependence on scientific breakthroughs and major consumer dietary shifts. The estimates
include fermented and cell-based alternative proteins but exclude plant-based products.
Three key factors are likely to determine the segment’s growth. First, the market’s trajectory is
shaped by the increasing share of protein in global diets, which correlates closely with economic
development and per capita GDP growth. According to the OECD and the Food and Agriculture
Organization of the United Nations, meat consumption could increase 2 percent by 2033, driven
primarily by population and income growth in middle-income countries, and 79 percent of total
growth could come from middle-income countries, even as consumption in most high-income
countries is expected to stagnate. The increasing demand for meat, particularly in developing
countries, could create an opportunity for alternative proteins to play an important role in meeting
global meat and protein needs.
Second, alternative proteins could significantly increase their share of total protein consumption
in coming decades. For now, they constitute just 0.8 percent of the overall protein market, though
that share could grow to between 1.0 and 9.0 percent by 2040. However, reaching the higher range
of these estimates would depend largely on overcoming formidable barriers such as consumer
acceptance and taste and texture preferences. The public’s acceptance and widespread use of
alternative proteins could hinge on R&D efforts focused on making products that match or surpass
the taste and satisfaction provided by traditional protein sources. As these challenges are addressed,
alternative proteins would be poised to play a more substantial role in future dietary choices and
global food systems.
Third, price stands out as the most significant barrier to widespread adoption. A 2022 paper found
that in 2013, the cost of producing cultured meat297 in the lab was $2.3 million per kilogram, though
with technological breakthroughs, that cost could be significantly reduced to $63 per kilogram, still
about ten times as expensive as the wholesale price of animal meat.298 High production costs are a
function of both operating costs (for example, growth factors) and inadequate yields (meat produced
per volume of bioreactor), both areas of active research and continued progress. In tandem with
further efficiencies and learning from operating at large scale, advances in these areas can further
drive down unit costs considerably. Crucially, the trajectory of regulatory frameworks, such as the
promotion of carbon pricing and land conservation measures, will heavily influence the comparative
price dynamics of alternative and conventional proteins. These regulations could improve or hinder
the cost competitiveness of alternative proteins, shaping their market viability and adoption rates in
the years to come.
Biomaterials and biochemicals. Biomaterials and biochemicals are substances or materials
produced by microorganisms or created through biological processes such as fermentation and
biocatalysis. These include biopolymers as well as enzymes for use as feed additives or required
for industrial processes. The end-use cases include chemical processing and materials production,
household products, agriculture, food, and nutrition. For example, corporate players and start-ups
have developed and commercialized biotechnology-based additives for use in food, animal feed, and
microbial products to improve yield or nutrition. The biomaterials and biochemicals market’s revenues

Arena-
grow from $41 billion in 2022 to $70 billion in 2040 in our lower range of scenarios and to $100 billion
in 2040 in our higher range of scenarios, a CAGR of 3 to 5 percent. In the higher range, this market
grows and biomaterial or biochemical substitutes would increase their share relative to traditional
materials. In the lower range, the market grows steadily but biomaterial and biochemical penetration
would remain the same.
The market trajectory of the biomaterials industry could depend on two main factors: whether
customers will be willing to pay a higher price for substitutes over traditional (synthetic) materials,
and the extent to which the properties of novel biosubstitutes will be comparable or superior to
traditional materials.
First, biomaterials and biochemicals are often more expensive than traditional materials, which
can affect their market penetration. For instance, biodegradable plastics derived from renewable
biomass sources such as cornstarch, sugarcane, and potato starch are designed to be broken
down using standard procedures. By contrast, traditional plastics are nondegradable, durable,
and persistent, and accumulate in the environment. However, biomaterial substitutes often cost
approximately 20 to 30 percent more because of the higher cost of raw materials and complex
production processes.
Second, biomaterials often face challenges due to inferior material properties that can hinder their
adoption rate. Biodegradable plastics, for instance, are engineered for functions similar to those
performed by products made from polyethylene, polypropylene, and polyvinyl chloride, such as
plastic bags, containers, and pipes, but the biodegradable materials may be less durable or have
other performance limitations. For example, starch-based biodegradable plastics are known to have
poor mechanical properties, hydrophilicity, and low thermal stability, while polylactic acid bioplastics
can exhibit brittleness and elevated moisture uptake.299 These factors influence market dynamics,
shaping both consumer preferences and industry innovation aimed at enhancing the viability
and competitiveness of biomaterial solutions. Moreover, according to a McKinsey report, “not all
biomaterials are created equally,” and the end-user adoption rate of biomaterials will depend upon
the ability of companies to understand the technical differentiation of different products and select
the appropriate end market.300
Consumer products and services. In our estimates, this market’s revenues grow from $36 billion in
2022 to $71 billion in 2040 in our lower range of scenarios and to $170 billion in 2040 in our higher
range of scenarios, a CAGR of 4 to 9 percent.
The market could include a wide variety of goods and services created by industrial and consumer
biotechnology: direct-to-consumer genetic testing; meals and vitamins based on people’s genes;
microbial skin-care products; gene therapy for cosmetic purposes, such as reversing hair loss;
genetic tests and treatments for pets; and perhaps even genetically modified pets. Most of these do
not exist yet, though some do; for example, 23andMe has offered consumers saliva-based genetic
testing since the mid-2000s, and companies are launching subscription-based meal plans and
dietary supplements tailored to customers’ DNA and microbiomes.
The personalized nature of these innovative products and services and their potential to enhance
individual well-being could fuel consumer interest in solutions tailored to genetic or physical profiles,
dietary needs, and aesthetic preferences. For example, companies like Ancestry and 23andMe
have capitalized on the growing demand for personalized healthcare by offering genetic analysis
services that could aid the early detection of hereditary diseases. These factors will likely continue
to drive future demand as customers increasingly seek tailored products that leverage technological
advances to better meet their needs.
The market’s revenues will depend significantly on whether companies can address current
regulatory concerns about the collection of genetic information and create enticing value

Arena-
propositions at reasonable prices. Revenues will also depend on whether companies can accept
the long development times that frequently characterize these kinds of products and services. For
example, businesses are already planning ways to monetize the genetic data collected by some of
these technologies—an endeavor that could prompt privacy and regulatory concerns.
Four key enablers could accelerate growth in all four segments. First, advanced computing, data
analytics, machine learning, AI, and biological engineering are accelerating R&D. For example, in
2023, Google DeepMind introduced GNoME, a new deep learning tool that can help discover new
materials by predicting their stability. The AI tool was used to find 380,000 stable materials that
could hold promise for future technologies.301 Second, efficient commercialization to scale from lab
to mass production is vital. Third, public and government investments, such as the US government’s
$2 billion in funding for biotechnology and biomanufacturing, could further support these efforts.
Fourth, continued corporate commitments to and consumer interest in sustainability will set the
pace of growth in the industrial and consumer biotech market. The market for biomaterials is likely
to expand more rapidly in scenarios where sustainability commitments are emphasized. Conversely,
market growth could be slower if these commitments wane, as price and convenience could outweigh
environmental considerations for both consumers and businesses.
Dynamism
The industrial and consumer biotechnology industry is young, and thus likely to show high dynamism,
as companies seek to develop viable technologies and find sustainable business models for
revenues and profits. For now, the industry is fragmented and contains small, science-focused start-
ups; incumbents in industries that make products that biotechnological advances would replace;
and cross-sector platform companies that provide product-development services such as R&D,
commercialization, and manufacturing. However, it is possible that just a few larger companies will
emerge as the industry matures. An early-mover advantage would give these larger players the
scale and financial reserves to increase their investments in R&D and bring new products to market
faster. Larger companies with early successes would also have the cash and reputation to find novel
biotechnologies by acquiring smaller upstarts, as occurred in the pharmaceutical industry. The
need for high R&D investment could create barriers to market entry and cause a few large players to
emerge over time.
Larger companies could acquire agricultural biotechnology start-ups to bring new innovations to
market at scale. For example, Cardea Bio led the research initiative to develop a biocompatible
semiconductor that could translate real-time multiomics signals into digital information. Cardea was
acquired by Paragraf, an integrated device manufacturer in the semiconductor industry, to accelerate
the mass production of these devices.
However, the direction and pace of scientific advancement are not always linear, and higher
R&D budgets are no guarantee of bringing a successful product to market. The number of larger
companies might also be limited if smaller upstarts continue to make breakthroughs that can be
brought to market relatively easily. In addition, platform players that provide biotech services such
as strain engineering, fermentation optimization, and product development could support a more
fragmented market by providing pathways for product development and commercialization, allowing
smaller players to access the resources required to scale their innovations.

Arena-
Swing factors
— As new biomaterials are developed with improved performance characteristics (for example,
environmental impact, yield, and strength), how fast can they be commercialized at scale?
— Can new biotechnology alternatives achieve the consumer demand or comfort to become
competitive with conventional products such as meat or plastics?
— As biotechnology has grown in scale and importance, consumers in many countries have
expressed concerns such as whether genetic testing companies can keep their genetic data
secure and whether companies should have the right to profit from customers’ genetic results.
How will public confidence evolve?

Arena-
Robin Riedel 16. Future air mobility
A variety of new future air mobility use cases—in transportation of people and goods and in
surveillance—are emerging as advances in technology enable new types of aircraft and operations.
The market segments in the industry include urban air mobility, regional air mobility, last-mile delivery
drones, and supersonic and hypersonic transportation. We focus on two of them: passenger urban air
mobility and last-mile drone delivery. If their makers’ ambitions come to pass, electric vertical takeoff
and landing vehicles (eVTOLs) and delivery drones could become a common sight in many cities,
turning skylines into scenes from science fiction.
The future air mobility industry is fueled by four major trends: the rapid advances of technologies
such as batteries, autonomy, and digital engineering; the global focus on sustainability; the increasing
prominence of shared mobility; and the need to create solutions to traffic congestion. Indicators
suggest that the industry could become an arena by 2040. Annual disclosed funding for future air
mobility grew from less than $600 million in 2017 to $5 billion in 2023, with cumulative disclosed
investment of $22 billion. From 2013 to 2023, about $18 billion was invested in passenger eVTOL and
last-mile delivery drone businesses.302 Moreover, several eVTOLs (such as aircraft from Archer, Joby,
Lilium, Volocopter, and others) are undergoing flight testing for certification, and companies have
already started making last-mile deliveries by drone, including more than one million deliveries in
2023 for last-mile logistics, restaurants, groceries, and healthcare providers.303
Most passenger eVTOLs would typically carry two to six passengers and are generally designed
for intracity trips (up to 50 kilometers) and urban or suburban trips (50 to 120 kilometers). Last-mile
delivery drones are far smaller and typically carry packages lighter than 10 kilograms containing
goods such as groceries, medicine, mail, and retail items. Both eVTOLs and delivery drones are
powered by battery-electric power trains; the drones would mostly take off and land vertically.
Growth
The future air mobility industry’s revenues are almost nonexistent today and could grow substantially
in the coming years, with projections ranging from $75 billion by 2040 in the lower range of scenarios
to $340 billion by 2040 in the higher range of scenarios.304 The industry could see rapid growth
through 2035 due to its nascency, but growth could moderate once the industry achieves some
scale. However, even at scale, the industry would still represent only a small share of the larger
transportation market; for example, in these estimates, passenger eVTOLs would account for the
equivalent of less than 1 percent of all trips undertaken via ground transportation.305
We measure the industry’s revenues as the amount the service operators would take in. We
estimate that 20 to 35 percent of the total revenues of passenger eVTOL operators could be paid to
manufacturers (with much of that paid to suppliers of components, such as batteries, airframes, and
aircraft systems), and 10 to 40 percent would be allocated to suppliers of mobility platforms (including
software for ticketing, scheduling, and navigation), licensing, financing, insurance, and air traffic
management.
dynamism of the future air The growth of the industry will likely depend on three main factors: price, customers’ preferences,
mobility arena. It should not and the rate of technological advancement in battery performance and autonomy.306 However,
account of the industry. To two additional factors could limit that growth: regulation and the development of necessary
learn more about future infrastructure.
air mobility and associated
Price: For the industry to grow toward the higher end of our range of revenues, the price of an eVTOL
Center for Future Mobility. trip or drone delivery would have to become competitive with that of other modes of transportation.

Arena-
For example, the median price of a regular ride-hailing service in 2022 varied highly by location,
from as low as $0.50 per kilometer in much of South and Central America to as high as $3.50 per
kilometer in New York and $4.30 per kilometer in Bern, Switzerland.307 Our estimates suggest that for
eVTOLs, the price once the market reaches scale could be between $2.00 and $4.50 per passenger
per kilometer, driven by a combination of component costs, operational costs like pilot salaries and
maintenance, infrastructure costs, and the economics of round-trip traffic, charging, maintenance
times, and peak-hour utilization.308 This suggests that eVTOLs could offer a service that would be
competitively priced compared with premium ground transportation, especially in areas where
ground transportation is expensive. And even at a higher price, some customers might be willing to
pay for passenger eVTOLs if they are more convenient or faster than the alternatives.
Last-mile delivery drones may also become cost competitive. Current unit delivery cost by drone
is higher than delivery by van, especially if the van is carrying multiple packages. However, this is
partly because in some jurisdictions, regulations permit a person to operate only one drone at a time
within visual line of sight. But if an operator could oversee 20 drones beyond visual line of sight,
drone delivery costs could become comparable to van delivery costs, according to our estimates.
Completely automated drones could drive those costs down even further.309 Minimizing drones’
footprints (by optimizing their use of airspace or changing drone size) and maximizing throughput (by
improving utilization) could also affect unit delivery cost by drone.
Customers’ preferences: A McKinsey survey of about 5,000 respondents indicated that although
meaningful numbers of customers would consider eVTOLs for commuting, running errands, business
and leisure travel, or getting to the airport, many were concerned about aircraft safety and trip
price.310 The survey also found that time savings was by far the main reason most respondents gave
for considering eVTOLs, suggesting that people in cities with severe traffic congestion or geographic
obstacles might be more open to using them. Current users of limousine and ride-hailing services
were also likelier to say they might use passenger eVTOLs.
Customer experience will help determine the extent to which eVTOLs are adopted. The aircraft might
save time, but car travel also has advantages.311 For example, taxis and ride-hailing services operate
on demand, whereas some passenger eVTOL flights are expected to follow a schedule. In addition,
most eVTOL passengers could still need to arrange ground transportation for the first and last mile
of their journeys. Reliability might also be an issue, because aircraft require more maintenance and
stricter safety protocols than cars, and their operations could be subject to weather cancellations.
The future of last-mile delivery drones also depends on how well they can meet a variety of customer
needs and overcome the public’s hesitation. On the one hand, drones could reach remote areas with
poor roads and could circumvent traffic congestion in cities, satisfying customers’ preference for
delivery speed and convenience. According to a McKinsey survey, 56 percent of consumers would
choose drone delivery if it was faster than traditional delivery and cost the same. On the other hand,
McKinsey research also suggests that the safety of drone deliveries, the noise they produce, and the
potential invasion of privacy are of concern to consumers.
Battery technology: While companies have already addressed many of the key technical challenges
for eVTOL flight, batteries, in particular, pose a large challenge. Technical challenges remain in areas
such as battery density (how much energy and thus range of flight a battery can supply per unit of
weight), cycle life (lifespan before replacement), recharge time, and performance in cold weather,
when batteries underperform. For example, enabling longer regional commutes via eVTOLs could
require higher-density batteries.
Moreover, for eVTOL services to reach profitability, their batteries would have to be able to charge
and discharge more quickly than electric car batteries and have longer lifetimes. Similarly, while
current battery technology is able to support drone use cases involving flights up to 60 minutes,

Arena-
last-mile delivery drones would benefit from further improvements in battery technology that enable
longer flights. The good news is that battery technology is advancing rapidly (see the “Batteries”
entry in this compendium). Growth in future air mobility will depend on the speed of technological
advancement in battery technology and the cost at which batteries can be provided.
Regulation: For eVTOLs to scale up, regulations allowing and governing their deployment will be
necessary. In 2023, the US Federal Aviation Administration (FAA) announced the steps needed to
achieve a competitive, scaled-up air taxi market by 2028.312 The European Union Aviation Safety
Agency has been developing a set of conditions for the certification of new passenger eVTOLs.313
Some companies, such as AutoFlight and EHang, have already received airworthiness certification
in China.314 But fully building out a mature eVTOL-specific regulatory environment may take time.
Some eVTOLs could be operated autonomously one day, presenting a whole new set of regulatory
considerations.
Regulatory standards for delivery drones are also starting to evolve. For instance, Australia has
approved Wing for drone deliveries in North Canberra, and Logan and Swoop Aero for drone
deliveries within 60 kilometers of bases in Toowoomba and Goondiwindi.315 Meituan has been
approved in China and made more than 100,000 drone deliveries in Shenzhen in 2022.316 In addition,
SF Express received a license for delivery drones in January 2022 and has established more than
60 routes covering areas in Longgang, Nanshan, Luohu, Longhua, and Pingshan. In the United
States, the FAA authorized operations beyond visual line of sight at seven test sites in September
2023 for UPS Flight Forward and Zipline. Since then, the number of companies with specific
approvals has grown sharply and now includes Amazon, Wing, and DroneUp, among many other
operators, though they are mostly operating at a small number of facilities and locations. Authorities
in the European Union and the United Kingdom have also made progress.317
Infrastructure: In some cases, passenger eVTOLs could take off and land at existing regional airports
or heliports. But larger-scale passenger operations would require dedicated boarding and loading
facilities, where the eVTOLs would also be parked and serviced. More important, establishing
physical infrastructure such as takeoff and landing spots along passengers’ preferred routes will
be crucial. As detailed in a McKinsey article, without a robust network of facilities, known as ports,
“flying-vehicle transport could follow a pattern similar to that seen in today’s helicopter market, where
the number of potential destinations is limited.”318 These networks would require land and investment.
In addition, existing facilities would need retrofitting to accommodate the charging infrastructure
needed for eVTOLs. Investment in more extensive facilities could follow as the market takes root
and demand grows. An urban network of facilities with 85 to 105 landing pads could cost between
$230 million and $300 million to build and between $165 million and $195 million annually to operate.
There are already some plans in North America and Europe to start developing such networks, and
about 20 ports were under way in each region as of 2022.319
Delivery drones also require some specialized infrastructure. However, since it can be built with
smaller footprints in areas outside dense urban centers and the facilities may need fewer or smaller
landing pads, the challenges for building physical infrastructure aren’t as limiting. At the same time,
advancements in digital infrastructure, such as more sophisticated unmanned traffic management
(UTM) systems, are key to realizing innovative drone delivery applications. Advancements in UTM
technology that enable sophisticated drone tracking and coordination with air traffic control systems
can help enable higher-altitude drone flights than are currently feasible.320
Dynamism
Because the industry is nascent, any growth would constitute a major share shift, increasing the
likelihood that future air mobility would qualify as an arena. In the meantime, it is possible to glimpse
the contours of the industry that could emerge.

Arena-
While passenger eVTOLs are still in their infancy, the prospect of scaling them up could be attractive
to new entrants, especially as the vehicles spread to new regions. Once the market reaches a certain
size, economies of scale and customer acquisition would probably lead to consolidation of market
share among leading players as providers drop out, much as we have seen in the ride-hailing industry,
as players without significant capital backing exit the market.
It is unclear whether the most successful business model will be a vertically integrated one, in
which eVTOLs are built and operated by the same company, or a more traditional supply chain, in
which manufacturers build eVTOLs and sell them to operators. If vertically integrated companies
manage to operate more efficiently and beat out smaller companies, the industry could favor fewer
large players. But if the ability to operate in heterogeneous markets promotes a structure in which
manufacturers sell eVTOLs to operators, the industry could instead become more fragmented,
with a smaller set of manufacturers but a large set of local operators. Business models may also be
unique to different regions and their regulatory and cost landscape, so a one-size-fits-all solution
may not prevail globally.
Prospects for last-mile delivery drones are somewhat easier to observe in the short term because
they are already operating commercially, unlike eVTOLs. Some companies are doing trials, and others
are already scaling up operations to deliver medical supplies, retail purchases, and food. In the most
common business model, drone manufacturers operate last-mile delivery services in partnership with
retailers, medical organizations, or third-party logistics providers. For example, in March 2024, Wing
and the food-delivery service DoorDash announced a pilot program in Virginia, following a 2022
pilot in Australia that has since grown to include more than 60 restaurants and other merchants.321
Walmart has partnered with four drone delivery companies and established 36 stores as drone
delivery hubs.322 In addition, major retail and logistics companies are developing in-house drone
delivery technology; Amazon Prime Air is an example.
In the long term, the drone-delivery market could have a small number of large players regionally
or even globally, depending on the extent to which different regions require different operations.
Business models may also evolve as the market matures and companies develop deep expertise and
specialization. Since the hardware is likely to become commoditized, companies that can operate
scaled-up networks cheaply, optimize logistics, and satisfy complex regulatory requirements could
have an advantage. Scale will help delivery drone operators build these advantages, conferring a
first-mover advantage and raising barriers to entry for new players.
Swing factors
— What steps might regulatory bodies (such as the FAA) take to enable the industry to operate at
scale (for example, greater coordination with industry)?
— Can passenger eVTOL companies build a sustainable business model and achieve profitability,
and what timeline would let them avoid remaining an expensive niche product?
— How will public acceptance of these innovations evolve and affect the market, especially when
it comes to concerns about safety and noise? Would consumers be willing to pay a premium for
future air mobility over other modes of transportation?

Arena-
Josh Sternberg,
Corina Curschellas
17. Drugs for obesity and
related conditions
As the world grows more prosperous and people live longer, the already large impacts of chronic
conditions such as cardiovascular disease, chronic obstructive pulmonary disease (COPD), cancer,
diabetes, and brain diseases (including Alzheimer’s disease) are expected to increase.323 In the
United States alone, the number of people over 50 with at least one chronic disease is projected to
double from 72 million in 2020 to 143 million by 2050.324 Globally, chronic diseases were causes or
contributing factors in 75 percent of deaths in 2010 and 79 percent in 2020; by 2030, experts predict
that they could be a factor in up to 84 percent of global mortality as more countries develop.325
A 2010 report from the Harvard School of Public Health and the World Economic Forum calculated
the global economic burden of chronic diseases by using a cost-of-illness approach that includes
both direct costs (those associated with diagnosis, treatment, and care) and indirect costs (such
as lost income). The report found that cardiovascular disease, cancer, COPD, diabetes, and mental
illness cost a combined $6.25 trillion globally in 2010. Using an alternative approach, the report
found that the total lost economic output due to reduced labor and capital inputs caused by chronic
disease mortality could cumulatively reach nearly $47 trillion from 2011 to 2030.326 This economic
impact is surely eclipsed only by the human cost of these diseases, which diminish both longevity
and quality of life. These chronic conditions require a strong response aimed at “adding years to life
and life to years.”327
Obesity, the prevalence of which has skyrocketed over the past 20 years, has been directly
linked to some of these chronic conditions, contributing to type 2 diabetes, heart disease, and
some cancers.328 A report by the World Obesity Federation calculates that the economic impact
attributable to obesity alone was $2.0 trillion in 2020 and is projected to increase to $4.3 trillion
by 2035.329 A McKinsey Health Institute (MHI) analysis calculated the obesity-attributable disease
burden measured in disability-adjusted life years (DALYs), which combine the years of life lost to
premature mortality with years of life lost to states of less than full health or of disability.330 One DALY
represents the loss of the equivalent of one year of full health. According to MHI, 153 million global
DALYs could be attributed to obesity in 2019. Many of these DALYs were due to obesity’s connection
to chronic conditions: 41 million from ischemic heart disease and 34 million from type 2 diabetes.331
The increasing prevalence of chronic diseases and obesity has led pharmaceutical companies into a
race to develop more effective therapeutics that could bring about step changes in the treatment and
management of these conditions.332 One class of drugs, glucagon-like peptide-1 agonists (GLP-1s)—
alone and in combination with gastric inhibitory polypeptides (GIPs)—appears to have the potential
to greatly change the trajectory of obesity-related ailments. Starting in 2021, these drugs (originally
developed to treat diabetes) began receiving approvals for treatment of obesity as well. The results
have been striking: sales of GLP-1-based therapies such as Novo Nordisk’s Ozempic reached
Note: This section describes $24 billion in 2022 alone, and the drugs have proved so popular that manufacturing supply has not
the potential growth and kept up with demand.
dynamism of the drugs
for obesity and related The first wave of GLP-1s has already driven substantial growth for drugmakers. In the seven years
conditions arena. It should not
be read as a comprehensive since Novo Nordisk launched Ozempic as a drug for type 2 diabetes, the company’s value has
account of the industry. To increased sevenfold to a market capitalization greater than $490 billion, and it has become the
learn more about drugs
for obesity and associated world’s 12th-largest company and Europe’s largest, surpassing giants such as LVMH.333 Eli Lilly,
with a market capitalization of more than $800 billion, recently became the world’s most valuable
Life Sciences Practice. pharmaceutical company, partly as a result of its GLP-1 agonist, Mounjaro/Zepbound.334 In addition,

Arena-
the opportunities in the market have already drawn potential entrants: both Amgen and Pfizer had
GLP-1 weight loss drugs in Phase II trials as of June 2023, while AstraZeneca and Roche concluded
multibillion-dollar deals to secure weight loss drugs currently in development.335
Some studies suggest that these drugs could have broader uses, including the treatment of other
chronic diseases. According to a 2019 study, GLP-1 agonists could help reduce inflammation, lower
blood pressure, and produce neuroprotective effects, all of which are correlated with a reduced
prevalence of Alzheimer’s disease, hypertension, and fatty liver disease.336
The rapidly growing market for obesity-related drugs is poised to expand further as pharmaceutical
companies race to develop new versions of the GLP-1 agonists. Moreover, GLP-1s are just one class
of drugs that could improve outcomes related to obesity. As the pharmaceutical industry invests
in medical solutions across the range of chronic diseases, the potential for more breakthroughs
indicates that these therapies are likely to become an arena.
Growth
While nascent demand across the range of chronic diseases is high, we focus on the obesity drug
industry—which has already seen a breakthrough—as a real-life example of what could happen with
treatments for chronic diseases. We anticipate that growth in this industry and its rapid technological
innovations may derive from consistent demand due to rising rates of obesity and other chronic
ailments, the price sensitivity of pharmaceuticals, and rates of patients’ adoption and treatment. We
measure the industry’s revenues as the total global sales of GLP-1s and other obesity drug therapies
for patients who are diagnosed with obesity and are prescribed the drugs as part of their treatment.
Some of these patients may use the drug in its injectable form, and others could eventually take an
oral solid pill if companies are successful in developing one.
In our modeling, the market could grow from $24 billion in 2022 to $120 billion by 2040 in a lower
range of scenarios and to $280 billion by 2040 in a higher range, a 9 to 15 percent CAGR.337 This
growth rate would imply a market size of $50 billion to $70 billion in 2030 assuming continuous
growth from 2022 to 2030, though the market may grow faster until 2030 than it would from 2030
to 2040. Faster growth through 2030 seems possible, and projections from Goldman Sachs, JP
Morgan, and others have estimated that the market could exceed $100 billion by 2030.338
The key driver of the range of market projections is the price of obesity drugs. While our analysis
assumes that improved clinical outcomes, reduced manufacturing and production costs,
increased competition, and the loss of exclusivity will drive down prices over time, the extent of
the decrease, along with its associated volume impacts (the number of patients accessing obesity
drugs), will determine the size of the market. This price sensitivity is due to the unique nature of the
pharmaceuticals market, where prices are not solely the outcome of a supply and demand market
equilibrium but are driven by a country-specific mix of intellectual property, healthcare regulatory and
reimbursement policies, and local competitive dynamics.
The adoption by the public of new GLP-1-based drug therapies or other innovative formulations is
fueled by the limitations of existing obesity and weight loss therapies. The new GLP-1 agonist-based
drugs such as Ozempic—which seem to emulate the effects of hormones that make patients feel
sated, though their exact mechanism is still being researched—have transformed the expectations
of weight loss efficacy, with some medications promising reductions in body weight of more than
15 percent.339 That is nearly double the efficacy of previous GLP-1 weight loss drugs, such as Novo
Nordisk’s liraglutide-based Saxenda, which was associated with less than 10 percent weight loss.
Rising rates of obesity point to a growing global need for effective treatments. Obesity already affects
about 16 percent of the global adult population (defined as those age 20 and over) and is projected
in the World Obesity Federation’s 2023 Atlas to affect about 25 percent of adults worldwide by

Arena-
2035.340 To calculate the 2040 obese population, we used the obesity rates in the World Obesity
Atlas projected for 2035 along with the medium fertility variant from the UN’s World Population
Prospects for the total adult population by region.341 This showed that more than 1.5 billion adults
could be affected by obesity in 2040, in an estimated global population of 9.2 billion. This scenario
is segmented by regional estimates of the prevalence of obesity, from North America and Latin
America, where nearly 50 percent of the population could experience obesity in 2040, to Asia and
Oceania, where only 13 percent and 17 percent of the population, respectively, would experience the
disease in 2040.
Our scenarios for possible market sizes in 2040 consider the following factors: the rising prevalence
of obesity globally, the annual cost of drugs per patient, and the subsequent obesity diagnosis and
drug treatment rates. To account for price elasticities, our approach explores two price scenarios and
makes some assumptions about the volume implications for each. We start with price, which could
be the biggest factor determining how widespread these drugs become because there is so much
latent global demand for obesity treatments. The number of patients with access to the drugs could
expand significantly if the annual cost per patient is relatively low. Other interdependent factors could
affect this market’s shape in 2040. The trajectory of research into new formulations—for example,
the development of an effective oral solid obesity pill rather than the current injectables—and these
treatments’ effects on related conditions like heart disease, along with regulatory outcomes and
competitive moves by players in the market, are impossible to predict but will play a large role in
shaping the future of these drugs globally.
However, we can imagine a scenario in which improved clinical outcomes, reduced manufacturing
and production costs, increased competition, and the loss of exclusivity drive down the price
of obesity drugs over time, in which case the addressable market would go from niche to truly
mainstream. Wegovy, the leading GLP-1 agonist prescribed for weight loss, is currently listed at
about $16,000 a year in the United States, though the net price received by the manufacturer after
discounts and coupons is about $8,400, according to a 2023 study by the American Enterprise
Institute.342 The cost of these medications could decline as a result of improvements to production
and innovative formulations (think oral solid pills).
As many pharmaceutical companies research and release new obesity drugs, competition in the
industry is likely to increase, which would result in lower prices. The Centers for Medicare and
Medicaid Services may choose Ozempic (which is not approved for weight loss) for price negotiations
under the Inflation Reduction Act.343 Patent protection for the branded injectable GLP-1 agonist
compound used in Wegovy and Ozempic is set to end in the early 2030s, when generic producers
will step into the market, increasing downward pressure on prices. How low the price could fall may
depend on how much producers can lower their production costs, as well as the number of new
entrants and the intensity of competition in the market. Outside of price and its impacts on volume,
other factors also affect diagnosis and treatment rates, including but not limited to impediments to
physicians prescribing the drugs, limited coverage for obesity therapies, and concerns about the
costs and side effects of long-term obesity treatment.
In the higher range of scenarios we model, prices drop radically, to an average of $500 a year per
patient worldwide, across all available obesity medications. We chose $500 as a comparable price
to other widely used drugs treating chronic conditions, such as statins,344 and slightly less than
AARP’s estimate of the average annual cost of therapy for a widely used generic drug at $679 per
year.345 In this scenario, it is likely that pharmaceutical companies would succeed in further reducing
the costs of production and that there would be a diverse market with branded and generic GLP-1s
competing for patient access. In this scenario, we assume diagnosis and treatment rates similar to
those for heart disease patients who take statins and similar medications to manage blood pressure.
We model a scenario in which approximately 75 percent of North Americans with obesity would take

Arena-
the drugs, along with 50 percent of Europeans and 20 to 50 percent of populations in Asia, Africa,
Latin America, and the Middle East, leading to a total of about 550 million obese adults being treated
globally (just under 40 percent of the global adult obese population).
In this scenario, the resulting market size would be $280 billion in 2040, with many patients taking
the drugs at a lower cost. A number of other assumptions would also need to be true to achieve
this market size: manufacturers would need to scale up massively to produce doses for 550 million
patients a year globally, and they would need to remove the current production bottleneck in the
fill-finish production step. In addition, competition would have to be sufficiently intense to bring
down the price from today’s level, healthcare payers worldwide would have to recognize the drugs’
benefits and offer coverage, side effects would have to be mild enough to promote widespread
adherence, and there would have to be widespread recognition from patients and doctors that
obesity is a treatable condition.
The lower range of scenarios assumes that the blended price of obesity therapies (blended across
all branded and generic options) in the United States in 2040 would fall, but not as low as $500 a
year per patient. Instead, the scenario assumes a price of $2,500 a year per patient, similar to the
average annual cost of current diabetes medications, while each patient in most other regions would
pay about $900 per year for the same treatment.346 This also assumes that 57 percent of the global
obese population is diagnosed347 and that of those diagnosed, about 20 percent of obese patients in
North America and Europe would take these drugs, while about 10 percent of those diagnosed would
take them in the rest of the world.348
Altogether, these numbers point to a total global patient population of about 100 million obese
adults (just under 7 percent of the global adult obese population) using the drugs. In this scenario,
the market size for obesity drugs in 2040 could be $120 billion, growing at a 9 percent CAGR from
today. The higher price and lower volume assumptions in this scenario could be a result of some
combination of the following four contributing factors, among others:
— Continued challenges prolong current production bottlenecks.
— There is no widely available small-molecule oral pill.
— Some patients end treatment as a result of unpleasant side effects and other causes.
— Healthcare payers are willing to offer some coverage, but not at the same rate as in the lower
price scenario.
While $120 billion to $280 billion in revenues would require sustained growth and massive increases
in patient access, these drugs could prove a relatively cost-effective intervention in the context of
global pharmaceutical spending considering their potential healthcare impact. Global spending on
drugs reached $1.48 trillion in 2022 and is forecast to increase 7 percent a year until 2028.349 While
we cannot know the exact trajectory of the drug market in the future, if this growth rate continues
through 2040, $120 billion to $280 billion of obesity drugs would reflect less than 5 percent of global
pharmaceutical spending, less than the relative market size of oncology drugs today (currently about
12 percent).
Dynamism
The introduction of new obesity drug therapies has already begun to shift market share toward the
first movers in the space while simultaneously pushing incumbents in other pharma submarkets to
develop their own obesity therapies. This new area of competition could reset the S-curve for obesity
treatments and offer incentives to new players to enter the field.

Arena-
The early commercial success of GLP-1s is triggering escalatory dynamics as pharma majors
invest in R&D to develop the next generation of branded drugs. As with other drugs, GLP-1s require
substantial up-front investment due to unpredictably long development and approval timelines. Given
the likelihood of increased pricing pressures once generics are introduced, current players will likely
be encouraged to develop new branded drug therapies with higher efficacy, fewer side effects, or
less intrusive ingestion methods—for example, oral solids replacing injectables.
However, with more than six applications already filed for generic semaglutide, a large number of
generics producers will likely emerge after Novo Nordisk loses exclusivity of its first-to-market
compound by the early 2030s in major markets such as the United States and Europe. That milestone
would drive the market toward relative fragmentation. While pharmaceutical companies may try to
develop second- or third-generation branded obesity treatments, the drugs would have to show
superior performance to be preferable to low-cost generics with clinical profiles similar to those
of existing options. And even if these new second-generation drugs succeed in the market, the
introduction of generics based on the first-generation drugs could exert considerable downward
pricing pressure on the market.
The discovery of step-change treatments for other chronic illnesses would likely set off a similar
escalation of R&D investment, with major pharmaceutical companies entering the arena, followed
by fragmentation with the entry of generics producers. For example, 1,600 treatments and vaccines
for cancer were in development in 2023, a more than twofold increase from 2014.350 As the global
burden of chronic diseases becomes increasingly significant, the potential impact of pharmaceutical
solutions increases, too.
Swing factors
— Researchers are investigating whether GLP-1-based therapies can be used to treat other
chronic diseases. If that research proves successful, how much more widespread could
GLP-1 treatments become?
— As the prevalence of obesity increases globally, what impact might the global recognition of
obesity as a disease by medical stakeholders (governments, public health agencies, payers,
and R&D companies) have on the market in price and volume?
— The convenience and tolerability of the drugs likely will have a large impact on their uptake.
Can research identify options for obesity treatments that are more convenient and have fewer
side effects?
— If considering the broader set of chronic conditions that make up a large share of the healthcare
burden, what could be the impact of developing similarly efficacious therapies for other
chronic conditions?

Arena-
Chad Cramer 18. Nuclear fission
power plants
The nuclear fission power plant industry dates to the 1950s. It encompasses the construction of
power plants that leverage nuclear fission to produce energy, including traditional large-scale
reactors (LSRs) as well as small modular reactors (SMRs) and microreactors, which are more recent
technological advancements. Nuclear fission works by splitting an atom into two smaller atoms,
releasing energy in the process. Fuel sources for nuclear fission today include uranium-235,
plutonium-239, and thorium-232 isotopes. Nuclear power can generate one million times more
energy than fossil fuels per mass. Today, nuclear fission reactors are a leading source of electricity
generated without emissions of greenhouse gases.351
As the demand for energy rises with economic development, electricity generation is projected to
roughly double from 2022 to 2040.352 In addition, with efforts to decarbonize, the energy mix could
significantly change as energy users switch to low-emissions sources from emissions-intensive
sources such as coal and other fossil fuels. Nuclear power has high potential to fill these energy needs.
It is already able to provide stable baseload electricity to the grid, unlike solar, which cannot produce
energy at night, and wind, whose generation depends on variable weather patterns. Furthermore, the
rapid expansion of solar and wind power could slow down because of factors such as the scarcity of
land for solar and wind farms, the time that it takes to build the necessary transmission infrastructure,
and the difficulty of scaling up electricity-storage capacity quickly enough.
Both private investors and governments in many markets are driving momentum in support of
the expansion of nuclear generation. Investors are showing interest, even though the industry
has high capital costs and investments take a long time to produce returns. Global venture
investment funding of nuclear companies shot up from $60 million in 2018 to $390 million in 2022.
Companies are also facing higher electricity demand caused by the increasing prevalence of
artificial intelligence (AI) and supercomputing, alongside sustainability goals. Nuclear power offers
a potential solution to meet these needs. For example, Microsoft is considering nuclear energy to
power its AI data center operations.353
Several countries are starting new nuclear programs or expanding existing ones. The US Department
of Energy has awarded about $3 billion to license, build, and demonstrate two next-generation
plants, and the Inflation Reduction Act provides either an investment tax credit of up to 50 percent
or a production tax credit of up to $36 per MWh for the first ten years of a new plant’s operation. In
2022, the United Kingdom announced a $150 million fund to support new nuclear projects.354 South
Korea plans to increase the share of nuclear power in the country’s energy mix to 30 percent or more
by 2030.355 In Finland, after the Olkiluouto 3 nuclear plant went into operation, the contribution of
nuclear power to total electricity generation increased from 33 percent in 2021 to more than 40
percent in 2023.356 Adding to this momentum, 25 countries collectively made commitments at the
COP28 climate summit in 2023 to triple global nuclear energy capacity by 2050.357
dynamism of the nuclear Growth
fission power plants arena.
It should not be read as a Nuclear fission’s growth may depend on decreasing the capital expenditures and time needed
comprehensive account of the to build large reactors and SMRs as technology advances. Growth could also depend on the
nuclear fission power plants, decarbonization paths followed by countries, which will each require varying amounts of the
the McKinsey Global Energy &
emissions-free “firm” energy production that nuclear power can provide.358 The industry’s revenues
Materials Practice. are projected to grow from $18 billion in 2022 to $65 billion by 2040 in a lower-range scenario and

Arena-
to $150 billion by 2040 in a higher-range scenario, a 7 to 13 percent CAGR. A further breakthrough

scenario could be possible if the cost of building nuclear drops significantly, leading to a market
size of $400 billion in 2040, a 19 percent CAGR from 2022. These figures include revenues for
the builders of nuclear fission power plants and do not include the cost of operations. The revenue
estimates factor in existing operational capacity net of suspended and permanently shut-down
facilities, capacity under construction, and countries’ commitments to achieving net-zero emissions
by 2050.
In the overall energy system, global electricity generation in 2022 was 27,000 TWh. Ten percent of
that output was from nuclear sources, about 30 percent came from other lower-emissions sources,
such as solar and wind, and the remaining 60 percent was from higher-emissions sources, such
as coal and natural gas. In our estimates, nuclear power could generate between 8 and 19 percent
of global electricity by 2040.359 Under a breakthrough scenario we modeled in which capital
expenditure per megawatt significantly drops, nuclear power could generate up to 43 percent of
global electricity. Historical precedent also suggests that nuclear power could play a larger role in
the energy mix than the 10 percent it provides today. Nuclear power underwent a construction boom
from the 1970s to the early 1990s, when its share of total power generation globally increased from
2 to 18 percent.360
The contribution of nuclear power to the energy mix depends on three factors: building costs,
whether solar and wind technologies become uninterrupted based on advances in storage and
transmission, and the decarbonization scenario. First, capital expenditure for plant construction
is a main driver of energy cost for nuclear power. Historically, the price ranged from a lower end of
$2 million to $3 million per megawatt—for the P4 reactors built in France in 1984 and the APR-1400
reactor built in South Korea in 2010—to a higher end of $5 million to $13 million per megawatt for
reactors built in the United States.361
As more countries build nuclear power plants and the global market grows, learning-curve efficiencies
could push capital expenditure lower for large reactors, SMRs, and microreactors. For example, the
Barakah nuclear plant in the United Arab Emirates achieved a 32 percent reduction in cost with each
doubling of deployment volume.362 Additional factors specific to LSRs, SMRs, and microreactors also
apply. For example, SMRs and microreactors that can be assembled in factories from prefabricated
components benefit from being constructed out of small, repeatable, and standardized components at
scale, lowering production costs. Furthermore, SMR and microreactor construction involves repeating
many smaller construction projects rather than managing one much larger project, enabling quicker
learning iterations. The adoption and standardization of SMR and microreactor technology could also
reduce costs, affecting the number of nuclear power plants that would be built. In the case of LSRs,
improved economies of scale might drive long-term cost-effectiveness.
Second, the amount of firm power363 required by a lower-emissions grid is an important factor. A
less carbon-intensive grid that relies more heavily on renewables would also need a baseline of firm
power to balance out energy demand and ensure reliable service. Currently, nuclear technology and
hydropower are the only proven options for providing emissions-free firm energy production. Looking
ahead, advancements in long-duration storage for solar and wind, geothermal energy, and carbon
capture of emissions from fossil fuel power sources like natural gas could offer potential alternatives
for supplying emissions-free firm power. Technology advancements in these sources will affect their
cost, performance, and public acceptance compared with nuclear power, and thus affect the future
energy mix.
Finally, the decarbonization path that countries choose significantly affects electricity demand,
generation levels, and the energy mix because decarbonization involves the electrification of major
energy use categories such as industrial processes, transportation, and heating and cooling. Under
current policies, which may not be sufficient to achieve net zero, electricity generation is projected to

Arena-
Exhibit
The number of new nuclear power plants depends on their cost relative
to the cost of other sources of energy and decarbonization scenarios.
Nuclear energy market size, $ billion
2022 2040
Three scenarios Lower growth Higher growth Breakthrough
400 Capital expenditure/MW, $ million
6–8 4–6 Up to 4
Decarbonization scenario
Countries maintain Countries Countries

some of their maintain stronger maintain stronger
momentum toward momentum toward momentum toward
150 decarbonization, decarbonization, decarbonization,
and/or other sources and nuclear and other clean
of firm clean power becomes a major power sources
65 become widely used. source of firm are not scaled up
clean power. as planned.
18
Incremental annual capacity, GW
CAGR, 2022–40, %
10 30 100
8 13 19
Potential number of reactors added annually
Share of global electricity
generation, %
Up to 8 large Up to 26 large Up to 83 large
8–11 14–19 33–43 or 31 small or 100 small or 330 small
Source: McKinsey Energy Insights Global Energy Perspective 2022; McKinsey Power Model; Net Zero country commitments, IAEA PRIS; McKinsey
Global Institute analysis
McKinsey & Company
reach 50,000 TWH by 2040. In a scenario with more targeted policies and increased commitments,
including from new adopters of nuclear power, electricity generation could reach 59,000 TWh.364
Higher-range and lower-range scenarios incorporating different assumptions for the three factors
above yielded the market size and rationales illustrated in the exhibit above.
In the lower-growth scenario, industry revenues would expand from $18 billion in 2022 to $65 billion
in 2040 at a CAGR of 8 percent. The capital expenditure per megawatt underlying this range falls at
the higher end of estimates, between $6 million and $8 million. Under the scenario, 10 GW of capacity
is constructed annually. This pace of construction assumes a mix of the following assumptions: that
all nuclear plants already under construction or planned to be built by 2040 will be completed, that
countries will maintain a similar pace of construction, and that countries follow through on some or all
of their commitments to net zero.
This scenario also involves building up to eight large reactors or 31 small reactors annually, bringing
nuclear capacity to 550 GW or 5,000 TWh by 2040. In that case, nuclear power could generate
between 8 and 11 percent of global electricity, depending on the total energy produced under
different decarbonization scenarios, similar to the 10 percent share it generates today.365

Arena-
In the higher-growth scenario, industry revenues would expand from $18 billion in 2022 to $150
billion in 2040 at a CAGR of 13 percent. The capital expenditure per megawatt in this scenario is
below the figure for the lower-growth scenario and falls between $4 million and $6 million. This
scenario assumes that 30 GW of capacity is constructed annually. Growth hinges on favorable
developments of the three factors mentioned earlier: the adoption of newer technologies such as
SMRs to reduce costs and enhance safety, increased reliance on nuclear power as a substantial
source of firm power compared with alternatives, and more ambitious decarbonization goals through
support from both the public and governments.
Moreover, this scenario relies on sustained commitment from countries deeply engaged in nuclear
power, particularly China, which plans to build 150 new nuclear reactors by 2035. It has 27 under
construction, far exceeding Russia, the second-largest builder, which currently has only four
reactors under construction.366 The scenario also anticipates increased adoption of nuclear power by
countries with minimal current capacity. It involves building as many as 26 large reactors or 100 small
reactors annually, bringing nuclear capacity to 940 GW or 8,000 TWh by 2040. If those conditions
are met, nuclear power could generate between 14 and 19 percent of global electricity, depending on
the total energy produced under different decarbonization scenarios.367
That scenario aligns with the historic peak of 30 GW added per year in the 1980s. By contrast, a
breakthrough scenario could produce even faster growth. For example, capital expenditure per
megawatt could decrease significantly, possibly to less than $4 million per megawatt and closer to
the $2 million to $3 million per megawatt cost of South Korea’s APR-1400 reactor, making nuclear
power cost-effective compared with other energy sources. Other emissions-free energy sources
could fall short in supplying and distributing enough electricity at the right time to meet demand.
There could be continued momentum in achieving global commitments for decarbonization by 2050.
Nuclear energy could take on a larger role in off-grid power, heat, desalination, and AI data centers.
Our breakthrough scenario could lead to a market size of $400 billion by 2040, and a 19 percent
CAGR from 2022. This scenario entails the addition of 100 GW added per year and involves building
as many as 83 large reactors or 330 small reactors per year, resulting in 2,200 GW of nuclear energy
capacity or 19,000 TWh by 2040. In this scenario, nuclear power could generate between 33 and 43
percent of global electricity, depending on the total energy produced under different decarbonization
scenarios. While this would represent a historically large share of the mix for nuclear power globally,
the share in some countries, such as Finland, France, and Switzerland, already exceeds 30 percent.368
Other factors influencing the growth of nuclear power include readiness of the supply chain to
secure fuel sources and meet stringent standards for multiple nuclear components, availability of
labor for plant construction, complexities in construction that may cause delays, and the ability to
deal with nuclear waste. The amount of waste generated through nuclear power generation is small,
but it remains highly radioactive; the nuclear industry has so far been successful in managing this
issue. Government support can also affect growth through regulations and subsidies. The current
regulatory framework applies to large reactors. The higher range of scenarios assumes that these
regulations are adjusted to enable SMR adoption. Governments have historically used public
subsidies to pursue nuclear power.
Dynamism
The nuclear fission power plant industry has a few well-established players with large market shares,
although new entrants are making inroads with new technologies. The high R&D spending required
for the design and construction of nuclear reactors, along with strict regulatory and licensing
requirements, favors large companies with big market shares, such as Hitachi, Westinghouse, and
China’s State Nuclear Power Technology Corporation. At the moment, there are about 120 owners
and 90 operators of nuclear power plants globally. Although the market for reactors is international,

Arena-
nuclear power players tend to build only in countries to which they have geopolitical ties. Those
factors limit the extent to which market share could shift. However, there is potential for larger players
to extend to developing economies, particularly countries without major nuclear-energy companies.
The lengthy development process of next-generation nuclear fission power plant technology might
also favor existing companies with large R&D budgets. New technologies require a long R&D phase
to build experimental “lab-scale” reactors that prove scientific feasibility. Then companies must
demonstrate a small-scale proof-of-concept reactor with sufficient stability of design. After these
rounds of testing, the companies need to build the full-scale reactor to validate performance before
full commercial deployment.
Nonetheless, newer entrants are developing the technology and could further scale it or sell it to
utility companies. For example, TerraPower raised $830 million in late-stage venture capital (VC)
funding in 2022.369 That was supplemented by more than $2 billion from the US Department of
Energy to build a nuclear power plant near a retired coal facility in Wyoming.370 In Sweden, LeadCold
is developing a Gen IV microreactor and received $2 million in late-stage VC funding, with an
additional $11 million from the Swedish Energy Agency. Both companies have plans for plant site
deployment in the late 2020s or early 2030s. Companies that are able to successfully develop newer
technologies with advantages such as lower production costs and enhanced safety will likely have
a significant first-mover advantage in capturing market share and could collect higher returns as
front-runners in commercializing the technology. However, larger legacy players are likely to retain
significant market presence given their scale, existing development contracts, and experience with
government funding.
Swing factors
— How will global energy demand shift in the future as a result of policies related to emissions and
other factors?
— How might technological advancements such as long-duration storage for solar and wind,
geothermal energy, and carbon capture of fossil fuel power sources such as natural gas affect
firm power supply and the uptake of nuclear power?
— How fast might companies accelerate learning-curve efficiencies in nuclear reactor construction
to reduce capital expenditure for nuclear plants, and how sensitive would demand be to such
price shifts?
— How much will governments support R&D and construction for nuclear power compared with
other renewable energy sources? How will this backing vary across countries based on broader
industrial policy? How might public perceptions about the safety of nuclear power evolve and
affect public and government support for nuclear power?
— How active will large-scale off-takers—such as data-center hyperscalers and high process-
heat demand manufacturers, for example—become in the funding, construction, and potential
ownership of new nuclear power plants?

Arena-
Acknowledgments
The research was led by Chris Bradley, drugs for obesity and related conditions, Kelkensberg, Gayatri Shenai, Hejin Kim,
McKinsey Global Institute director and Josh Sternberg, a partner in the Greater Humayun Tai, Jakob Fleischmann, James
senior partner in the Sydney office; Kweilin Boston office, and Corina Curschellas, a Kaplan, Jamie Vickers, Jeffrey Algazy,
Ellingrud, MGI director and senior partner partner in Zurich; and for nuclear fission Jeremy Schneider, Jesse Klempner,
in Minneapolis; Michael Birshan, senior power plants, Chad Cramer, an associate Jonathan Tilley, Jordan Bank, Jose
partner in London; Michael Chui, a senior partner in Columbus. Luis Blanco, Jochen Latz, Kabir Ahuja,
fellow at QuantumBlack, AI by McKinsey; Kersten Heineke, Kevin Wei Wang, Lars
The project team was led by Alexis Yumeng
and Kevin Russell, an MGI senior fellow in Hartenstein, Laura Millroy, Lei Xu, Marc
Yang, Giorgi Suladze, Peter Kaufman,
Charlotte, North Carolina. Brodherson, Martin Kellner, Martin
Sophie Chinchilla, and Suhayl Chettih. We
Linder, Mike Friedman, Nick Inchastegui,
We give particular thanks to McKinsey are grateful to Max Berley, senior editor,
Philipp Kampshoff, Quentin George, Ryan
colleagues who co-authored write-ups and Benjamin Plotinsky, senior editor, who
Brukardt, Russell Hensley, Sari Varpa,
exploring the 18 arenas of the future: for helped write and edit the report.
Timo Möller, Ting Wu, Veronica Retana,
e-commerce, Roberto Longo, a partner in
The project team was supported by and Will Ardron.
Milan; for AI software and services, Ben
Alexa Wright, Anna Chopp, Aurola Qin,
Ellencweig, a senior partner in Stamford; In MGI’s publishing team, we would like
Chaeyoon Kim, Daniel Paletz, David Randl,
for cloud services, Bhargs Srivathsan, to thank Rachel Robinson, Rishabh
Delfin Villafuerte, Edward Kalaydjian,
a partner in the Bay Area; for electric Chaturvedi, and David Batcheck. For data
Emily Chen, Emmanuel Daniel-Aguebor,
vehicles, Moritz Rittstieg, a partner in visualization, we would like to thank Chuck
Feyi Edun, Finn Wang, Geoffrey Gao,
Chicago, and Patrick Schaufuss, a partner Burke. In communications, our thanks
Grace Kamara, Jose Betancourt,
in Munich; for digital advertising, Craig go to Rebeca Robboy, Nienke Beuwer,
Katharina Marquat, Lorenzo Flores,
Macdonald, a partner in London, and Lia Shannon Ensor, Ashley Grant, Crystal
Nine Chantawasinkul, Onyx Bengston,
Grigg, associate partner in Denver; for Zhu, and Cathy Gui. Thanks also go to
Richard Yang, Robin van Aeken, Rodrigo
semiconductors, Nikolaus Lehmann, an McKinsey’s design team, especially Diane
Garcia, Shania Robinson, Shekhar Kumar,
associate partner in Munich; for space, Rice, Nathan R. Wilson, Stephen Landau,
Shreyvardhan Sharma, Thiago Cersosimo,
Andrew Sierra, an associate partner in Janet Michaud, and Sean M. Conrad. We
Venassa Omoruna, Will Lepry, Yixuan
Chicago; for cybersecurity, Jeffrey Caso, are also grateful for the collaboration
Chen, Yoana Sidzhimova, and Zander
a partner in Washington, DC; for batteries, of McKinsey’s digital production team,
Cowan.
Alexandre van de Rijt, an associate partner including Mary Gayen, Katie Shearer, Dana
in Stuttgart; for modular construction, Jan For kindly sharing insights, we thank Sand, Charmaine Rice, and Regina Small.
Mischke, a partner in Zurich, and Dave adviser Matthew J. Slaughter, Dean
As with all MGI research, this work
Dauphinais, an associate partner in the of the Tuck School of Business,
is independent and has not been
Greater Boston office; streaming video, Dartmouth College.
commissioned or sponsored in any way
Clayton O’Toole, a partner in Minneapolis; Many McKinsey colleagues gave us input by any business, government, or other
for video games, Andrew Chang, an and guidance. We want to thank: Aamer institution. While we gathered a variety
associate partner in Tokyo; for robotics Baig, Alex Singla, Alok Singh, Anna of perspectives, our views have been
and shared autonomous vehicles, Ani Stepien, Arun Arora, Ben Ellencweig, independently formed and articulated in
Kelkar, a partner in the Greater Boston Bill Lacivita, Brad Herbig, Brian Gregg, this report. Any errors are our own.
office; for industrial and consumer Brooke Stokes, Cary Mei, Chandra
biotechnology, Tom Brennan, a partner in Gnanasambandam, Chhavi Adtani,
Philadelphia; for future air mobility, Robin Chris Donahue, Deston Barger, Erik
Riedel, a partner in the Bay Area office; for Johnson, Erik Sjödin, Erlend Spets, Freek

Arena-
Endnotes
12 25
Because many companies fit into multiple arenas, Data from PitchBook Inc. Analysis was not reviewed
the market caps listed here are weighted to reflect by PitchBook analysts.
26
the splitting of market cap across multiple lines of Cody Slingerland, “101+ cloud computing statistics
business or business units. For example, the largest that will blow your mind (updated 2024),” CloudZero,
market cap would be $2.3 trillion if we counted Apple December 2023.
as one unit rather than dividing it, as we have done,
into computing hardware and consumer electronics.
See the technical appendix for more detail on our
Chapter 3
weighting methodology. 27
We use after-tax profit margins to refer to net
13
Home regions are defined by the location of the operating profit less adjusted taxes (NOPLAT). The
Chapter 1 company’s headquarters. margins we considered were based on observed
1
NOPLAT margins in industries most similar to our
In this report we use the word “markets” and phrases
potential future arenas. In some instances, the
including the word “market,” such as “market Chapter 2
benchmarks are clear, as in software, e-commerce,
segment” and “market niche.” We may also show 14
Individuals using the internet (% of population)— and pharmaceuticals (for the drugs for obesity
various calculations of “market shares.” In doing so,
United States, World Bank, accessed July 2, 2024. and related conditions arena). In other instances,
we are referring to various groups of products sold 15
Revenues and market cap figures cited in this we considered a wide range of industries given
globally or in a particular region that we think are
section refer to our analyzed sample of about the nascency of the potential future arenas. For
useful to consider together. We do not intend to use
3,000 companies. example, modular construction took into account the
the word “market” in any formal or legal sense. 16
2
Netflix, Inc. 2003 annual report, Netflix, accessed construction engineering, building materials, and
Some industry classifications, such as software or
July 2, 2024; Netflix 2007 annual report, Netflix, software industries, as well as NOPLAT margins of
telecommunications, map one-to-one to potential
accessed July 2, 2024. existing modular construction players.
arenas. However, competitive battles happened 17 28
Gartner®, “Forecast: Semiconductor Capital McKinsey analysis incorporating data from Statista.
either within or across more established industry 29
Spending, Wafer Fab Equipment and Capacity, The analysis is limited to the sample of more than
classifications, so they can sometimes be averaged
Worldwide, 2Q24 Update,” by Bob Johnson, Gaurav 3,000 companies we considered, of which more than
out (for example, EVs within the automotive sector).
Gupta, Menglin Cao, Thomas Chiang, 8 July 2024. 600 were tagged as belonging to arenas.
Moreover, more than 300 companies operate in 30
Gartner®, “Forecast: Semiconductor Capital McKinsey analysis incorporating data from
multiple industries (such as Amazon in e-commerce
Spending, Wafer Fab Equipment and Capacity, PitchBook Inc.; analysis not reviewed by PitchBook
and cloud) but tend to be placed in a single industry
Worldwide, 4Q22 Update,” by Bob Johnson, Gaurav analysts. We mapped and tallied PitchBook’s
index. To address this, we created a segmentation
Gupta, Menglin Cao, Thomas Chiang. Gartner®, industry and vertical classifications that are closely
of 57 distinct industries and categorized the world’s
“Forecast: Semiconductor Capital Spending, attributable to arenas. Minor double counting of
top 3,000 companies from 2005 to 2020 within
Wafer Fab Equipment and Capacity, Worldwide, transactions may have resulted in instances where
these segments. This approach involved splitting the
4Q16 Update,” by Bob Johnson, Gaurav Gupta, a deal is attributable to multiple arenas, such as
revenues of 300 very large multi-industry companies
Menglin Cao, Thomas Chiang. Gartner®, “Forecast: e-commerce and AI software.
into their respective segments. 31
3
Semiconductor Capital Spending, Wafer Fab Connecting for inclusion: Broadband access for all,
These numbers cover only the e-commerce
Equipment and Capacity, Worldwide, 4Q11 Update,” World Bank, accessed July 8, 2024.
component of Amazon’s valuation. 32
4
by Bob Johnson, Gaurav Gupta, Menglin Cao, Thomas We recognize that in some cases boundaries are
LG closed its smartphone business in 2021.
5
Chiang. Gartner®, “Forecast: Semiconductor Capital blurred, especially due to the interconnected nature
We conducted the Exhibit 5 analysis for 2005–19
Spending, Wafer Fab Equipment and Capacity, of different arenas and technologies such as AI.
to isolate the impact of the COVID-19 pandemic on 33
Worldwide, 4Q06 Update,” by Bob Johnson, Gaurav Retail media networks allow retail companies
economic profit. We also conducted the analysis for
Gupta, Menglin Cao, Thomas Chiang. GARTNER is a to sell advertising space on their websites or
2005–20, and the results are similar, with all arenas
registered trademark and service mark of Gartner, applications. This is discussed further in the digital
improving their ranking over the time period.
6
Inc. and/or its affiliates in the U.S. and internationally advertisements arena.
Marc de Jong, Tido Röder, Peter Stumpner, and Ilya 34
and is used herein with permission. All rights “Projecting the global value of cloud: $3 trillion is up
Zaznov, “Working hard for the money: The crunch
reserved. Charts/graphics created by McKinsey & for grabs for companies that go beyond adoption,”
on global economic profit,” McKinsey Quarterly,
Company based on Gartner research. McKinsey, November 2022.
April 2023. 18 35
7
Kasper Karlsson, “Understanding the semiconductor “The state of AI in 2023: Generative AI’s breakout
Only data from the US National Center for Science
value chain,” Quartr, March 21, 2024. year,” McKinsey Global Survey, August 2023.
and Engineering Statistics was used given the 19 36
See “Arenas are where investment and innovation “The state of AI in early 2024: Gen AI adoption spikes
inconsistency in data quality of publicly reported
happen” in chapter 1. and starts to generate value,” McKinsey Global
R&D spend across our sample. 20
8
GDP growth annual (annual %), World Bank, Survey, May 2024.
National Center for Science and Engineering 37
accessed July 2, 2024. EMARKETER.
Statistics and Census Bureau, Business Enterprise 21 38
The distinction between these two categories is not Paresh Dave, “Google search’s new AI overviews will
Research and Development Survey, 2020; McKinsey
rigid as in both cases new products and services are soon have ads,” Wired, May 21, 2024.
Global Institute analysis. 39
9
displacing old ones, but it is a qualitatively different “Where does shared autonomous mobility go next?”
“Semiconductor design and manufacturing:
shift in the latter case. McKinsey, January 2023.
Achieving leading-edge capabilities,” McKinsey, 22 40
This highlights the fact that the potion is not Space: The $1.8 trillion opportunity for global
August 2020.
10
definitional for an arena but rather a falsifiable theory economic growth, McKinsey, April 2024.
Research and development in the pharmaceutical 41
about what creates arenas, supported by evidence in Ibid; Space IQ: Space Investment Quarterly, Q1 2023.
industry, Congressional Budget Office, April 2021. 42
11
almost every case, but to varying degrees. The role of cybersecurity and data security in the
As noted, we started from a list of the world’s 3,000 23
“Siemens driving digital transformation with $10 digital economy, UN Capital Development Fund,
largest companies. In this case, of the 29 companies
billion investment in US software companies since June 2022.
we identified in the consumer internet industry, 16 43
2007,” Siemens, March 27, 2017. Although video games are connected to the video
either did not exist or did not report any market cap 24
Sunil Gupta, “Lessons from Amazon’s early growth and audio entertainment industry, this is not a spin-
in 2005.
strategy,” Harvard Business Review, April 24, 2024. off arena because we excluded gaming from our

Arena-
analysis of video and audio entertainment. Other 64

Anutosh Banerjee, Robert Byrne, Ian De Bode, and 80
Euromonitor International Passport: Retail,
future arenas may also have associations with arenas Matt Higginson, “Web3 beyond the hype,” McKinsey, E-commerce, 2023 edition.
of today, but they represent distinct industries and September 2022. 81
Mercado Libre annual reports, 2019–24.
markets and do not fall within our more granular 65
These themes, as well as trends in demographics 82
“What is AI (artificial intelligence)?” McKinsey,
definitions of the arenas. and capitalization that will also shape the global April 2023.
44
A classification of games with the highest budgets economy more broadly, are discussed further in 83
“The state of AI in 2023: Generative AI’s breakout
and production values. Chris Bradley, Jeongmin Seong, Sven Smit, and Lola year,” McKinsey Global Survey, August 2023.
45
World obesity atlas 2023, World Obesity, March 2023. Woetzel, On the cusp of a new era? McKinsey Global 84
Daan van Rossum, “Generative AI top 150: The
46
In this section, swing factors are defined as highly Institute, October 2022. world’s most used AI tools,” FlexOS, February 2024.
uncertain developments that affect these arenas’ 66
Generative AI and the future of work in America, 85
“Why agents are the next frontier of generative AI,”
prospects for high growth or dynamism. McKinsey Global Institute, July 2023. McKinsey, July 2024; “What is artificial general
47
“Consumer electronics—worldwide,” Statista, 67
The hard stuff: Navigating the physical realities of intelligence (AGI)?,” McKinsey, March 2024.
accessed July 6, 2024. the energy transition, McKinsey Global Institute, 86
“Exploring opportunities in the generative AI value
48
Kevin Parmenter, “Navigating the industrial August 2024. chain,” McKinsey, April 2023.
electronics market: Trends, innovations and growth 87
If hardware is included, revenues in 2022 totaled
opportunities,” Power Systems Design, March 2024. $118 billion; “Worldwide AI and generative AI
49
“Global energy perspective 2023: Hydrogen Arenas of tomorrow compendium spending guide,” International Data Corporation,
outlook,” McKinsey, January 2024. 68
Crowdsourced delivery methods involve using 2024. McKinsey analysis incorporating data from
50
European Hydrogen Bank, European Commission, members of the general public as delivery persons. International Data Corporation and McKinsey Tech
accessed July 6, 2024. 69
“Social commerce: The future of how consumers Council analyses on investments.
51
Ajsa Habibic, “Germany grants €3.5 billion for interact with brands,” McKinsey, October 2022.
88
The economic potential of generative AI: The next
green hydrogen import program,” Offshore Energy, 70
Defining this arena as retail and food excludes some productivity frontier, McKinsey, June 2023.
February 14, 2024. categories of online purchases where the players
89
In the report The economic potential of generative
52
“Biden-Harris administration announces $7 billion are meaningfully different, such as automotive, AI, the total impact of these techniques is $11 trillion
for America’s first clean hydrogen hubs, driving travel e-commerce (online travel purchases made to $17.7 trillion. In estimating the specific impact
clean manufacturing and delivering new economic by consumers outside their country of residence, of gen AI, we deducted 15 percent of the total
opportunities worldwide,” US Department of Energy, purchases of travel services, and purchases of estimated value to account for value captured by
October 13, 2023. travel insurance), and mobility e-commerce (online non-AI capabilities.
53
“Safeguarding green steel in Europe: Facing the purchases of transportation- and mobility-related
90
Don Nguyen, “How AI can help diagnose
natural-gas challenge,” McKinsey, June 2021. products, including shared mobility services, mass rare diseases,” Harvard Medical School,
54
“On course for large-scale production from 2025,” transit, and other commuting expenditures). October 18, 2022.
H2 Green Steel, July 7, 2022; “H2 Green Steel 71
McKinsey analysis incorporating data from
91
The economic potential of generative AI, June 2023.
starts a new chapter,” North Sweden Business, Euromonitor International Passport: Retail,
92
Lamarre, Smaje, and Zemmel, Rewired. 2023.
March 23, 2024. E-commerce, 2023 edition.
93
Jonathan Vanian, “Mark Zuckerberg indicates Meta
55
“The resilience of steel: Navigating the crossroads,” 72
Analysis based on data from eMarketer, World Bank, is spending billions of dollars on Nvidia AI chips,”
McKinsey, April 2023. ArcGIS, and GSMA Intelligence. CNBC, January 18, 2024.
56
“Korian 2020 half-year results: Operational 73
Neira Hajro, Kate Smaje, Benjamin Vieira, and
94
“AI startups funded by Y Combinator 2024,”
resilience throughout the COVID-19 crisis with Rodney Zemmel, “Digital resilience: Consumer Y Combinator, March 2024.
increased investment in local care pathways,” Korian, survey finds ample scope for growth,” McKinsey,
95
Kim Martineau, “What is retrieval-augmented
July 30, 2020. October 2022. generation (RAG)?,” IBM Research, August 22, 2023.
57
Artem Vlasov, “Tokamaks, stellarators, laser-based 74
The food segment includes online ordering of food
96
Jack Flynn, “25 amazing cloud adoption statistics
and alternative concepts: Report offers global by consumers, regardless of whether the payment [2023]: cloud migration, computing, and more,”
perspective on nuclear fusion devices,” International occurs online or offline. Food prepared outside Zippia, June 22, 2023; “2024 state of the cloud
Atomic Energy Agency, February 2023. the home is included, as are groceries delivered by report,” Flexera, accessed July 3, 2024.
58
Magnets are required to contain super-hot plasma services such as Instacart.
97
Suresh Kumar, “Blazing a trail in cloud computing;
during the nuclear fission process. The first wall 75
Korea online penetration data excludes motor how Walmart built one of the largest hybrid clouds
in a fusion reactor is the surface of the inner wall vehicles, motor vehicle parts and accessories, travel in existence,” Walmart Global Tech via LinkedIn,
that is directly in contact with the plasma. The first- arrangement and transportation services, culture and June 23, 2022.
wall material must withstand a high neutron flux, leisure services, e-coupon services, food services,
98
“In search of cloud value: Can generative AI
high temperatures, and intense magnetic fields. and miscellaneous sales and travel; Country statistics transform cloud ROI?” McKinsey, November 2023.
Tokamaks are machines that confine plasma and are are from Euromonitor Country Retail Report, 2023,
99
“Projecting the global value of cloud: $3 trillion is up
the leading confinement concept for nuclear fusion and exclude auto sales and travel. for grabs for companies that go beyond adoption,”
power plants. 76
Total sales of consumer goods in December 2023, McKinsey, November 2022.
59
The benefits and costs of green power, US National Bureau of Retail Statistics of China,
100
McKinsey analysis incorporating data from
Environmental Protection Agency, accessed January 2024. International Data Corporation.
August 7, 2024. 77
“Quick commerce pushes the limits on grocery
101
“IDC’s forecast for worldwide IT spending in 2023
60
“AR & VR—worldwide,” Statista, accessed delivery,” McKinsey, August 2022. continues to slowly trend downward,” International
April 25, 2024. 78
Analysis based on data from Euromonitor Data Corporation, April 5, 2023.
61
“Global cryptocurrency market cap charts,” International Passport: Retail, E-commerce,
102
McKinsey CloudSights survey of 64 enterprises.
CoinGecko, accessed July 6, 2024. 2023 edition, and Kantar Retail IQ, May 2024.
103
Lauren Leffer, “The AI boom could use a shocking
62
Web3 developer report (Q3 2022), Alchemy Insights, 79
Euromonitor International Passport: Retail, amount of electricity,” Scientific American,
October 2022. E-commerce, 2023 edition; merchandise October 13, 2023.
63
Jason Brett, “Congress has introduced 50 digital value defined as gross merchandise value
104
“Data centers and data transmission networks,”
asset bills impacting regulation, blockchain, and after promotions, discounts, and other fees are International Energy Agency, accessed
CBDC policy,” Forbes, May 19, 2022. subtracted. April 11, 2024.

Arena-
105
McKinsey analysis incorporating IDC data. the evolving consumer ecosystem, Interactive 148
“Where does shared autonomous mobility go next?”
106
Data sovereignty regulations refer to the laws and Advertising Bureau, November 2021. McKinsey, January 2023.
130
regulations that govern the storage and processing Elijah Andes, “Carvana creates 1.3M+ unique 149
Jennifer Elias, “Alphabet to invest $5 billion in self-
of data in cloud services. AI-generated videos for customers,” Carvana, driving car unit Waymo,” CNBC, July 23, 2024.
107
“Average price and driving range of BEVs, May 9, 2023. 150
Levels of driving automation are as follows: level
131
2010–2019,” International Energy Agency, Andrea Day and Chris DiLella, “How Virgin Voyages zero, no driving automation; level one, driver
May 2020. is using A.I.—and a partnership with J. Lo–to boost assistance; level two, partial driving automation;
108
Kate Abnett, “EU lawmakers approve effective bookings,” CNBC, July 29, 2023. level three, conditional driving automation; level
132
2035 ban on new fossil fuel cars,” Reuters, “Video game market revenue worldwide in 2022, four, high driving automation; and level five, full
February 14, 2023. by segment (in billion U.S. dollars),” Statista, driving automation. For more information, see
109
“BMW Group continues on BEV growth path: August 29, 2023. “SAE levels of driving automation™ refined for
133
Milestone of one million fully electric vehicles “Q2 2024 update,” Spotify, July 23, 2024. clarity and international audience,” SAE
134
reached,” BMW Group, April 10, 2024. Omdia. International, May 2021.
135
110
“Ford hybrids, EVs, Transit set records; Q1 sales top McKinsey analysis incorporating data from 151
We did so because the competitive arena here
industry, up 7%,” Ford Motor Company, April 3, 2024. EMARKETER, Magna, and 10-K filings. is defined by the players offering a shared
136
111
Includes light-, medium-, and heavy-duty trucks and Josh Horwitz, “Fridges, microwaves fall prey to mobility service, rather than a product, so private
light commercial vehicles. global chip shortage,” Reuters, March 29, 2021. autonomous vehicles constitute a different industry.
137
112
“China market hits new milestone as EVs, hybrids QuettaFLOPs are 1030 floating-point operations, a We did not include private autonomous vehicles in
make up half of July sales,” Reuters, August 8, 2024. measure of computing power. our short list of arenas because there is uncertainty
138
113
Ziu Liu et al., “Global carbon emissions in 2023,” “Generative AI: The next S-curve for the over whether it will emerge as a distinct industry
Nature, April 2024. semiconductor industry,” McKinsey, March 2024. from automotive with new players or as a feature that
139
114
Annual MCFM Mobility Consumer Survey 2024, Bits&Chips; and Thomas Alsop, “Capital expenditure automotive players would add to their products. For
McKinsey, February 2024. in the global semiconductor industry from 2000 to more on private autonomous vehicles, see “Private
115
Tyler Graham, Dan Avery, “12 US states are planning 2023,” Statista, May 2024. autonomous vehicles: The other side of the robo-taxi
140
to ban the sale of gas-powered cars,” CNET, Ondrej Burkacky, Julia Dragon, and Nikolaus story,” McKinsey, December 2020.
June 12, 2024. Lehmann, “The semiconductor decade: A trillion- 152
“What’s next for autonomous vehicles?” McKinsey,
116
Bart Ziegler, “Commercial trucks are a key part of dollar industry,” McKinsey, April 2022; McKinsey December 2021.
EV adoption. What’s holding them back?” Wall Street analysis incorporating data from Statista and IDC. 153
Ibid.
141
Journal, July 23, 2023. Chris Donkin, “DeepMind CEO expects Google 154
Ibid.
117
Liz Najman, “How long do electric car batteries to top $100 billion in AI spending,” Mobile World 155
Ibid.
last?” Recurrent, May 20, 2024. Live, April 16, 2024; and Florian Zandt, “Data centers 156
“Autonomous vehicles move forward: Perspectives
118
The hard stuff: Navigating the physical realities of almost sole driver of Nvidia’s revenue boom,” from industry leaders,” McKinsey, January 2024.
the energy transition, McKinsey Global Institute, Statista, May 24, 2024. 157
“Where does shared autonomous mobility go next?”
142
August 2024. Tobias Mann, “Microsoft, OpenAI may be dreaming McKinsey, January 2023.
119
This includes only direct capital expenditures on of $100 billion 5GW AI ‘Stargate’ computer,” The 158
Costs of operating private vehicles include fixed
charging infrastructure and does not include the Register, April 1, 2024. costs (for example, acquisition, insurance, taxes,
143
cost of grid upgrades that may be required. Gartner®, Market Share: Semiconductor Wafer Fab tolls, and parking) and variable costs (such as
120
The International Energy Agency’s World Equipment, Worldwide, 2023, Bob Johnson, Gaurav depreciation, maintenance, cleaning, and fuel);
Energy Outlook 2023 puts global electricity Gupta, Menglin Cao, May 2024. exclude cost of the driver’s time.
production in 2030 at approximately 36,000 to GARTNER is a registered trademark and service 159
“Where does shared autonomous mobility
38,000 terawatt-hours. mark of Gartner, Inc. and/or its affiliates in the go next?” McKinsey, January 2023. Our estimates
121
“Alternative fuels data center,” US Department of U.S. and internationally and is used herein with are in line with those of other researchers.
Energy, accessed August 1, 2024. permission. All rights reserved. See Olivia Fiol and Sophia Weng, “Shared
144
122
“Passenger cars market size & share analysis— Omdia 2001 to 2023 Semiconductor Market autonomous vehicles could improve transit
growth trends & forecasts up to 2030,” Mordor Share (2024). access for people with disabilities if regulated
145
Intelligence, accessed August 1, 2024. Suhas A R, Joel Martin, and Niti Jhunjhunwala, appropriately,” Urban Institute, October 2022; and
123
McKinsey analysis based on data from EV Volumes (2024). “Semiconductors—the next frontier of geopolitics,” Patrick M. Bösch et al., “Cost-based analysis of
124
Ibid. HFS Research, March 2024; and Judy Lin, “TSMC autonomous mobility services,” Transport Policy,
125
See the discussion of the consumer internet arena of dominates global foundry market with 61% market volume 64, May 2018.
today in chapter 1. share in 4Q23, says Counterpoint,” DigiTimes Asia, 160
Due to the complexities of modeling passenger
126
One difference between the digital advertising arena May 3, 2024. demand for annual road miles traveled, our analysis
146
of the future and the current consumer internet Joann Muller, “Robo-taxis hit the accelerator in takes into account only growth of passenger miles
arena is that the latter also includes revenues from growing list of cities nationwide,” Axios, August 29, due to expected GDP and population growth.
paid subscriptions to services such as AOL that were 2023; Evelyn Cheng, “Baidu says it can now operate 161
“The road to affordable autonomous mobility,”
popular in the early years of the 21st century. robo-taxis in Beijing with no human staff inside,” McKinsey, January 2022.
127
McKinsey analysis incorporating data from CNBC, March 17, 2023. 162
Carmakers’ overall vehicle sales through 2040
147
EMARKETER; Statista; Oxford Economics; PwC Kersten Heineke, Benedikt Kloss, Timo Möller, are projected to stay consistent with their historical
Global Entertainment & Media Outlook 2023–2027, and Charlotte Wiemuth, “Shared mobility: Where it trajectory. That is, demand for SAVs would replace
PwC; Moffett Nathanson; and Magna. stands, where it’s headed,” McKinsey, August 2021. demand for other cars rather than driving up
128
McKinsey analysis incorporating data from World Robo-taxis are autonomous vehicles that function total demand. However, automakers may be
Data Lab; the top ten countries are India, China, similarly to taxis. They carry four to five passengers able to benefit from the trend toward autonomous
Bangladesh, Indonesia, Pakistan, Vietnam, the on an on-demand basis on personalized routes. vehicles if they become preferred partners of
Philippines, Mexico, Egypt, and Nigeria, ranked by Robo-shuttles generally transport 20 or more SAV service providers or if they offer superior
the increase of middle-class population. passengers on predetermined routes with autonomous driving capabilities in people’s
129
Brand disruption 2022: The IAB annual report on scheduled stops. personal vehicles.

Arena-
163 190 212

Kristen Korosec, “Volkswagen to start testing Based on 10-K company reports. Front-of-the-meter installations include energy grid
191
self-driving ID Buzz vans in Austin,” TechCrunch, Darrell Etherington, “Spire’s ship tracking satellite storage facilities; behind-the-meter installations
July 6, 2023. data makes it easier to monitor vessels from space,” include commercial and residential storage systems,
164
“Waymo and Uber partner to bring Waymo’s TechCrunch, August 29, 2017. often used with solar panels, as backup systems, or
192
autonomous driving technology to the Uber Space: The $1.8 trillion opportunity, 2024. with EV charging infrastructure.
193 213
platform,” Uber Technologies, May 23, 2023; “Uber Acquired Ball Aerospace. Batteries and secure energy transitions, IEA,
194
and Cruise to deploy autonomous vehicles on the The role of cybersecurity and data security in April 2024.
214
Uber platform,” Uber Technologies, August 22, 2024. the digital economy, United Nations Capital For further discussion of managing variability in
165
Geofencing refers to the use of GPS or radio Development Fund, June 2022. renewable energy generation, read The hard stuff:
195
frequency identification to create virtual geographic “Technology patents by top themes, 2017–2021— Navigating the physical realities of the energy
boundaries and trigger a software response when Thematic research,” GlobalData, May 2022. transition, McKinsey Global Institute, August 2024.
196 215
that boundary is crossed. McKinsey analysis incorporating data from IDC and Renewables 2023—analysis, International Energy
166
“Launches,” SpaceX, June 23, 2024. the McKinsey Capability Maturity Model. Agency, January 2024.
167 197 216
“Virgin Galactic spaceplane takes tourists on flight,” We estimate that all IT spending will grow at a CAGR Frequency regulation is the process of maintaining a
Reuters, June 8, 2024. of about 6.5 percent between 2022 and 2040. That stable frequency in the electrical grid by smoothing
168
Ariel Cohen, “China’s space mining industry is is a global estimate; the rate would vary significantly fluctuations due to variations in electricity demand
prepping for launch—but what about the US?” by region. and supply.
198 217
Forbes, October 26, 2021. Bharath Aiyer, Jeffrey Caso, Peter Russell, and Thermal storage refers to the process of
169
William Harwood, “3 companies win NASA contracts Marc Sorel, “New survey reveals $2 trillion market capturing and storing thermal energy for later use.
to develop new Artemis Moon rover designs,” CBS opportunity for cybersecurity technology and Compressed-air storage uses compressed air to
News, April 3, 2024. service providers,” McKinsey, October 2022. store energy. Hydrogen power-to-gas involves
170 199
Siladitya Ray, “India’s ambitious Moon mission cost Bharath Aiyer, Jeffrey Caso, and Marc Sorel, “The storing energy for extended periods by creating
less than Hollywood space films—here’s what you unsolved opportunities for cybersecurity providers,” hydrogen through electrolysis.
218
need to know,” Forbes, August 9, 2023. McKinsey, January 2022. Wolfgang Fengler et al., How the world consumer
171 200
Space: The $1.8 trillion opportunity for global “Cybersecurity trends: Looking over the horizon,” class will grow from 4 billion to 5 billion people by
economic growth, McKinsey and World Economic McKinsey, March 2022. 2031, Brookings Institution, July 2023.
201 219
Forum, April 2024. Daniel Howley, “Microsoft commits to spend “Global supply chains of EV batteries,” IEA, accessed
172
Space launch to low Earth orbit: How much does it $20 billion on cybersecurity over five years,” Yahoo! July 1, 2024.
220
cost? Center for Strategic and International Studies, Finance, August 25, 2021. Ibid.
202 221
accessed July 8, 2024. Kent Walker, “Why we’re committing $10 billion to Sam Adham and Arena Yang, “Oversupply in China’s
173
Space: The $1.8 trillion opportunity, 2024; and Space advance cybersecurity,” Google, August 25, 2021. battery cell industry will lead to consolidation,” CRU,
203
Investment Quarterly, Q1 2023. Dean Walter, Sam Butler-Sloss, and Kingsmill Bon, August 25, 2023.
174 222
Space: The $1.8 trillion opportunity, 2024. “The rise of batteries in six charts and not too many Biden-Harris administration announces $3.5 billion
175
McKinsey analysis incorporating data from numbers,” RMI, January 2024. to strengthen domestic battery manufacturing, US
204
McKinsey’s Aerospace & Defense Practice. “Lead acid battery market,” Straits Research, Department of Energy, November 15, 2023.
176 223
Market size estimates include backbone applications accessed July 1, 2024. Reinventing construction through a productivity
205
and exclude reach applications. Includes heavy-, medium-, and light-duty trucks as revolution, McKinsey, February 2017.
177 224
“Next generation satellite systems continue to well as light commercial vehicles. ‘Superstars’: The dynamics of firms, sectors, and
206
disrupt satellite capacity pricing landscape,” Utility-scale storage systems play a crucial role in cities leading the global economy, McKinsey Global
Euroconsult, February 2024. ensuring grid stability and backup power. This is Institute, October 2018.
178 225
Space: The $1.8 trillion opportunity, 2024. especially important because renewable energy “UN expert urges action to end global housing
179
Ibid. sources such as wind and solar are intermittent and shortage,” Office of the United Nations High
180
Ibid. can cause fluctuations in power supply. Commissioner for Human Rights, October 20, 2023.
181 207 226
Using satellite imagery for emergency disaster Traditional household batteries were excluded from Ezra Greenburg, Erik Schaefer, and Brooke Weddle,
response, British Geological Survey, June 2021. the analysis given the sector’s low growth potential “Tradespeople wanted: The need for critical trade
182
“Space agencies around the world,” Space Crew, and stagnant innovation. skills in the US,” McKinsey, April 2024.
208 227
March 2024. Gigafactories are specialized factories producing Building value by decarbonizing the built
183
Hari Kumar et al., “India is on the moon: Lander’s electric batteries at large scale. environment, McKinsey, June 2023.
209 228
success moves nation to next space chapter,” New A cell is the basic unit that generates electrical “Modular construction: From projects to products,”
York Times, August 23, 2023. energy; a battery pack is a collection of cells. McKinsey, June 2019.
184 229
“Government space spending of the leading Battery cell cost is roughly 10 to 15 percent lower Jose Luis Blanco, Dave Dauphinais, Garo Hovnanian,
countries in the world,” Statista, March 2024. than the battery cell price, which also includes and Rob Palter, “Making modular construction fit,”
185
“Euroconsult predicts 10-year growth cycle for the manufacturer’s profit margin. Our revenue McKinsey, May 2023.
230
government space program,” Euroconsult, July 2019. forecasts are based on the battery cell price that The penetration rate is measured here as the
186
Steven J. Butow et al., State of the space industrial manufacturers charge their customers. Battery cell total dollar value of modular construction as a
base: Winning the new space race for sustainability, costs are around 10 percent more than the battery percentage of the total cost of construction
prosperity and the planet, United States Space raw materials (seen in Exhibit 2). projects. The percentage of projects that use some
210
Force, Defense Innovation Unit, Department of “Global EV Outlook 2023 executive summary,” type of modular construction is higher than the
the Air Force, and Air Force Research Laboratory, International Energy Agency (IEA), April 2023; penetration rate.
231
August 2022. Joel Jaeger, These countries are adopting electric McKinsey analysis incorporating data from
187
Space: The $1.8 trillion opportunity, 2024. vehicles the fastest, World Resources Institute, MarketsandMarkets and Oxford Intelligence.
188 232
The role of space in driving sustainability, security, September 2023. Jose Luis Blanco, David Rockhill, Aditya Sanghvi,
211
and development on Earth, McKinsey, May 2022. Nelson Nsitem, “Global energy storage market and Alberto Torres, “From start-up to scale-up:
189
Aashna Shah, “Venture capital fuels space startup records biggest jump yet,” BloombergNEF, Accelerating growth in construction technology,”
dreams,” Bloomberg, May 15, 2024. April 25, 2024. McKinsey, May 2023.

Arena-
233
Video and audio entertainment, a closely related 255
Taylor Lyles, “Over half of US kids are playing Roblox, be defined as people’s negative reaction to certain
industry, is described in chapter 1 as one of and it’s about to host Fortnite-esque virtual parties lifelike robots.” Rina Diane Caballar, “What is the
today’s arenas. too,” The Verge, July 21, 2020. Uncanny Valley?” IEEE Spectrum, May 24, 2024.
272
234
Petroc Taylor, “Number of fixed broadband 256
Mark Delaney, “This Fortnite creative mode is as Venture capital figures from McKinsey analysis
subscriptions worldwide from 2005 to 2023, by popular as Epic’s own modes–and bigger than the incorporating data from PitchBook Inc.; analysis
region,” Statista, March 28, 2024. game that inspired it,” Gamespot, July 5, 2023. not reviewed by PitchBook analysts. “Covariant
235
“Roku hits major milestones, surpassing 80 million 257
Andrea Knezovic, “IP-based mobile games: adds $75 million in Series C funds to meet customer
active accounts and more than 100 billion streaming Why they’re dominating the industry,” Udonis, demand for scaled AI robotics deployments,”
hours in 2023,” Roku, February 15, 2024. February 6, 2024. Covariant, April 4, 2023.
273
236
McKinsey analysis incorporating data from Magna 258
Eliza Crichton-Stuart, “Using brand & IP “Hyundai Motor Group completes acquisition of
(December 2023): Omdia, Oxford Economics, collaborations to boost engagement & revenue,” Boston Dynamics from SoftBank,” Hyundai Motor
and Statista. gam3s.gg, November 2023. Group, June 21, 2021; “Agility Robotics: Company
237
“TV & online video intelligence database,” Omdia, 259
“Esports–worldwide,” Statista, accessed diligence,” Agility Robotics, accessed July 11, 2024;
June 11, 2024. July 9, 2024. “Figure AI raises $675 million at $2.6 billion valuation
238
Brad Adgate, “NBA new media rights agreement 260
Fabio Duarte, “Top 23 esports statistics in 2024,” and signs collaboration agreement with OpenAI,”
reflect the growth of streaming,” Forbes, July 29, Exploding Topics, April 29, 2024. Figure AI, February 29, 2024.
274
2024; and Mark Sweeney, “Amazon and BBC break 261
Tom Wijman, “NewZoo’s video games market size “Amazon begins using humanoid robot in
BT stranglehold on Champions League football,” estimates and forecasts for 2022,” NewZoo, warehouse,” Bloomberg, March 24, 2024; Noelle
Guardian, July 1, 2022. May 17, 2023. Mateer, “Hotel tech-in: The robot butlers taking over
239
Danielle Commisso, “One subscription too many— 262
McKinsey analysis incorporating data from hotel deliveries,” Hotel Dive, July 31, 2023; Gerrit
video streaming reaches an inflection point as PitchBook Inc. Analysis was not reviewed by DeVynck and Rachel Lerman, “Google is training its
consumers report feeling ‘subscription fatigue,’” PitchBook analysts. robots to be more like humans,” Washington Post,
CivicScience, August 16, 2023. 263
Jordan Novet, “Microsoft closes $69 billion August 16, 2022; and “Nvidia announces project
240
“Ad-supported streaming monthly viewers,” acquisition of Activision Blizzard after lengthy GR00T foundation model for humanoid robots
EMARKETER, September 3, 2024. regulatory review,” CNBC, October 13, 2023. and major Isaac robotics platform update,” Nvidia,
241
Laura Ceci, “Mobile video in the United States— 264
Developers must be formally authorized, sign a March 18, 2024.
275
statistics & facts,” Statista, December 18, 2023. nondisclosure agreement to receive the necessary Technologies that create the experience of touch,
242
Bill Raymond, “MrBeast vs. Netflix: The rise of UGC,” hardware, pay licensing fees, and meet the forces, and motions.
276
CannellaMedia, updated March 27, 2024. console company’s technical requirements. In Jobs lost, jobs gained: What the future of work will
243
Jennifer Maas, “Netflix lost 970,000 subscribers in PC game development, no legal agreements are mean for jobs, skills, and wages, McKinsey Global
Q2, beating its estimate by more than 1 million subs,” required, there are many options for open-source Institute, November 2017; and The economic
Variety, July 19, 2022. development, and developers have many choices potential of generative AI: The next productivity
244
Jon Brodkin, “Netflix stock plummets 37% as for distribution, including third-party platforms and frontier, McKinsey Global Institute, June 14, 2023.
277
CEO says company plans ad-supported tier,” Ars direct sales to consumers. Based on $15.8 billion in industrial robotics from
Technica, April 20, 2022. 265
Aisha Malik, “Take-Two completes $12.7B acquisition World Robotics 2023 Report: Asia ahead of Europe
245
“Machine learning,” Netflix, accessed July 3, 2024. of mobile games giant Zynga,” TechCrunch, May 23, and the Americas, International Federation of
246
“Disney and Epic Games to create expansive 2022; and “Activision Blizzard to acquire King Digital Robotics, September 2023; and $5.3 billion in
and open games and entertainment universe Entertainment for $5.9 billion,” Activision Blizzard, automated guided vehicles and autonomous mobile
connected to Fortnite,” The Walt Disney Company, November 2, 2015. robots from Interact Analysis: The Mobile Robot
February 7, 2024. 266
Rebekah Valentine, “The average cost to acquire Market 2023, October 2023.
278
247
Andrea Knezovic, “Gen Alpha and Gen Z a mobile game user plummeted this year,” DWAs include tasks such as cleaning work areas,
gamers: How they engage with games,” Udonis, GamesIndustry.biz, September 18, 2020. setting up merchandise displays, and moving
December 19, 2023. 267
We exclude robots for transportation (for example, materials, equipment, or supplies.
279
248
“Registered users of Fortnite worldwide from August self-driving, delivery drones, and self-piloting Although the analysis was based on work activities
2017 to March 2023,” Statista, March 22, 2023; and planes) from our discussion of market sizing and occupations in data from the US Department
“Revenue generated by Fortnite worldwide from and market dynamics. The deep dive on shared of Labor, it provided a template to estimate the
2018 to 2022,” Statista, August 23, 2023. autonomous vehicles (SAVs) can be found in the automation potential and create timing scenarios
249
Anticipated acquisition by Microsoft of Activision SAVs arena write-up in this compendium. of automation for 45 other economies representing
Blizzard, Inc., final report, UK Competition & Markets 268
Robot history, International Federation of Robotics, 80 percent of the global workforce.
280
Authority, April 2023. accessed April 2, 2024. A future that works: Automation, employment,
250
McKinsey analysis incorporating data from S&P 269
Yanfan Liu, “Impact of industrial robots on and productivity, McKinsey Global Institute,
Global Market Intelligence. environmental pollution: Evidence from China,” January 2017.
281
251
“US consumer spending on video games inched Scientific Reports, November 2023. Both scenarios assume that robots already match
up in 2023,” Marketing Charts, February 1, 2024, 270
Jyoti Mann, “Sam Altman’s OpenAI wants to get top-quartile human performance for gross motor
and Circana. humanoid robots talking. Here’s why,” Business skills. The higher range of scenarios also assumes
252
McKinsey analysis incorporating data from PwC Insider, February 29, 2024. that robots reach top-quartile human performance
Global Entertainment & Media Outlook 2023–2027, 271
“The uncanny valley is a concept first introduced in for fine motor skills by 2025 and for mobility by
PwC, accessed July 11, 2024; NewZoo; Omdia; S&P the 1970s by Masahiro Mori, then a professor at the 2037. The lower range of scenarios assumes robots
Global Market Intelligence; and Statista. Tokyo Institute of Technology. Mori coined the term wouldn’t attain top-quartile human performance for
253
Damjana Cikaric, “15 stats on mobile gaming . . . to describe his observation that as robots appear fine motor skills until 2040; reaching that level of
demographics for 2024,” PlayToday.Co, more humanlike, they become more appealing—but mobility wouldn’t occur until after 2050.
282
February 23, 2024. only up to a certain point. Upon reaching the uncanny Help wanted: Charting the challenge of tight labor
254
“How user-generated content is changing the game valley, [humans’] affinity descends into a feeling of markets in advanced economies, McKinsey Global
industry for good,” GamesBeat, October 27, 2022. strangeness, a sense of unease, and a tendency to Institute, June 2024.
be scared or freaked out. The uncanny valley can

Arena-
283
Frederico Berruti, Damian Lewandowski, and 301
Amil Merchant and Ekin Dogus Cubuk, “Millions The future of unmanned aerial systems,” McKinsey,
Jonathan Tilley, “The robot renaissance: How of new materials discovered with deep learning,” December 2017.
321
humanlike machines are reshaping business,” DeepMind, Google, November 29, 2023. Aisha Malik, “DoorDash begins piloting drone
McKinsey, March 22, 2024. 302
Funding includes venture capital, R&D, private deliveries in the US,” TechCrunch, March 21, 2024.
322
284
“How to boost productivity with collaborative investment in public equity, and funding from Andrea Cornell, Sarina Mahan, and Robin Riedel,
robots,” Universal Robots, accessed July 11, 2024. special purpose acquisition companies, as of “Commercial drone deliveries are demonstrating
285
“The robot renaissance,” March 22, 2024. July 17, 2023; Saskia Boeck, Sarina Mahan, continued momentum in 2023,” Future Air Mobility
286
“Hyundai Motor Group completes acquisition of Ann-Sophie von Gaisberg, Tore Johnston, Blog, McKinsey, October 6, 2023.
323
Boston Dynamics from SoftBank,” Hyundai Motor Stephan Lidel, and Robin Riedel, “FAM funding: Cardiovascular disease includes heart disease
Group, June 21, 2021. Capital flows regain momentum despite and stroke. COPD is commonly referred to as
287
“Hyundai Motor Group launches Boston Dynamics challenges,” Future Air Mobility Blog, McKinsey, lung disease.
324
AI Institute to spearhead advancements in artificial June 16, 2023. John P. Ansah and Chi-Tsun Chiu, “Projecting the
intelligence & robotics,” Hyundai Motor Group, 303
“Joby completes third stage of FAA certification chronic disease burden among the adult population
August 12, 2022. process,” Joby Aviation, February 21, 2024; and in the United States using a multi-state population
288
Irina Gostimskaya, “CRISPR–Cas9: A history of its “Archer receives FAA certification to begin operating model,” Frontiers in Public Health, volume 10, 2022.
325
discovery and ethical considerations of its use in commercial airline,” Archer Aviation, June 5, 2024. Chirag Adatia, Ralf Dreischmeier, Samarta Shah,
genome editing,” Biochemistry (Moscow), volume 87, 304
Analysis incorporating data from McKinsey and Kirtika Sharma, “The health benefits and
number 8, August 2022. Advanced Air Mobility research. business potential of digital therapeutics,” McKinsey,
289
“2024 annual synthetic biology investment report,” 305
We define eVTOLs’ revenues as the sale of January 2023.
326
SynBioBeta, accessed August 1, 2024. Totals cited passenger trips and cargo transportation, and we The global economic burden of noncommunicable
above exclude healthcare and biopharma segments. include autonomous and human-piloted eVTOLs. We diseases, Harvard School of Public Health and World
290
Gené Teare, “Global startup funding in 2023 clocks define delivery drones’ revenues as the revenues of Economic Forum, September 2011.
327
in at the lowest level in 5 years,” Crunchbase, the services that operate them. We exclude military Adding years to life and life to years, McKinsey
January 4, 2024. and surveillance uses from our analysis. Health Institute, March 2022.
328
291
McKinsey analysis incorporating data from the 306
Battery autonomy refers to battery duration at a About obesity, US Centers for Disease Control and
McKinsey Global Institute report The bio revolution: specific load level. Prevention, January 23, 2024.
329
Innovations transforming economies, societies, and 307
Matt Pelkey, “Cost of an Uber in cities around the World obesity atlas 2023, World Obesity Federation,
our lives, May 2020; the McKinsey industrial biotech world,” NetCredit blog, July 6, 2022. March 2023.
330
market model; and the Vivid Economics alternative 308
Trip distance measured per flight-path kilometer. For more information on the DALY metric, see
proteins model. 309
Andrea Cornell, Benedikt Kloss, D. J. Presser, and “Disability-adjusted life years,” World Health
292
Richard Waite and Alex Rudee, 6 ways the US can Robin Riedel, “Drones take to the sky, potentially Organization, accessed April 19, 2024.
331
curb climate change and grow more food, World disrupting last-mile delivery,” Future Air Mobility According to the McKinsey Health Institute: all-
Resources Institute, August 20, 2020; and Nana Blog, McKinsey, January 23, 2023. causes DALYs attributed to high body mass index
Afranaa Kwapong, Does production of biofuel 310
Benedikt Kloss and Robin Riedel, “Up in the air: among adults (15+ years old) in 2019 from Institute
mean less food production? Biomass Connect, How do consumers view advanced air mobility?” for Health Metrics and Evaluation. Used with
April 25, 2023. McKinsey, June 1, 2021. permission.
332
293
Erik Stokstad, “New genetically modified corn 311
Ramya Mahalingam, Adam Mitchell, and Robin For more information on opportunities to
produces up to 10% more than similar types,” Riedel, “How customer experience will make or improve drug development, see Guarav Agrawal,
Science, November 2019. break advanced air mobility,” Future Air Mobility Harriet Keane, Maha Prabakaran, and Michael
294
Erik Stokstad, “European Parliament votes to Blog, McKinsey, March 7, 2023. Steinmann, “The pursuit of excellence in new-drug
ease regulation of gene-edited crops,” Science, 312
“FAA issues implementation plan outlining steps development,” McKinsey, November 2019.
333
February 7, 2024. to usher in advanced air mobility,” Federal Aviation Sanne Wass, Naomi Kresge, and Bloomberg, “Novo
295
Mayank Bhardwaj, “Activists accuse India of lapses Administration, July 2023. Nordisk’s market value of $570 billion is now bigger
in genetically modified mustard approval,” Reuters, 313
“Vertical take-off and landing (VTOL),” European than the entire Danish economy—creating a ‘Nokia
January 7, 2023; and Matt Reynolds, “As Kenya’s Union Aviation Safety Agency, accessed risk’ for Denmark,” Fortune, May 1, 2024.
334
crops fail, a fight over GMOs rages,” Wired, August 1, 2024. Angel Adegbesan, “Eli Lilly surpasses UnitedHealth
March 3, 2023. 314
Evelyn Cheng, “China gives Ehang the first industry as world’s biggest healthcare firm,” Bloomberg,
296
J. M. Ribaut, M. C. de Vicente, and X. Delannay, approval for fully autonomous, passenger-carrying July 5, 2023.
335
“Molecular breeding in developing countries: air taxis,” CNBC, November 8, 2023; Charles Alcock, Jon Haworth, “Pfizer targets obesity with new
Challenges and perspectives,” Current Opinion in “AutoFlight delivers first prosperity eVOTL and once-daily danuglipron drug,” ABC News, July 11,
Plant Biology, volume 13, number 2, April 2010. certifies freighter,” AIN Media Group, April 4, 2024. 2024; Jonathan Gardner, “Amgen shares soar as
297
Produced by cultivating animal cells directly. 315
“Drone delivery services,” Australian Civil Aviation executives outline obesity drug push,” BioPharma
298
Greg L. Garrison, Jon T. Biermacher, and B. Wade Safety Authority, accessed July 17, 2024. Dive, May 3, 2024; Ludwig Burger, “Roche joins
Brosen, “How much will large-scale production of 316
Zeyi Yang, “Food delivery by drone is just part of daily race for obesity drugs with $2.7 billion Carmot deal,”
cell-cultured meat cost?” Journal of Agriculture and life in Shenzhen,” MIT Technology Review, May 2023. Reuters, December 4, 2023.
336
Food Research, volume 10, December 2022. 317
Axel Esqué, Tore Johnston, and Robin Riedel, “Major Qiu-Xuan Li et al., “GLP-1 and underlying beneficial
299
Ping Fan et al., “A review on the occurrence and developments in 2022 and significant milestones actions in Alzheimer’s disease, hypertension and
influence of biodegradable microplastics in soil ahead,” Future Air Mobility Blog, McKinsey, NASH,” Frontiers in Endocrinology, volume 12, 2021.
337
ecosystems: Are biodegradable plastics substitute January 19, 2023. McKinsey analysis incorporating data from Clarivate,
or threat?” Environment International, volume 163, 318
Tore Johnston, Robin Riedel, and Shivika Sahdev, World obesity atlas, and UN World Population
May 2022. “To take off, flying vehicles first need places to land,” Prospects 2022.
338
300
Tom Brennan, Michael Chui, Wen Chyan, and McKinsey, August 2020. The increase in appetite for obesity drugs, JPMorgan
Axel Spamann, “The third wave of biomaterials: 319
Ibid. Chase, November 29, 2023; “Why the anti-obesity
When innovation meets demand,” McKinsey, 320
Pamela Cohn, Alastair Green, Meredith Langstaff, drug market could grow to $100 billion by 2030,”
November 2021. and Melanie Roller, “Commercial drones are here: Goldman Sachs, October 30, 2023.

Arena-
339 352 362

For more information on how GLP-1 agonists work, Global Energy Perspective 2022, McKinsey, Costs declined from $5,500 for the first unit to
see Logan Collins and Ryan A. Costello, Glucagon- April 2022. $4,300 for the second unit to $2,500 for the fourth
353
like peptide-1 receptor agonists, National Institutes Jennifer Hiller, “Microsoft targets nuclear to power unit. On average, cost falls by 32 percent with each
of Health, February 2024. AI operations,” Wall Street Journal, doubling of production; Unlocking reductions in the
340
World Obesity Atlas, 2023. December 12. 2023. construction costs of nuclear: A practical guide for
341 354
UN World Population Prospects 2022, medium Future Nuclear Enabling Fund, UK Department for stakeholders, OECD, 2020.
363
fertility variant 2022–2100, 2022. Business, Energy & Industrial Strategy, May 2022. Glossary, US Energy Information Administration,
342 355
Benedic N. Ippolito and Joseph F. Levy, Estimating Nuclear power in South Korea, World Nuclear accessed August 13, 2024.
364
the cost of new treatments for diabetes and obesity, Association, May 2024. Global Energy Perspective 2022, McKinsey,
356
American Enterprise Institute, September 2023. “Finland’s nuclear and renewable power strengths April 2022.
343 365
Juliette Cubanski et al., A new use for Wegovy opens provide a solid foundation for reaching its ambitious Current percentage of energy mix calculated from
the door to Medicare coverage for millions of people climate targets, IEA review says,” IEA, May 2023; and Global Energy Perspective 2022 estimates; 2040
with obesity, KFF, April 2024. “Distribution of electricity generation in Finland by estimates are derived from the sum of additional
344
Hershel Jick et al., “Comparison of prescription drug 2023, by source,” Statista, June 28, 2024. nuclear builds and current capacity, along with
357
costs in the United States and the United Kingdom, “At COP28, countries launch declaration to triple electricity generation forecast under different
Part 1: statins,” Pharmacotherapy, January 2012. nuclear energy capacity by 2050, recognizing the decarbonization pathways.
345 366
Leigh Purvis and Stephen Schondelmeyer, Trends key role of nuclear energy in reaching net zero,” US Under construction, IAEA Power Reactor Information
in retail prices of prescription drugs widely used by Department of Energy, December 1, 2023. System, accessed August 13, 2024; and Stephen
358
older Americans, 2006 to 2020, AARP Public Policy Firm power refers to power-producing capacity Ezell, How innovative is China in nuclear power?
Institute, January 16, 2024. that is available at all times with a guaranteed Information Technology and Innovation Foundation,
346
Matej Mikulik, “Global spending on medicines in commitment to deliver, as opposed to common June 2024.
367
2010, 2020, and a forecast for 2027,” Statista, renewable sources such as solar and wind that have Current percentage of energy mix calculated from
May 2024. varying generation capacities depending on weather Global Energy Perspective 2022 estimates; 2040
347
Diagnosis rates are calculated by region; McKinsey conditions and time of day. estimates are derived from the sum of additional
359
projection to 2040 incorporating data on obesity Current percentage of energy mix calculated from nuclear builds and current capacity, along with
diagnosis rates from 2022 to 2035 from Clarivate, McKinsey’s Global Energy Perspective 2022 electricity generation forecast under different
© 2022 DR/Decision Resources, LLC. estimates; 2040 estimates are derived from the sum decarbonization pathways.
348 368
The assumption that 20 percent of North American of additional nuclear builds and current capacity, Nuclear reactors in Finland, World Nuclear
patients would take the drugs is roughly half of along with electricity generation forecasts under Association, accessed August 13, 2024.
369
today’s treatment rate for type 2 diabetes patients different decarbonization pathways. “TerraPower announces $830 million secured in
360
who use branded medications for treatment, which World electricity generation mix by fuel, 1971–2019, 2022,” TerraPower, November 3, 2022.
370
have a similar per-patient annual cost today. IEA, August 2021. Success stories: TerraPower nuclear plant,
349 361
The global use of medicines 2024: Outlook to 2028, The costs of the nuclear power sector, Cour des Interagency Working Group on Coal and Power
IQVIA Institute, January 2024. Comptes, January 2012; Jessica R. Lovering, Arthur Plant Communities and Economic Revitalization,
350
Medicines in development, PhRMA, accessed Yip, and Ted Nordhaus, “Historical construction July 9, 2024.
April 19, 2024. costs of global nuclear power reactors,” Energy
351
Use cases for nuclear fission could expand beyond Policy, volume 91, April 2016; and Abdalla Abou-
pure power to pink hydrogen, district heating, and Jaude et al., Meta-analysis of advanced nuclear
process heating, among others. reactor cost estimations, Idaho National Laboratory,
July 2024.

Arena-
Technical appendix
Chapter 1 Data set
The companies we analyzed were selected from more than 20,000 companies tracked in the
McKinsey Value Intelligence database, applying the following filters:
— Data quality: Data at least included detailed reporting of income statement and balance sheet.
— Parent companies only: No double counting of subsidiaries.
— Minimum market cap: $3.5 billion in 2005 or $5 billion in 2020.
We augmented this initial database to account for IPOs and other M&A activities from 2005 to 2020,
based on Capital IQ data. Seventy private companies with at least $5 billion in valuation as of 2020
were also included, along with their 2005 data.
For companies with missing market cap values for 2005, we looked at their 2020 market-cap-to-
revenue multiples and applied this multiple to the companies’ 2005 revenues to approximate 2005
market caps. All market cap numbers were measured at the end of the relevant calendar year.
This data set yielded 2,970 unique companies. Those that operated in several industries were split
into multiple discrete entries. For example, Microsoft was split into four entries: cloud services,
consumer electronics, consumer internet, and software. Amazon was split into two: cloud services
and e-commerce. Publicly available data such as annual reports and press releases was used to
calculate how to split these companies. This yielded a subset of 150 companies. We used the revenue
split of individual business segments as the baseline to calculate the market cap split adjusted by the
industry’s multiple.
To illustrate our process with a hypothetical example, Theta Company has operations in both
consumer electronics and cloud services. In this scenario, the consumer electronics segment
generates $8,000 of revenue, and the hypothetical assumes that the industry has an average
market-cap-to-revenue multiple of two. In cloud services, Theta has $2,000 in revenues, and the
hypothetical assumes that the industry has an average market-cap-to-revenue multiple of eight.
By multiplying the revenues and multiples, we find that Theta’s market cap split between consumer
electronics and cloud services is 50 percent each. We can then multiply Theta’s market cap for
the year by 50 percent to produce its market cap for each industry. For this model, each company-
industry combination was treated as a unique data point. The final count of data points was 3,135.
The data set encompassed 57 different industries. Twelve of these were the current-era arenas we
identified. These 57 industries were defined by starting with the McKinsey Value Intelligence data
set, which as of 2023 contains 29 level two subindustries defined by McKinsey Industry Classification
(MIC). These MIC subindustries differ from the standard Global Industrial Classifications (GIC)
system by having three levels of granularity (13 level one industries, 29 level two subindustries,
and 124 level three subindustries) compared with GIC’s four levels (11 sectors, 25 industry groups,
74 industries, and 163 subindustries). Some of these MIC industries were then split into more granular
industry definitions, depending on the company. These definitions were used in our final data set.
For example, the MIC industry definition would have classified Tesla as belonging to the automotive
and assembly industry. In our data set, companies in the automotive and assembly industry were
classified as automotives, electric vehicles, or industrial equipment. As a result, Tesla was mapped

Arena-
into electric vehicles instead of into a broader automotive and assembly industry. Some companies
may have had a different MIC level two classification but were folded into the same industry group
(for example, United Airlines was in the MIC level two air and travel classification, while Aéroports de
Paris was in MIC level two transport and infrastructure, but they both are mapped to our air services
industry). We conducted this kind of reclassification in six cases: consumer durables, air services,
healthcare, consumer internet, freight and logistics, and information-enabled business services.
Exhibit A1 sets out the reclassification of companies from their MIC level two designations
(29 definitions as of 2023) to their final industry designations.1
Exhibit A1
McKinsey Industry Classification (MIC) Industries New industry mapping used in final data set
Advanced electronics • Consumer durables • Industrial machinery

• Industrial electronics • Semiconductors
Aerospace and defense • Aerospace and defense
Agriculture and food production • Agriculture
Air and travel • Air services

• Travel and leisure
Apparel, fashion, and luxury • Apparel and luxury
Automotive and assembly • Automotives • Industrial equipment

• Electric vehicles
Banks • Retail and commercial banks
Basic materials • Building materials and products • Forest products and packaging
• Construction engineering and • Mining
services
Business services • Information-enabled business • Other business services

services
• Recruiting and staffing services
Chemicals • Chemicals
Conglomerates • Conglomerates
Consumer durables • Consumer durables

• Consumer electronics
Consumer packaged goods • Beverages • Packaged food

• Home and personal goods • Tobacco
• Leisure products
1
Because of the six industries that appear in multiple MIC level two classifications, this list has 63 items instead of 57, with six duplicates.

Arena-
Exhibit A1 (continued)
McKinsey Industry Classification (MIC) Industries New industry mapping used in final data set
Consumer services • Consumer services

• Restaurants and food service
Diversified financials • Nonbank financials

• Payments
Healthcare providers • Healthcare
Healthcare supplies and distribution • Healthcare
High tech • Cloud services • IT solutions and services

• Hardware • Software
• Information-enabled business
services
Insurance • Insurance
• Life insurance
Logistics and trading • Freight and logistics

• Wholesale trading
Media • Consumer internet • Media

• Gaming • Video and entertainment
Medical technology • Medical technology
Oil and gas • Oil and gas
Pharma and biotech • Biopharmaceuticals

• Pharmaceuticals
Real estate • Chinese real estate

• Real estate (excl China)
Retail • Consumer internet

• E-commerce
• Everyday and general retail
Telecom • Telecom
Transport and infrastructure • Air services

• Freight and logistics
• Rail and transport
Utilities • Electric power

• Other utilities

Arena-
Computing growth rates and shuffle rates

The growth rate measures the increase in the overall revenue or market capitalization share of an
industry. For example, the industry with the highest revenue growth rate in the period we considered
was e-commerce, which had a 0.08 percent share in 2005 and a 2.40 percent share in 2020, an
increase of 2.32 percentage points—the growth rate metric for e-commerce. The industry with the
lowest revenue growth rate was oil and gas, with a 5.48 percentage-point decrease. E-commerce
had the highest growth rate for market cap, at 3.82 percentage points. Retail and commercial banks
had the lowest growth rate for market cap, at −6.45 percentage points.
A negative growth rate did not necessarily mean that an industry was shrinking, but rather that its
shares of the total revenues in our database had decreased. Oil and gas registered revenue growth
from 2005 to 2020, for example, but other industries outpaced it. Similarly, retail and commercial
banks increased their market cap from 2005 to 2020 but grew at a slower pace than other industries.
The relevant formulas are as follows:
The shuffle rate measures the change in players’ market shares within an industry. An example
appears in chapter 1, in the sidebar “Industry dynamism measured by share shifts.” To compute the
shuffle rate of a given industry, we started with the percentage-point changes in revenue or market
cap share of each player from 2005 to 2020. We then added up all the positive percentage-point
changes for each player.
For example, cloud services had the highest revenue shuffle rate at 87 percentage points, while
aerospace and defense had the lowest revenue shuffle rate at 15 percentage points. In market cap
shuffle rate, electric vehicles had the highest at 97 percentage points, while aerospace and defense
had the lowest at 28 percentage points.
The formulas appear below:
One can also think of the growth rate as a metric that measures change across industries, while the
shuffle rate measures change within industries.
In this report, we showed combined industry shuffle rates and industry share growth rates of market
capitalization and revenue as bubble charts. Exhibits A2 and A3 are individual charts for each of the
four dimensions relevant to demonstrating the arenas’ growth and dynamism.

Arena-
Exhibit A2
In terms of market cap, arenas of today ranked high in both

industry share growth rates and shuffle rates.
Market cap industry share growth rates and shuffle rates for 57 industries
Industry share growth rate Shuffle rate (increase in market cap share
(change in share of global increase in market cap), 2005–20, among companies in each industry), 2005–20,
percentage points percentage points
4 E-commerce Electric vehicles

Software Cloud services
Consumer internet Payments
3 Semiconductors E-commerce
Payments Consumer
internet
Consumer electronics
Consumer
2 Biopharma
electronics
Electric vehicles
Biopharma
Cloud services
Video and audio
Industrial electronics entertainment
1
Video and audio entertainment Industrial
electronics
Information-enabled
0 business services
Software
Information-enabled
business services Semiconductors
–1
–2
–3
–4
–5 0 25 50 75 100
–6
McKinsey & Company

Arena-
Exhibit A3
In terms of revenue, current arenas mostly ranked high in both

industry share growth rates and shuffle rates.
Revenue industry share growth rates and shuffle rates for 57 industries
Industry share growth rate Shuffle rate (increase in revenue share

(change in share of global increase in revenue), 2005–20, among companies in each industry), 2005–20,
percentage points percentage points
4 E-commerce Cloud services

Consumer internet Consumer
internet
Industrial electronics
E-commerce
3 Biopharma
Consumer
Software
electronics
Semiconductors
Software
2 Consumer electronics
Electric
Video and audio entertainment vehicles
Electric vehicles Biopharma
Cloud services Payments
1
Payments Video and audio
entertainment
Information-enabled
0 business services
Industrial
Information-enabled
electronics
business services
Semiconductors
–1
–2
–3
–4
–5 0 25 50 75 100
–6
McKinsey & Company

Arena-
Selection of the 12 arenas of today

Three criteria were used in selecting the 12 current arenas: a high growth rate, a high shuffle rate, and
a minimum market cap in 2020.
We first ranked the 57 industries by their growth rates and shuffle rates for both revenue and market
cap, with one the highest and 57 the lowest. We then created a composite score by adding the
four ranks of each industry—their growth and shuffle rates for both revenues and market cap. In
theory, this would mean that if an industry ranked the highest in growth rate and shuffle rate for both
revenues and market cap, it would have a score of one for all four metrics, for a total composite score
of four. By contrast, if the industry ranked the lowest for all four metrics, or a score of 57 for each, it
would have a total composite score of 228.
We filtered out arenas with market capitalization of less than $800 billion in 2020.
Exhibit A4 lists the top 20 industries by composite scores.

Arena-
Exhibit A4
Growth rate rank Shuffle rate rank Market

Composite capitalization,
Industry Revenues Market cap Revenues Market cap score 2020, $ billion
E-commerce 1 1 4 6 12 3,308
Consumer internet 3 3 3 11 20 3,460
Chinese real estate 2 18 2 5 27 471
Cloud services 17 9 1 2 29 1,225
Electric vehicles 15 8 9 1 33 941
Consumer electronics 11 6 5 12 34 2,502
Payments 18 5 14 3 40 1,643
Biopharmaceuticals 8 7 13 14 42 2,289
Agriculture 12 24 6 8 50 231
Industrial electronics 7 10 19 17 53 2,000
Software 9 2 8 34 53 3,432
Video and entertainment 13 11 16 15 55 1,500
Consumer durables 22 34 10 7 73 245
Information-enabled 23 14 18 23 78 888
business services
Construction engineering 6 41 12 25 84 316

and services
Travel and leisure 41 25 7 13 86 788
Gaming 25 23 20 22 90 396
Semiconductors 10 4 44 37 95 3,495
Recruiting and staffing 29 32 29 9 99 119

services
Healthcare 4 21 55 20 100 786

Arena-
The 12 industries with the lowest composite scores (indicating highest growth and shuffle rates) and a
market capitalization of at least $800 billion in 2020 made up our final list of 12 current arenas.
Other analyses
GDP contribution analysis. We first identified the overall industry sizes of the 12 arenas in 2005
and 2020. This was required as the revenues presented in chapter 1 only include companies in the
analyzed sample, and hence represent only a subset of the total industry. Overall industry sizes
were obtained from McKinsey databases, such as the Global Payments Map, as well as third-
party information sources, such as Statista . For instances where 2005 data was unavailable, the
respective arena 2005–20 CAGR was applied to 2020 data to determine the 2005 industry size by
working backward. As a result, the overall revenue size of the 12 arenas is estimated at $2.8 trillion in
2005 and $13.4 trillion in 2020.
To estimate the GDP contribution of the 12 arenas, we first tried to identify the appropriate ratio to apply
to revenues, to arrive at gross value added (the “ratio”). We used data from the US Bureau of Economic
Analysis (BEA), which contains value-added and gross-output data from 58 industries. We calculated
the ratio between value added and gross output across these 58 industries from 2005 to 2020.
We then mapped each of the 12 arenas with the most closely linked BEA industries and identified the
ratio most appropriate for each arena. For instance, the consumer electronics arena was mapped to
the BEA industry labeled “computer and electronic products.” This industry had an average ratio of
0.66 from 2005 to 2020. The industrial electronics arena was mapped to the BEA industry labeled
“machinery,” which had an average ratio of 0.38. We also looked at the variances of the ratios across
years. While some industries (such as motion picture and sound recording) showed volatility, those
contributed a relatively small proportion of the sample.
We then applied the respective ratios to each of the 12 overall arena industry sizes to identify
the estimated GDP equivalent of arenas. When compared with GDP, this estimated arena GDP
contribution came to $1.6 trillion in 2005 and $7.7 trillion in 2020.
Sensitivity analysis. To test for sensitivity, we replicated analyses of the six key metrics that
differentiated arenas from other industries listed in chapter 1 to account for significant global
events such as the financial crisis of 2007–08 and the start of the COVID-19 pandemic in 2020. For
example, we looked at the movements in economic profit and revenue and market cap growth rates
from 2005 to 2019 (instead of 2020), which resulted in the same outcomes. The takeaways from all
analyses held across all of these checks.
R&D analysis. Company reported R&D expenses were both incomplete and inconsistent at the
firm level within the top 3,000 companies’ databases. We instead used publicly available data. We
used publicly available data from the US National Center for Science and Engineering Statistics,
specifically its Business Enterprise Research and Development (BERD) Survey. We took the total
domestic US private-sector spending on R&D in 2005 and 2020 at the most granular level possible.
This resulted in 54 different industries defined by the BERD survey. We mapped these 54 industries
as either arenas (and arena-adjacent) or non-arenas. For example, we mapped an industry named
“software publishers” in the BERD survey to an arena (software). For large industries such as
pharmaceuticals and medicines, we divided these data points into two categories: pharmaceuticals
(not an arena) and biopharmaceuticals (an arena). The proportion of this split was based on the
revenues of these two industries as listed in the main McKinsey Value Intelligence data set. The
same logic was used to calculate R&D to revenue ratios for arenas and non-arenas, using publicly
available data from the same source on 2020 R&D intensity. R&D intensity is defined as the cost of
US spending on R&D divided by US net sales of companies that performed or funded R&D.

Arena-
Spawning giants analysis. To analyze the share of total market cap of large market cap companies
within arenas and non-arenas, we have used the same McKinsey Value Intelligence company-level
data as was used to identify arenas of today. Companies were grouped by market cap size according
to the following criteria: i) market cap lower than $50 billion; ii) market cap between $50 billion and
$200 billion; iii) market cap equal to or larger than $200 billion. To get to the percentage share of
these companies in arenas and non-arenas, we divided the number of companies in each of these
three market cap size categories by the total number of companies in arenas and non-arenas. To
get to the percentage share of total arenas’ and non-arenas’ market cap held by companies in these
three categories, we have divided the market cap value held by each of these categories by total
arenas and non-arenas market caps.
Revenue sources by geographies analysis. To analyze the source countries of our companies’
revenues, we also used the McKinsey Value Intelligence database, which breaks down company
revenues by country. We then mapped these data points individually into two groups: domestic, if the
source country was where the company was headquartered, and international, if the revenues were
from outside the home country. In the rare instances where the split was unavailable, we would by
default select the revenue earned domestically.
Chapter 2 Early markers of arenas analysis

We used data from PitchBook to identify venture capital (VC) and IPO flows from 2003 to 2007.
PitchBook categorizes VC and IPO flows in two ways: by industry and by vertical. PitchBook has
215 industries and 59 verticals. A transaction is mapped to an industry and at least one vertical.
For example, if VC flows went to a company that provides software for virtual reality gaming, the
company would be mapped to application software (industry) and tagged to gaming (vertical) and
virtual reality (vertical).
We mapped these 215 industries and 59 verticals to our 12 arenas to the extent possible. In a few
instances, the industries or verticals were mapped to several arenas. For example, PitchBook
has an entertainment software industry. This was mapped to two arenas: software, and video and
entertainment. PitchBook also has a mobile vertical. This was mapped to two arenas: consumer
electronics and consumer internet.
We removed instances when a transaction was counted twice because the industry-vertical
combination was mapped to the same arena. For instance, a transaction tagged to the application
software industry (mapped to the software arena) with an AI and machine-learning vertical (also
mapped to the software arena) was counted only once for the software arena’s total.
However, some double counting was unavoidable when a transaction was mapped to several
verticals mapped to different arenas. For example, if the transaction went to a company that provided
an e-commerce application on phones, it could be tagged to three verticals: e-commerce (mapped
to the e-commerce arena), mobile (mapped to the consumer electronics arena), and software as a
service (mapped to the software arena). In this case, while there was no double counting within the
three arenas mapped, the transaction was included in the totals of these three arenas. We estimated
these instances of multiple counting of arenas at around 20 percent of the overall value; they did not
materially change the takeaways of the analysis.
Chapter 3 NOPLAT estimates used

Various benchmarks were used to identify the net operating profit less adjusted taxes (NOPLAT)
margin assumptions used for chapter 3, Exhibit 14. The primary intent of using NOPLAT estimates

Arena-
was to give readers an important additional layer of considerations aside from the potential revenues
presented in the exhibit.
For most of these arenas, all three of our general approaches to benchmarking used historical 2022
NOPLAT margins.
The first approach was used for continuing and spin-off arenas. The benchmark for these arenas
was the originating arena. Seven arenas fell into this category: e-commerce, AI software and
services, electric vehicles, digital ads, cloud services, semiconductors, and streaming video. We
then adjusted our assumptions based on the 2022 NOPLAT of the originating arena, given the
uncertainty surrounding the evolution of the economics of these arenas. For example, e-commerce
saw a 3 percent NOPLAT in 2022 from the same data set we used in chapter 1. We then ranged
this 3 percent and used a 2 to 5 percent assumption. Similarly, we used the NOPLAT benchmark of
13 percent for video and audio entertainment (an originating arena) as our base assumption for the
streaming video potential future arena. This was then ranged to 10 to 15 percent.
The second approach was for potential future arenas that were closely linked to any of the
57 industries in our chapter 1 data set (the 57 are identified above). Seven arenas belonged to this
category: shared autonomous vehicles, space, cybersecurity, video games, industrial and consumer
biotech, modular construction, and future air mobility. We took all relevant industries connected to
these potential future arenas and used the range of those numbers as our benchmark. In instances
when only one industry was used as the benchmark, we ranged those numbers as well, in a similar
way to the ranging we did in the first approach in the previous paragraph. For instance, aerospace
and defense saw an 8 percent NOPLAT in 2022, which we ranged from 5 to 10 percent.
The last approach was for potential future arenas that had existing companies that could be
reasonably used as benchmarks. We considered companies that were relatively mature, because
their economics and margins were likely more stable than those of newer companies or start-ups.
Four potential future arenas were in this category: batteries, drugs for obesity and related conditions,
nuclear fission power plants, and robotics. For example, we used Eli Lilly and Novo Nordisk as our
benchmarks for drugs for obesity and related conditions.
GDP contribution analysis

To estimate the GDP contribution of the potential arenas of tomorrow in both 2022 and 2040,
we followed the same general approach discussed above in chapter 1. We first tried to identify
the appropriate ratio to apply to revenues to estimate value added and applied those ratios to the
estimated arena sizes in 2022 and 2040.
We used data from the US BEA, which contains value-added and gross-output data from
58 industries. We then mapped the 18 arenas of tomorrow with the most closely linked BEA
industries. For instance, the modular construction arena was mapped to the “construction” industry,
and the cybersecurity arena was mapped to the “computer systems design and related services”
industry. These mappings were then matched with the respective value-added to gross-output ratios
from 2005 to 2020, which were mostly stable. These ratios were then multiplied on the respective
low and high case revenue estimates for each of the 18 arenas. This resulted in an estimated total
GDP equivalent of $4.2 trillion in 2022 and $16.4 trillion to $26.9 trillion for the low and high cases,
respectively, in 2040.
To estimate overall GDP, we used GDP data from the World Bank for 2022 as our starting point for
global GDP. The growth rate we used was 2.85 percent, the rate estimated by the OECD for OECD
and G-20 countries from 2020 to 2040.

McKinsey Global Institute
October 2024
Copyright © McKinsey & Company
Designed by the McKinsey Global Institute
mckinsey.com/mgi
@McKinsey_MGI
@McKinseyGlobalInstitute
@McKinseyGlobalInstitute
Subscribe to MGI’s podcast, Forward Thinking:
mck.co/forwardthinking

The Next Big Arenas of Competition Final

Uploaded by

Copyright:

Available Formats

The Next Big Arenas of Competition Final

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Next Big Arenas of Competition Final

Uploaded by

Copyright:

Available Formats

The next

Copyright © 2024 McKinsey & Company.

The next big arenas of competition ii

MGI Directors MGI Partners

The next big arenas of competition 1

Chapter one. The arenas of today 17

Arenas of tomorrow compendium 83

The next big arenas of competition 2

The next big arenas of competition 4

The next big arenas of competition 6

The arenas of today

In 2005, arenas generated less than

The next big arenas of competition 8

Market cap shuffle rates and growth share for 57 industries

Industry share growth rate (change in share of global increase

McKinsey & Company

Today’s arenas stand out from other industries in six ways.

The next big arenas of competition 9

The next big arenas of competition 10

The origins of arenas

The next big arenas of competition 11

The arenas of tomorrow

Some arenas of today will

The next big arenas of competition 12

The 18 potential arenas of tomorrow could generate $29 trillion to

18 potential arenas Revenue, 2040 estimate (CAGR, Profit, 2040

Space 300 960–1,600 (7–10) 50–160 (5–10)

Cybersecurity 160 590–1,200 (8–12) 90–240 (15–20)

Batteries 98 810–1,100 (12–14) 40–110 (5–10)

Modular construction 180 540–1,100 (6–10) 20–220 (4–20)

Streaming video 160 510–1,000 (6–11) 50–150 (10–15)

Video games 230 550–910 (5–8) 80–180 (15–20)

Robotics 21 190–910 (13–23) 20–180 (10–20)

Industrial and consumer biotech 140 340–900 (5–11) 10–270 (4–30)

McKinsey & Company

The next big arenas of competition 13

The next big arenas of competition 14

Five arenas of today—biopharmaceuticals, consumer electronics, information-enabled business

The next big arenas of competition 15

The arenas of today

Arenas are exceptional industries

The next big arenas of competition 17

Today’s 12 arenas grew quickly and were dynamic

The next big arenas of competition 18

Company ranking by market share of revenue

Change in ranking Revenue share, % Positive shifts,

Source: McKinsey Value Intelligence; McKinsey Global Institute analysis

McKinsey & Company

The next big arenas of competition 19

Company ranking by market share of revenue

Change in ranking Revenue share, % Positive shifts,

Source: McKinsey Value Intelligence; McKinsey Global Institute analysis

McKinsey & Company

The next big arenas of competition 20

Market cap shuffle rates and growth share for 57 industries

Industry share growth rate (change in share of global increase

McKinsey & Company

The next big arenas of competition 21

Revenue shuffle rates and growth share for 57 industries

Shuffle rate E-commerce, 888