2001 11 Skate To Where The Money Will Be
2001 11 Skate To Where The Money Will Be
2001 11 Skate To Where The Money Will Be
Reprint r0110d
ILLUSTRATION: BILL HALL
72 Copyright 2001 by Harvard Business School Publishing Corporation. All rights reserved.
As hockey great Wayne Gretzky used to say,
the key to winning is getting rst to where the puck
is going next. The same could be said
about succeeding in business, and a new theory
of protability could help you do just that.
Skate
to Where the
Money
Will Be
by Clayton M. Christensen,
Michael Raynor,
and Matthew Verlinden
november 2001 73
S kat e t o W h e re t h e M o n ey Wi l l B e
During the early days of the computer industry, for ex- service more quickly, with fewer concerns about the un-
ample, when mainframes were not yet powerful or fast intended consequences of reconguring their own cen-
enough to satisfy mainstream customers needs, an inde- tral-office facilities. Regulatory mandates cannot decou-
pendent contract manufacturer assembling machines ple an industry at an interdependent interface. As long
from suppliers components could not have survived be- as DSL service is not good enough to satisfy most users,
cause the way the machines were designed depended on the integrated telephone companies will be able to pro-
the way they were manufactured and vice versa. Nor vide better, more reliable service than nonintegrated
could an independent supplier of operating systems, core competitors.
memory, or logic circuitry have survived because these
key subsystems had to be designed interdependently, too.
When the product isnt good enough, in other words,
Going to Pieces
being an integrated company is critical to success. As the Product performance almost always improves beyond the
most integrated company during the early era of the com- needs of the general consumer, as companies stretch to
puter industry, IBM dominated its world. Ford and Gen- meet the needs of the most demanding (and most prot-
eral Motors, as the most integrated automakers, domi- able) customers. When technological progress overshoots
nated their industry during the era when cars were not what mainstream customers can make use of, companies
good enough. For the same reasons, RCA, Xerox, AT&T, that want to win the business of the overserved customers
Alcoa, Standard Oil, and U.S. Steel domi-
nated their industries at similar stages.
Their products were based on the sorts of
proprietary, interdependent value chains
The Disruptive Technologies Model
that are necessary when pushing the fron-
tier of what is possible. The disruptive technologies model contrasts the pace of techno-
When a nonintegrated company tries to logical progress with customers ability to use that progress. Ac-
compete under these circumstances, it usu- cording to the model, there are two types of performance trajec-
ally fails. Stitching together a system with tories in every market. One trajectory, depicted by the shaded
other partnercompanies is extremely dif- area, shows how much improvement in a product or service cus-
cult when the subsystems and expertise tomers can absorb over time. The other trajectory, shown by the
those companies provide are interdepen- solid lines, depicts the improvement that innovators in the in-
dent. We could offer numerous historical dustry generate as they introduce new and enhanced products.
examples, but there are plenty of illustra-
tions from industries that are still emerg-
Performance
ing. In the late 1990s, for example, many y
tor
nonintegrated companies attempted to ra jec ogy
t l
offer high-speed DSL access to the Internet ce no
m an tech
r t
over phone lines operated by telephone rfo en
Pe pres anding cust
omers Performance
of Most- dem
}
companies. Most of these attempts failed. that customers
in the main-
Many believe that low prices for DSL ser- stream market
vice that were rooted in regulatory pecu- cu st om ers can absorb
anding
Least-dem
liarities of the Telecommunications Act of ce
an
1996 are what drove the competitive local rm
e rfo
p
exchange carriers toward bankruptcy. This w ory
Disruptive Neaject
was only the proximate cause of their technology tr
demise, however. The fundamental issue is Time
that at this point in the industrys evolu-
tion, DSL technology isnt good enough Almost always, this second trajectory the pace of technological
yet, and there are, as a result, too many un- innovation outstrips the ability of customers in a given tier of
predictable interdependencies between the market to absorb it. This creates the potential for innovative
what focused DSL providers need to do companies to enter the lower tiers of the market with disruptive
and what the telephone companies must technologies cheaper, simpler, more convenient products or
do in response. The incumbent phone services. Almost always, the leading companies are so absorbed
companies capacity to span the whole with upmarket innovations addressed to their most sophisticated
value chain has been a powerful advan- and protable customers that they miss the disruptive innova-
tage. They understand their own network tions. Disruptive technologies have caused many of historys best
architectures and can consequently offer companies to plunge into crisis and fail.
november 2001 75
S kat e t o W h e re t h e M o n ey Wi l l B e
in less-demanding tiers of the market are forced to consequence. The exhibit The Dis-Integration of the
change the way they compete. They must bring more ex- Computer Industry illustrates how this happened in
ible products to market faster and customize their prod- that eld. During its early decades, the dominant com-
ucts to meet the needs of customers in ever smaller mar- panies were integrated across most value-chain links
ket niches. because competitive conditions mandated integration.
To compete on these new dimensions, companies must As the personal computer disrupted the industry, how-
design modular products, in which the interfaces be- ever, it was as if the industry got pushed through a bo-
tween components and subsystems are clearly specied. logna slicer. The dominant, integrated companies were
Ultimately, these interfaces coalesce into industry stan- displaced by specialists that competed in horizontal
dards. Modular architectures help companies introduce strata within the value chain.
new products faster because subsystems can be improved This shift explains why Dell Computer was so success-
without having to redesign everything. Companies can ful in the 1990s. Dell did not succeed because its products
mix and match the best components from the best sup- were better than those of competitors IBM, Compaq,
pliers to respond to the specic needs of individual cus- and the like. Rather, overshooting triggered a shift in
tomers. Although standard interfaces invariably force the basis of competition to speed, convenience, and
compromises in system performance, competitors aim- customization, and Dells business model was a perfect
ing at overserved customers can comfortably trade off match for that environment. Customers were delighted
some performance to achieve the benets of speed and to buy computers with outsourced subsystems, custom-
exibility. assembled to their own specications and delivered in-
Once a modular architecture and the requisite indus- credibly quickly at competitive prices. This also explains
try standards have been dened, integration is no longer how Cisco, with its disruptive router and its noninte-
crucial to a companys success. In fact, it becomes a grated business model, bested more integrated com-
competitive disadvantage in terms of speed, exibility, petitors like Lucent in the market for telecommunica-
and price, and the industry tends to dis-integrate as a tions equipment.
Mainframes and minicomputers were never good own proprietary designs and components. The PC,
enough or fast enough or cheap enough to create a mass though, very quickly became good enough for the
market and were therefore always the province of large, average consumer, giving rise to an army of specialized
integrated players who built their machines from their players.
november 2001 77
S kat e t o W h e re t h e M o n ey Wi l l B e
end-use products typically make the most money, for two cate anything you can do very quickly. And because most
reasons. First, the interdependent, proprietary architec- of the costs in an outsourcing-intensive business model
ture of their products makes differentiation straightfor- are variable rather than xed, there are minimal econo-
ward. Second, the high ratio of xed to variable costs, mies of scale, so that large and small competitors have
which is inherent to the design and manufacture of archi- similar costs. Making an undifferentiated product at un-
tecturally interdependent products, creates steep econo- differentiated costs is a recipe for earning undifferenti-
mies of scale. Larger competitors can amortize high xed ated prots.
costs over greater volume, giving them strong cost advan- So, whats our Compaq engineer to do? Shell put pres-
tages over smaller competitors. Making highly differenti- sure on her suppliers to invent faster microprocessors and
ated products with strong cost advantages is a license to higher-capacity, lower-cost disk drives.
print money, and lots of it. Overshooting at the system level often throws the sub-
Hence IBM, as the most integrated competitor in the system suppliers back to a stage where their product is
mainframe computer industry, made 95% of the indus- not good enough for what the system assembler needs.
trys prots from just a 70% market share. And from the Competitive forces consequently compel the subsystem
1950s through the 1970s, General Motors garnered 80% of suppliers to create architectures that are increasingly in-
the prots from about 55% of the U.S. auto market. Most terdependent and proprietary as they try to push the
of IBMs and GMs suppliers, by contrast, survived on sub- bleeding edge of performance. They have to do this to win
sistence prots year after year. the business of their immediate customers, who are the
But when the large integrated players overshoot what designers and manufacturers of modular products.
their mainstream customers can use, the tables begin to Hence, as a natural and inescapable result of the shift in
turn. Disruptive competitors begin to move upmarket, industry structure, the place where companies are used to
and the power to make money shifts away from compa- making a lot of money the end-user stage becomes un-
nies that design and assemble the end-use product toward likely to be the place where money will be made in the fu-
the back end of the value chain to those companies that ture. And, conversely, the places where attractive prots
supply subsystems with internal architectures that are were rarely made in the past components and subsys-
still technologically interdependent. tems often become highly protable.
A good way to visualize this is to imagine an engineer The exhibit Where the Money Went in the PC In-
employed at Compaq whose boss just told her to design dustry illustrates how this worked in the desktop com-
a desktop computer better than Dells, IBMs, or Hewlett- puter market in the 1990s. Initially, money owed from
Packards. How would she do it? When designing and as- the customer to the companies that designed and assem-
sembling a modular product, your competitors can repli- bled computers; but as the decade progressed, less and
Desktop computer
As PCs became good enough for main- designer/assemblers
stream users, prots owed from the
customers through the assemblers
(the IBMs and Compaqs of the world)
to lodge in the component makers
the operating system maker (Micro-
soft), the processor maker (Intel), and
initially to the memory chip makers
and disk drive manufacturers. But as Operating system Microprocessor DRAMs Disk drives
DRAM chips and drives became good
enough for the assemblers, the money
owed even further up the value chain
to DRAM equipment makers and head
and disk suppliers. Equipment makers Heads and disks
november 2001 79
S kat e t o W h e re t h e M o n ey Wi l l B e
less of it stopped there as prot. Quite a bit of this money off asset-intensive units that design and manufacture
owed over to operating system maker Microsoft and components to companies that see in those same opera-
lodged there. Another chunk owed to processor manu- tions the opportunity to create subsystems whose archi-
facturer Intel and stopped there. Money owed to the tectures are progressively more interdependentthus im-
DRAM chip makers such as Samsung and Micron Tech- proving the numerator of their ROA ratio. Lucents recent
nology as well, but not much of it stopped there. It owed spin-offs of its component and manufacturing operations
through them and accumulated instead at companies like is an example. This seems perfectly logical and necessary,
Applied Materials, which supplied the chip-manufacturing given the increasingly modular character of many of Lu-
equipment that the DRAM makers used. Similarly, money cents systems. But with perfect predictability, this pres-
owed right through the assemblers of disk drives such as sure from Wall Street to boost ROA forces companies to
Quantum and lodged at the stage where heads and disks skate away from the place where the money will be made
were made. in the future.
Whats different about the places where the money col- This scenario could soon play out in one of IBMs busi-
lected and those where it didnt? For most of this period, nesses. Through the 1990s, the capacity of the 2.5-inch
prots lodged with the products that were the ones not disk drives used in notebook computers tended to be in-
yet good enough for what their immediate customers adequate. True to form, their architectures were interde-
needed. The architectures of those products therefore pendent, and the design and assembly stage was very
like IBM can exibly couple and decouple its operations, change in the basis of competition, which has precipi-
rather than irrevocably sell off operations, it has greater tated a change in architecture, which has forced the dom-
potential than a nonintegrated company to thrive from inant, integrated rms to dis-integrate.
one cycle to the next. To become fast and exible, IBMs PC business out-
sourced its microprocessor to Intel and its operating system
Where Will the Money Be to Microsoft. But in the process, IBM hung onto where the
money had been the design and assembly of the com-
in the Auto Industry? puter system and put into business the two companies
We believe this model can help managers, strategists, and that were positioned where the money would be. GM and
investors in a wide variety of industries see into the future Ford, with the encouragement of their investment
with greater clarity than the traditional tools of historical bankers, have just done exactly the same thing. They have
data analysis have allowed. When we consider, for exam- spun out the pieces of the value chain where the money
ple, where the money in the automobile industry will go will be in order to stay where the money has been.
in the future, the car companies seem to be falling into ex- Ford and GM had no choice but to decouple their com-
actly the same trap that IBM did some 15 years ago. ponent operations from their design-and-assembly busi-
While automobiles often used to rust or fall apart me- nesses. Indeed, they gave their shareholders the option of
chanically well before their owners were ready to part owning one or both. But rather than an irreversible di-
with them, auto quality now has overshot what most cus- vestiture, they might have taken a page from IBMs re-
tomers want or need. In fact, the most reliable cars usu- cent forays into opportunistic decoupling, ignored the
ally go out of style long before they wear out. As a result, siren song of investment bankers, and found a way not to
the basis of competition is changing. Whereas it used to shed those asset- and scale-intensive businesses where
take six years to design a new car model, today it takes the numerator of the ROA ratio will likely be more at-
less than two. Car companies routinely compete by cus- tractive in the future. This will be especially true if shifts
tomizing features to the whims of smaller and smaller in customer demand mandate some sort of reintegration in
market niches. In the 1960s, it was not unusual for a the future.
model to sell a million units a year. Today, the market is Managers of the slimmed-down automakers can still do
far more fragmented: If you sell 200,000 units of a par- well, but theyll need to dramatically change the way they
ticular model, youre doing ne. Some makers now do business in the design-and-assembly stage. They need
promise that you can walk into a dealership, custom to do in their industry what Dell did in the computer in-
order a car exactly to your desired conguration, and dustry become consummately fast, exible, and conve-
have it delivered in ve days roughly the response time nient. Overshooting changes the game. If GM and Ford
that Dell Computer offers. can play this new game better than competitors, they can
To compete in this way, automakers are adopting mod- still prosper, much as Dell did in the 1990s against com-
ular architectures for their mainstream models. Rather petitors who hadnt mastered the new rules as effectively.
than knitting together individual components from di-
verse suppliers, theyre procuring subsystems from fewer The implications of these ndings are clear. The power
tier-one suppliers. The architecture within each subsys- to capture attractive prots will shift in the value chain to
tem braking, steering, chassis, and the like is becoming those activities where the immediate customer is not yet
progressively more interdependent as these suppliers satised with the functionality of available products. It is in
work to meet the auto assemblers performance and cost these stages that complex, interdependent integration oc-
demands. Inevitably, the subsystems external interfaces cursactivities that create steeper economies of scale and
are becoming more modular because the economics of greater opportunities for differentiation. The power will
using the same subsystem in several car models more shift away from activities where the immediate customer
than compensates for any compromises in performance is more than satised because it is there that standard,
that might result. modular integration occurs. In most markets, this power
As the basis of competition has shifted, the vertically in- shift occurs tier by tier in a way that is quite predictable.
tegrated automakers have had to break up their value Executives whose companies are currently making lots
chains so they can more quickly and exibly incorporate of money ought not to wonder whether the power to earn
the best components from the best suppliers. GM subse- attractive prots will shift, but when. If they watch for the
quently spun out its component operations into a sepa- signals, quite possibly they can prosper in all cycles, rather
rate company, Delphi Automotive Systems, and Ford has than in only one.
spun out its component operations as Visteon. Thus, the
same thing is happening to the auto industry that hap- Reprint r0110d
pened to computers: Overshooting has precipitated a To place an order, call 1-800-988-0886.
november 2001 81