Network effects

Building ecosystems of value

SANGEET PAUL CHOUDARY

Best-selling author of the books Platform Scale and Platform Revolution
 ABOUT THE AUTHOR

SANGEET PAUL CHOUDARY

is the founder of Platformation Labs and the best-selling author of the books Platform Scale and Platform Revolution. He has been ranked
as a leading global thinker for two consecutive years by Thinkers50, ranking among the top 30 emerging thinkers globally in 2016
(Thinkers50 Radar) and ranking among the top 50 thinkers of Indian origin in 2015 (Thinkers50 India).
He is the co-chair of the MIT Platform Strategy Summit at the MIT Media Labs and an Entrepreneur-in-residence at INSEAD Business
School. He is also an empaneled expert on the global advisory council for the World Economic Forum’s initiative on the Digital
Transformation of Industries. His work has been featured as the Spotlight article on Harvard Business Review (April 2016 edition) and the
themed Business Report of the MIT Technology Review (September 2015).
As the founder of Platformation Labs, Sangeet is an advisor to leading executives globally. He is also an empaneled executive educator with
Harvard Business School Publishing, and has advised the leadership of Fortune 500 firms, family-owned conglomerates, and key
government bodies.
He is frequently quoted and published in leading journals and media including the Harvard Business Review, MIT Technology Review, MIT
Sloan Management Review.The Economist, The Wall Street Journal, WIRED Magazine, Forbes, Fortune, and others. Sangeet is a
frequently sought after advisor to CXOs globally on the topic of digital transformation and also serves as a fellow at the Centre for Global
Enterprise in New York.
He is a frequent keynote speaker and has been invited to speak at leading global forums including the World Economic Forum’s Annual
Meeting of the New Champions (Summer Davos), the WEF ASEAN Summit, and the G20 Summit 2014 events. Sangeet has a bachelors in
computer science from IIT Kanpur and a masters in management from IIM Bangalore.

CONTRIBUTING AUTHORS ABOUT PLATFORMATION LABS

Marshall Van Alstyne and Geoffrey Parker Platformation Labs is C-level executive advisory firm and think-tank, focused on the analysis
are contributing authors to the research and implementation of platform business models and network effects towards the digital
published by Platformation Labs, including transformation of industries. Platformation Labs has advised governments, Fortune 100
the books Platform Revolution (co- firms and high growth startups in 40+ countries across the Americas, Europe, Africa and
authors) and Platform Scale. Asia-Pacific. Our thought leadership and intellectual capital are commissioned and licensed
by leading consulting firms globally and have been featured in leading global forums.
THE NETWORK EFFECT ISN’T
GOOD ENOUGH!
The concept of stored value: How network effects can be strengthened further.

Sangeet Paul Choudary

This essay originally appeared on TechCrunch. I had the pleasure of co-authoring it with Nir Eyal. Nir
runs a great blog about the intersection of psychology, technology, and business at NirAndFar.com.

If there is one altar at which Silicon Valley worships, it is the shrine of the holy network effect. Its mystical
powers pluck lone startups from obscurity and elevate them to fame and fortune. The list of anointed
ones includes nearly every technology success story of the past 15 years. Apple, Facebook, Microsoft,
eBay, and PayPal, have each soared to multi-billion-dollar valuations on the supreme power of the
network effect.

But today, the power of the network effect is fading, at least in its current incarnation. Traditionally
defined as a system where each new user on the network increases the value of the service for all
others, a network effect often creates a winner-takes-all dynamic, ordaining one dominant company
above the rest. Moreover, these companies often wield monopoly-like powers over their industries.

IN THE BEGINNING

Once, all a company needed to do to leverage the network effect was facilitate communication between
a critical number of customers. If enough people used a particular system to exchange information, a
leader would emerge and become the de facto platform. Companies who could either form a
marketplace or facilitate the flow of information between parties became tremendously powerful as
central hubs of data transfer.

In fact, the first network effects platform was Bell Telephone, which established a government-
sanctioned monopoly nearly 100 years ago. Since then, successful network effects businesses have
sung from essentially the same hymnal.

First, establish a medium of communication by building the required infrastructure or inventing a new
technology. For example, lay down telephone wires from coast to coast. Then, provide access to the
network to improve the ease of information transfer — say, by selling fax machines. Finally, race to grow
the user base before competing services do. If you get bigger faster than your competitors, voilà! You’re
inside the pearly gates.
RAPTURE
That’s the plan at least. But today, things are not quite so simple. For one, in the old days, consumers paid to
access the network through their upfront investment in hardware. These upfront costs locked users into the
network and once they were in, they were in for good, thus erecting barriers to entry for would-be competitors.
However, the cost of providing access to the network has fallen precipitously. The days of customers buying
expensive hardware to use a network are gone as is the correlating lock-in effect.

CONVERTS
In addition, to access costs falling to zero, another key component of what once kept users locked into a
network has vanished. Once, porting contacts onto a new network, like switching instant messaging services
from Yahoo! to AIM, was a non-trivial task.

Today, however, customers use their Facebook, Twitter or Google profiles to join a new service in seconds. A
burgeoning network, take Instagram or Pinterest, can leverage the single sign-on enabled by the social graph
to reach critical mass faster than ever before. Users not only port their personal information but bring their
connections as well. In the age of the social web, the convenience of the social graph has largely toppled the
lock-in that once kept users bound to one network over another.

TENDING THE FLOCK

Without the upfront investment in physical hardware and users’ newfound ability to port personal information
and contacts, how is a company to retain its users? Is the network effect’s ability to lock-in users dead?
Hardly.

The power to leverage the network effect now resides in “stored value.” Unlike network access costs, stored
value is an investment that comes in small increments with repeated use, increasing the importance of the
service the more a user engages with it.

STORED VALUE
Stored value comes in four forms, and companies leverage these tiny investments to build lock-in to their
service and retain users.
Creative content (e.g. Pinterest, Facebook, Instagram): Users invest in creating a portfolio of creative
content, which forms the basis of their interactions on the platform1. The quality and quantity of the
content result in more interactions with other users, which, in turn, provides greater value to the content
creator.

Reputation (e.g. TaskRabbit, Airbnb, Stack Overflow): Although marketplaces for physical goods, such
as eBay, have been around for some time, services marketplaces have grown in popularity lately. Trust
is an important component of this new breed of network effects business. As a result, reputation built on
the platform directly contributes to greater value for all users. Building reputation on a platform requires
consistent delivery of highly rated services and may also involve qualifying for some minimum criteria set
forth by the platform. Hence, once a service provider builds reputation on a platform, it prevents her from
migrating to a competing platform.

Usage Data: Users store value in the form of data, either by actively collecting information, such as in
the case of Dropbox or Reddit, or passively as their usage improves the service by offering more
relevant information, such as is the case with Quora, which delivers a personalized news feed based on
usage. The more a user consumes information through the platform, the more intelligent the algorithm
becomes in recommending pertinent content to the user. In both cases, the data set built by or for the
user delivers greater value with increased usage, something that won’t directly be available on a
competing platform.

Influence (e.g. Twitter, YouTube channel subscriptions): Networks that utilize a one-sided follow model
create an influence dynamic. Unlike importing contacts or “friending” people, collecting followers is
largely outside the direct control of the user. With the exception of sketchy tactics banned by the Twitter
terms of service, accruing more Twitter followers can only be done by tweeting content others find
interesting enough to share. As the user’s follower count grows, so does the stored value in the network
and the incentive to stay actively engaged.
KEYS TO THE KINGDOM

Creating a network effect is not what it used to be. Today, stored value created by the
users reinforces the power of the network effect to retain users and grow market share.
This dynamic makes creating user habits all the more important as investments of
stored value only occur through successive passes through the user experience (see
Nir’s previous article and video).

With the portability of the social graph and the fall of upfront costs to join a network,
companies must leverage new ways of acquiring and retaining users. Business models
that leverage a network effect plus stored value, hold the keys to the kingdom.3
If you liked reading this post, you might want to get a free digital copy of the forthcoming
book PLATFORMED
 THE NETWORK EFFECT PLAYBOOK:
SOCIAL PRODUCTS WIN WITH UTILITY,
NOT INVITES
When everyone is building new social products, how do you
get users to choose yours over the others?

Sangeet Paul Choudary

The proverbial chicken and egg problem of building a new social product is well understood among tech
startups, and it’s been commonplace to follow two contrasting mechanisms for getting traction.

Traditionally, startups have solved this problem by racing to connect users with each other, essentially
providing them the pipes to interact with each other. Twitter, Facebook and LinkedIn have grown big
with this connection-first model.

However, a new breed of networks is gaining ground with the content-first model. They provide users
with tools to create a corpus of content, and then enable conversations around that content. Behance,
Pinterest, Instagram, Dribble, Scoop.It have all gained traction by building a corpus of content before
building a social network.

The rules of building a social product are changing. It’s important to understand this shift to build social
products that can effectively gain traction on the internet today.The connection-first model is no longer
as effective as it used to be. As the social web grows, and a larger number of social products compete
for our attention, we are seeing a dramatic shift towards the content-first model. If you’re still getting
users to send out Facebook invites, you’re adding to the noise, instead of standing out and getting
noticed.

The Connections-first Social Product

Traditionally, the playbook for building network effects has been the following: Get users on board,
connect them to each other and have them create content and conversations.

Social networks like Bebo, Facebook and Twitter used this playbook to create their respective networks
leveraging address-book integrations and other hacks to rapidly build a large number of network
connections.
The importance of building connections, in this model, cannot be emphasized enough. In fact, the
growth teams at Facebook, Twitter and LinkedIn specifically aim for ‘X connections for a user within Y
days of sign-up’ to activate the user.
Since a critical mass of connections is required before users experience value, the key to building a successful
network is minimizing the friction in creating connections. Contact-list integration helped social networks like
Facebook and LinkedIn gain initial traction through the removal of sign-up friction.

In spite of growth hacks like contact-list integration, there is always a lead time in getting users on board and
reaching critical mass. This is the ‘gap’ where it becomes very difficult to demonstrate value in using the
product.

Frictionless sign-up + Virality = Network Effects? Or not

!
Startups often believe that removing friction in sign-up and creating some form of viral acquisition are the two
key elements to reaching critical mass. In fact, with the rise of Facebook Connect and the social graph, a large
number of social products have sprung up on the promise of frictionless sign-up and viral growth. However,
users on the internet have limited time and attention. As more startups leverage the social graph and flood
users with invitations to join their networks, users have started to develop invite fatigue.

Clearly, frictionless sign-up and virality are not the one-stop solutions we were hoping they would be.

The secret to network value

Startups often fail to appreciate the gap between technology and value proposition. For products like Evernote,
technology serves the entire value proposition. However, for social products, the value proposition is a
combination of technology and the content that users create on top of it. YouTube’s value lies in its hosting
and streaming capability, but more importantly in its vast repository of videos.

The secret to creating a social product that demonstrates immediate value is to enable content before creating
the network.

Content created on the network is the new source of competitive advantage. The videos on YouTube, the
pictures on Instagram, the answers on Quora are the primary source of value for users and the key driver of
competitive advantage for these platforms.
The Content-first Social Product
Today’s social startups don’t start off as networks. They start off as standalone apps. These products
enable users to create a corpus of content first. They then connect the users with each other as a
consequence of sharing that content.

Instagram started out as a photo-taking tool and built itself out into a social network subsequently. The
initial focus was entirely on the creation of content and the connections were formed over time
leveraging other social networks. It is unlikely that Facebook would have considered Instagram a direct
competitor in its early days, largely owing to its model of deferring network creation.

How to create a network in stealth mode

Instagram started off as a standalone tool. In doing so, the product provides ‘single-user’ utility to the
user even when other users aren’t around on the network. There are two aspects to building single-user
utility:

1. The single-user utility should allow creation of content that will ultimately form the core of the
network. The core of Instagram is pictures. Discussions are centered around pictures. Hence, the
single-user tool needs to allow creation of pictures. This is an extension of the OpenTable model, where
a restaurant first manages its real-time seating inventory on a single-user tool, before that very
inventory is exposed to consumers on a network, to allow them to reserve tables. Curation-as-creation
products like ScoopIt and Storify also use this model to curate content which will serve as the core for
network interactions.

2. The product should deliver greater value when users share their content with their
friends. The product builds out the network at the backend as more content is shared. Hence, the
social network gets created, effectively solving the chicken and egg problem. A new breed of curation-
as-creation startups (Scoop.It, Paper.Li etc.) is gaining traction on a similar model.
1. The new playbook for creating social products is essentially the following:

2. Have a vision for creating the network but do not start executing on network creation

Enable a single-user tool that creates content that is core to social interactions

Share this content on external networks (social networks, email, blogosphere)

Capture interactions around the content to build network linkages at the backend

Open out the network once a critical mass of linkages have been built

The rise of the content portfolio

Instagram demonstrates how a network is created around a portfolio of user-generated content. Behance and
Dribbble have followed similar strategies by providing a portfolio for hosting designs, before adding value
through the creation of a peer-review community. Initially, Pinterest appealed to the designer community as a
tool to ‘bookmark’ their favorite designs, before it built out the network. Early adopters found enough value in
the ability to store designs and pictures, to use the product before the network became active.
The new success factors
Frictionless sign-up and virality are important but they are no longer the key to building social products. The
following are key to building content-first social products:
THE SCALABLE STARTUP
TEST: BUMPS VS. ENGINES

Why do some startups grow faster than the rest?

Sangeet Paul Choudary

“A startup is a company designed to grow fast… The only essential thing is growth. Everything else we
associate with startups follows from growth.” - Paul Graham

Paul Graham differentiates between startups and new businesses on one particular parameter, the
potential for scale and hyper-growth. A startup’s potential to achieve hyper-growth and rapid scale is
largely dependent on the types of growth strategies it implements.

Scaling a startup is a lot like starting a car. There are two ways of starting a car and making it run. You
can drive another car and ‘bump’ into the first car to make the first car move. The car does move with
that ‘bump’, but that movement isn’t sustainable. To make a car run sustainably, you need to fit it with
an engine.

Startups grow in similar ways. Most growth strategies fall into one of two buckets: Bumps or Engines.

Bumps
‘Bump’ strategies – PR, advertising, events – give a startup some traction and exposure. However,
these strategies require commensurate investment of time, effort or money. Advertising works as long
as you pump in money. PR and events aren’t repeatable or scalable; you need to invest time and effort
every time you run them.

Engines
In contrast, startups that build growth ‘engines’ scale rapidly and achieve exponential growth. These
startups do not rely exclusively on external marketing channels; they build an engine for growth within
the product. They are designed to grow as a consequence of product usage.

Growth by Design
Startups with growth ‘engines’ experience growth that scales with adoption. As more users use the
product, the product’s growth rate increases. More pictures created and shared from Instagram expose
Instagram to even more users. As users create and send out surveys from SurveyMonkey, survey
recipients get exposed to the product and come on board to create their own surveys. KickStarter’s
project creators spread the word about KickStarter every time they promote their project. All these
products are designed to get greater exposure through usage. That is a common design pattern that we
see repeated across scalable startups. As more users use the product, the product gets exposed to
new users, leading to greater growth.
Growth-first Products
Startups that are designed for scale do not see growth as something divorced from the product. They
implement growth within the product, much like an engine sits within a car and powers the car forward. They
leverage users to expose their product to more users. Many startups erroneously believe that getting users to
send out invites equals creation of such a growth engine. On the contrary, startups designed for scalable
growth rarely rely on invites. They rely on users to share elements of the product with their network. YouTube
users share videos, KickStarter users share projects and AirBnB users share listings.

Intelligent Piggybacking
These startups also, invariably, piggyback on an external network to achieve rapid growth. Instagram used
Facebook to achieve rapid traction initially. AirBnB allowed users to cross-post listings to Craigslist, which
accelerated its initial growth1. Paypal piggybacked on eBay’s growth initially and YouTube first got traction
when MySpace users started using it to display videos on MySpace. In all these examples, users on the
external network get exposed to the product and transition to becoming active users of the product.

The Scalable Startup Test

A scalable startup can never achieve hyper-growth purely on the strength of strong marketing budgets and
external ‘bump’ strategies. It needs a growth ‘engine’ that powers growth from within. It needs to design
growth that accelerates with usage. Finally, it needs to create the hooks and motivations that will enable and
incentivize users to expose the product to others every time they use the product.

If you’re creating a marketing plan for your startup, think of what you’ve got and slot those down into bumps
and engines. Bumps are important to get initial traction, which then acts as fuel for the engine. But if bumps
are the only thing you’ve got, you probably aren’t building a startup that can scale.

The Curse of the Well-Funded Startup

Most startups lament the fact that they haven’t got the money to spend on marketing. Ironically, some of the
fastest growing startups of all time, e.g. AirBnB, Instagram, didn’t have the money to spend on marketing.
Marketing dollars make one complacent about growth. In fact, the lack of marketing dollars forces one to
figure out how to build engines for growth that will not require commensurate investment. More often than
not, well-funded startups fail because of an over-reliance on those very funds. Access to cash numbs the
need to create organic growth.
VIRALITY VS. NETWORK EFFECTS
While often confused, the two are actually very different.

Sangeet Paul Choudary

Network Effects and Virality are often confused in the online world, possibly because the two often
occur together and, in such cases, end up reinforcing each other.

Network effects and Virality are, however, completely different. There are many products which have
network effects but are not viral. Conversely, many viral products do not have network effects.

Quick Definitions
A product with network effects gets more valuable as more users use it. Network effects are achieved
only after a certain critical mass is reached but can prove to be a very strong source of value and
competitive advantage beyond that point.

A viral product is one whose rate of adoption increases with adoption. Within a certain limit, the product
grows faster as more users adopt it.

Same Same But Different

Both network effects and virality tend to magnify value and growth respectively as more users use the
product. This is probably why the two concepts are often confused. However, as elaborated above, the
two actually mean very different things.

In fact, there are many products that exhibit virality without exhibiting network effects. A case in point
being email and cross-platform communication products. A key feature here is that they are either
interoperable across networks (Hotmail) or leverage an underlying network for both the viral
transmission as well as delivery of the value proposition. In the case of SurveyMonkey, Eventbrite etc.,
that underlying network may be mail, a social network or even a blog.

There are many others that exhibit network effects without exhibiting virality. Products with indirect
network effects such as marketplaces may not grow virally. In such cases, network effects are a result
of aggregation of the two sides and while each side can be brought on virally through some incentive,
it’s very difficult to leverage the indirect network effect to get users on one side to come on through
invitations or interactions from the other side.
The following graphic shows a quick overview of how these products stack up and should help clarify the
difference between virality and network effects

Virality Network Effects

 HOW TO GET USERS TO
CONTRIBUTE ON A CREATIVITY
PLATFORM
A framework for structuring user-contribution ecosystems.

Sangeet Paul Choudary

Let’s loosely define creativity platforms as platforms that allow creators to express themselves, and
possibly get an audience for their ‘creations’. YouTube, Dribbble, 500px, Instagram, Flickr… the list
goes on.

A platform without creators is a ghost town and there is little incentive for consumers to use it.
Replicating the technology of YouTube is a considerably smaller challenge compared to replicating its
community of video creators.

The creators are active partners in creating (and delivering) the value proposition of the platform.
Hence, any startup building a creativity platform should:

1. Understand the motivations of the creators

2. Create enabling technology that caters to those motivations

3. Have a clear strategy to maximize the number of creators on the platforms.

The following 6 questions can help a platform think through these issues and enable it to successfully
create a platform that finds traction.

ARE YOU PROVIDING TOOLS OR PIPES? OR BOTH?

Creativity platforms may provide content creators the tools to enable them to be creative, or they could
supply the pipes to market their creations to an audience. Or both.

Tools: Platforms may provide creative and/or infrastructural tools. Vimeo gives anyone the ability to
host an HD video online and deliver a video quality superior to all competitors. Instagram enables users
to create beautiful photos without being a PhotoShop expert.

Pipes: In some cases, creativity platforms may provide pipes to a specific desired
audience. Dribbble allows users to upload their creations and provides them access to the right
professional community.
Tools+Pipes: One way to build a lasting platform is to supply both. That is what enabled Instagram, a late
follower, to disrupt Hipstamatic, a far superior product. Hipstamatic allowed users to apply filters to pictures
(initially), but Instagram created a thriving community around such photos. Facebook Photos, in a similar way,
disrupted Flickr to become the largest storehouse of photos on the internet. Facebook provided access to an
audience and their news feed while Flickr only provided hosting.

At the end of the day, if a tree falls in a forest and no one is around to hear it, does it make a sound?

HOW ARE YOU MAKING THE CREATIVE PROCESS EASIER THAN EXISTING
OPTIONS?
There is no dearth of choice on the internet. Competing platforms are a click away. In such a scenario, platforms
that allow easy creation and allow users with lower skills to create high-quality creations often achieve higher
traction. The number of people who tweet is orders of magnitude higher than the number of people who blog.

One of the contributing factors is the fact that Twitter provides pipes in addition to tools. However, the more
important factor is the lower skill and investment required to tweet, as compared to writing a blog post. In a
similar way, Instagram lowered the skills required to create beautiful pictures, a factor that led to its widespread
adoption.

IS THERE A ROBUST CURATION MODEL TO SEPARATE THE BEST FROM THE

REST?
Curation is critical when you’re providing a democratic platform. The platform should have a robust model to
separate the bathroom singers from the Grammy winners. There are typically three broad models of curation.
and a scalable platform usually has a combination of all three:

1. Algorithmic Curation: The internet is fundamentally about automation. The key ingredient of a scalable
model of curation is algorithmic detection of good versus bad, based on certain rules. There is, however, a
potential for false positives with algorithmic curation which might lead to good creations being rejected.
Algorithmic curation should, hence, be scaled carefully and should ‘learn’ and optimize with social and editorial
inputs.
2. Social Curation: You may call it the Digg model but it’s the model, of choice on all platforms today.
The community is provided with tools (voting, rating, flagging, etc.) to provide an input regarding the
quality of the creation and the aggregation of these inputs is used to sort and rank creations and
determine their relevance.

3. Editorial Curation: While tech entrepreneurs would want everything to be automated, manual
curation has a place on every platform, especially in the early days. Editorial curation helps to
understand patterns that can then be automated and scaled. In some cases, editorial curation can even
be used to kick start the platform when there aren’t enough creators on the platform. This is important
because creators power the value proposition for such platforms. The platform has little or no value
without the creations.

IS THERE A CLEAR, DEMOCRATIC, EQUAL-ACCESS PATH TO THE TOP?

Before launching a platform, you should understand the motivations that drive creators to contribute. A
common theme across all platforms is visibility, self-expression, and/or recognition. Since platforms
have a model of curation to separate the signal from the noise and since the creations that make it to
the top of the heap get consumed much more than those that don’t, creators should be given a clear
equal-access path to the top of the heap.

Just as every website publisher invests in SEO to score high on Google’s ‘curation’, creators need an
understanding of what it takes to rank high on a platform. If the mechanism of ranking high and gaining
visibility is unclear, creators may not be interested in participating on the platform. “We feature the most
voted creations on the front page” is a clear proposition as it specifies how the ‘best’ are separated from
the ‘rest.’

IS THERE AN INCENTIVE BEYOND SELF-EXPRESSION?

Art for art’s sake isn’t always good enough. While self-expression and the ability to gain an online
following and build a personal brand are great incentives for creators, having additional incentives can
provide a competitive advantage for the platform vs. competition.
Let’s look at designers and photographers. Different platforms have provided different incentives for these
people:

Threadless: Provides community recognition and curation + The ability to possibly monetize creations if voted
to the top

500px: Provides community recognition and curation + The ability to host an online portfolio

Dribbble: Provides community recognition and curation + Access to highly relevant job offers

DO YOU HAVE A PLAN TO CONVERT CONSUMERS TO CREATORS?

User generated content has come a long way on the internet. At one point, the 90-9-1 rule was often quoted to
explain the low levels of contribution in online communities but, of late, platforms have greater percentages of
contributors, especially owing to the rise of the smartphones which allow a greater number of users to create
anytime and anywhere.

However, the success of a platform still hinges on its ability to maximize the percentage of creators. Hence, in
its initial days, a platform needs to focus on attracting creators. YouTube, for example, did this through a
series of competitions for creators.

But once a critical mass of creators is on board, a second cycle needs to be started. Creators attract
consumers, and it is much more efficient for the platform to convert these consumers into new creators. A
platform, hence, needs to have a clear plan for converting consumers to creators to have a sustainable value
proposition.
THE THREE FORMS OF METRICS
THAT DRIVE SCALE
A thoughtful look at how scale definitions are changing in a networked world.

Sangeet Paul Choudary

Successful businesses are often not distracted by a hundred different metrics, but laser focused on one
metric that is the best predictor of scale. How does a business identify such a metric?

Most businesses are in a relentless pursuit of scale. Most of business education is built around creating
and understanding patterns in business scale. There is, however, a shift in the concept of scale in
business today and not all companies seem to embrace it appropriately.

The traditional definition of scale

Let’s call this Scale v1. The key inputs for business, traditionally, have been land, capital stock, and
labor. The business scales whenever one or more of these parameters scale. You hire more people, you
get more resources and better machinery and the business scales. However, there is a cost associated
with scaling each of these three variables. As a result, the primary focus while scaling is optimization.
Optimization involves creating repeatable processes which can be cost-effectively repeated over and
over again to grow the business. The two key aspects of scaling are:

A) Repeatability

B) Cost-effectiveness

A lot of business education is focused on strategies for optimization of these processes. An IT

outsourcing shop, for example, optimizes processes surrounding the labor variable to create scale. A
manufacturing business has to optimize processes involving all three variables (e.g. procurement,
production, distribution).

The new definition of scale

Let’s call this Scale v2. This is where the internet comes in. The internet, by definition, brings with itself
unlimited scale. Moreover, since networked businesses tend to deal more with bits than atoms, the
inputs to business are no longer the three variables above. The new inputs to business are data and
talent (intellectual capital, both internal and external).

Scale in a networked business is no longer dependent on processes within the business, it’s driven by
network processes. By network processes, I mean interactions that users on the network have with the
product. The business scales by scaling these interactions. Hence, the new counterpart of process
optimization is actually interaction design, and this is partly why designers are so important in networked
businesses because they directly contribute to scale much like MBAs did in traditional businesses.
Not every internet startup is all about Scale v2. e.g. an commerce company like Zappos has Scale v1 on the
supply side and Scale v2 on the demand side. The same applies to any online media company which still has
reporters sitting in a room and churning out stories. On the other hand, marketplaces and platforms like eBay,
Facebook and Airbnb have Scale v2 on both sides.

Google was probably the first business that achieved Scale v2 on both the demand AND the supply side. The
users are on self-serve, and so are the advertisers. This is why it is, even today, one of the most successful
internet businesses ever.

On the contrary, Groupon has Scale V2 on the demand side but Scale V1 on the supply side. It maintains an
ever-growing sales force to manage the merchant side of the business.

Amazon is one of those rare internet companies that did a fabulous job of mastering both Scale v1 (on the
supply side) as well as Scale v2 (on the demand side).

So what metrics should your business use?

Here’s the one line answer:

A business should focus on those metrics which help it create repetitive processes (Scale v1) or
repetitive interactions (Scale v2) that will ultimately build scale.

First, it’s important that a business knows which forms of scale it has on which sides. Second, most metrics
fall into three categories:

1. Per-unit economics of repeatable process OR interaction

2. Time between repeatable process OR interaction

3. % of inputs successfully being leveraged for repeatable process OR interaction

Scale v1: The metrics that determine scale are typically determinants of per-unit efficiency. These could
be one or more of the following:

1. Per-unit efficiency (Cost per input, per-unit production cost, Cost of moving a unit through a channel)

2. Turnaround efficiency (time to source)

3. Utilization efficiency (Inventory turnover)

Scale v2: The metrics that determine scale are typically determinants of engagement and repeat usage.
These could be one or more of the following:

1. Per interaction engagement (Length of visit etc.)

2. Time between interactions (Time between contributions, time between visits)

3. % Interacting (% active users, % of users who produce, etc.)

A business that has Scale v1 on one side (e.g. supply) and Scale v2 on the other side (demand) has to
look at both types of metrics. This is typically the case with an e-commerce company which looks at
improving time to source on the supply side and time between purchases on the demand side. A social
network like Facebook needs users to interact with each other as often as possible and hence focuses
on %interacting, more specifically DAU/MAU.

The governing principle is to understand the processes and interactions which drive scale and focus on
the metric that decides how those processes and interactions can be made repeatable.

What do you feel? Does it help to connect business metrics to scale goals and has it helped while
running your startup?
A SCALING FRAMEWORK FOR NETWORK
EFFECT PLATFORMS

A guide to scaling value and volume on platforms.

Sangeet Paul Choudary

One of the most common misconceptions about running a platform business is that it’s all straight and
easy once the initial chicken and egg problem of getting both producers and consumers is solved.
Ironically, in an age where kicking off new platforms isn’t quite as difficult as it used to be, the most
important platform management issues really come up once the platform starts working and delivering
value.

But let’s start this at the beginning:

The original platform hassle

Everyone loves a good network effect. The problem is often that it is quite difficult achieving it. But
without network effects, there is little or no value on the platform (unless the platform has some
alternate form of standalone value).

As we so often discuss on this blog, the goal of the platform is to enable interactions between producers
and consumers. Think of Dribbble, TaskRabbit or Salesforce, you see this theme repeatedly. As the
platform manager, you want producers to create value and consumers to consume value for the
platform to fulfill its purpose. Additionally, the platform needs some mechanism for curation; that
ensures quality.

Network effects are realized when there are enough producers and consumers with overlapping intent
for interactions to spark off between them. In such a scenario, the platform starts fulfilling its role of
enabling interactions.

From here on, the goal of the platform is to scale its ability to enable more and better interactions.

A scaling manifesto for platforms

If we were to condense it in one line:

A platform’s goal is to scale the quantity and the quality of interactions that it enables.

The first part is obvious but involves several nuances. The second part is less obvious and is often
ignored at the platform’s peril.

Let’s peel this further.

Scaling interactions
To start with, it goes without saying that the platform wants its core interaction to be repeated as often as
possible. More rides on Uber and more tasks booked on TaskRabbit is good news.

The Core Interaction consists of three actions: Creation, Curation and Consumption. Depending on the type of
the platform and the stage of evolution it is in, scaling interactions may involve scaling one or more of these
actions.

Scaling Quantity: Creation

Some platforms scale by focusing solely on scaling creation of new value. A true blue classifieds platform like
Craigslist largely cares about the depth and breadth of its listings.

A platform can stall despite having strong network effects if producers stop creating value on the platform.
Thankfully, this is one of the most obvious goals and metrics to watch out for and platforms rarely fail because
of an inability to focus here.

There are several ways platforms encourage a constant stream of value creation, a framework of which is
discussed in detail in the article here.

Scaling Quantity: Consumption

Scaling creation and consumption often go hand in hand. This is because producers are likely to participate on
a platform only when there is active participation from consumers as well. No one wants to speak to an empty
room. Hence, efforts to scale creation of new value are not going to be sustainable unless complemented by
efforts to scale consumption as well.

For demand-driven platforms like Upwork and 99designs, where the value is created in response to a request,
scaling consumption is a critical first step to scaling interactions. But even for a creation-centric platform like
YouTube, scaling consumption is very important, especially when producers participate with a need to self-
express or self-promote.

Of course, to scale consumption, the platform needs to capture better data about the consumer and use that to
make more relevant recommendations. No amount of marketing and alerts/notifications can substitute
relevance for consumers. This is also largely why we see the role of data scientist rapidly emerging in
importance, not only in the tech industry but also outside it.
A consequence of this is the fact that platforms need to start acquiring data about users the moment they
sign up. This translates itself into simple tweaks in the onboarding flow. e.g. Pinterest will ask you to like a
few boards and a few topics as part of its onboarding flow. It is data that helps it understand what to serve
you next so that you continue to consume. Increasingly, marketplaces aren’t focused on transactions
alone. Additionally, they encourage users to follow topics and listings, in the hope that such data on users’
interests can help create opportunities for transactions in the future without the user explicitly initiating
one.

Scaling Quality: Curation

Finally, as more value gets created and consumed on the platform, the platform also needs to get better
at its ability to curate and differentiate high-quality from low-quality. To ensure value and a desirable
experience to its users, the platform needs to ensure that it encourages actions that result in high-quality
creations and discourages actions that result in low-quality contributions.

In its early days, YouTube ran competitions where the most upvoted videos were rewarded. Wikipedia
blocks IPs and accounts that generate a lot of suspicious activity. Sittercity and TaskRabbit do intensive
background checks on service providers on their platform. Whether social feedback or editorial pre-
screening, curation is a necessary part of running a platform.

Curation is typically done in one of three forms:

1. Editorial: An editor, admin or community manager approves and disapproves contributions to the
platform.
2. Algorithmic: Algorithms take decisions on what’s desirable and what’s not based on certain parameters.
3. Social: The community curates through signals about quality, like rating, voting, etc.

Scaling curation may mean different things depending on which of these aspects are being scaled.

Scaling editorial curation: The brute force method to scale editorial curation, traditionally, has been to
get more editors on board. That never works well for a network effects platform. Editorial actions scale
only when they are gradually moved out to the community over time. The editors do not become
redundant; they simply take on more abstracted roles. This may involve educating the community on how
to curate and ensuring that the tools of curation (e.g. rating, review, reporting, etc.) are being used
correctly and often enough. In the case of platforms like Viki and Wikipedia, it may also mean creating a
hierarchy of sorts in the community to differentiate highly reputed users from less reputed ones to phase
out actions gradually, from internal editors to highly reputed users, and so on.
Scaling algorithmic curation: Very briefly, algorithmic curation may scale by improving the algorithms
themselves or improving the inputs to the algorithms. The inputs to the algorithms are provided by editors or by
the community and hence scaling algorithmic curation works very closely with scaling the other two forms of
curation.

Scaling social curation: Social curation scales either by permeating a reputation model through the
community or by merely relying on the strength of numbers. In the former, the opinions of experts are given
more weight than that of novices. In the latter, all opinions count towards the same. This is, of course, a
continuum rather than a duality and most platforms lie somewhere along the continuum. The more the
expertise required to make a judgment on curation, the more likely is it to rely on a reputation model. Users
who have curated well in the past have a greater say in future curation.

Scaling Quality – Overall Governance

Finally, this specifically involves looking for corner cases and situations where the platform is used/abused in
ways that the platform creator hadn’t planned for. Specifically, this involves identifying undesirable interactions
and ensuring that those are not repeated. A murderer using a dating site to find his next victim is an
undesirable interaction as is a contributor defacing the Wikipedia profile of a public figure.

In some cases, scaling quality through governance may have limits. No matter how much you govern, you may
still have the odd traveler trashing an Airbnb apartment. In such cases, creating centralized trust mechanisms
becomes critical to scaling the platform and ensuring widespread adoption among mainstream users. This
brings us to our final point on scaling platform activity.

Scaling Quality – Mitigating Interaction Risk

The final element of scaling quality of interactions has to do with the risk inherent to either side in participating
in an interaction. In an age where people actually die when meeting people through Craigslist, mitigating risk
becomes an important aspect of driving platform adoption. Not all interactions are created equal. Some are
riskier than others. Participating on Twitter doesn’t involve many risk for either side but participating on a
platform for discovering home cooked food may have higher risks associated. Depending on the degree of risk
involved, the platform may have to invest heavily in offering centralized guarantees and insurance. Most
sharing economy platforms invest in creating insurance and trust mechanisms to ensure that users aren’t
discouraged from participating.
 PLATFORM METRICS: THE CORE
METRIC FOR PLATFORMS, NETWORKS
AND MARKETPLACES
The one metric that tells you how well your platform is performing.

Sangeet Paul Choudary

You become what you measure. From my experience working with clients across enterprises and
startups, the most common reason for failure and inefficiency is the focus on convenient, but
inappropriate, metrics. Your technology doesn’t determine the business you build. Neither does your
organizational capability. The metric you optimize for is the single biggest factor that determines which
business you end up building.

METRIC DESIGN
The importance of choosing the right metric is more far-reaching than we often believe. A metric is a bit
like a commander’s intent in an army. At battle, there are a lot of variabilities and unexpected
contingencies which cannot be pre-planned for. The Commander’s Intent is a simple rule of thumb that
helps soldiers take local, individual decisions towards a cohesive, larger goal.

Metrics work in much the same manner. Once you set a metric, the entire team organizes its efforts
around it, and works relentlessly to optimize the business for that metric. It’s often fancy to have a large
dashboard with multiple graphs tracking hundreds of things. But to be truly effective, an
organization/team/individual should be solely focused on optimizing for one metric.

As a result, identifying and designing the right metric is critical for business success. More often than
not, I’ve seen the following general observation to hold true:

If you’re asking someone to optimize for more than one metric, you’re setting them up for failure.

Often, ratios help capture multiple movements in one metric. Whether you think of the financial ratios
that traditional business managers track or the DAU/MAU that app developers relentlessly track today,
ratios tend to be important as they explain concentration rather than quantity.

PIPE METRICS
This discussion of metrics is especially important in the world of platforms and networked businesses.
Platform businesses are a lot more complicated than traditional pipe businesses.Pipes optimize
unidirectional flow of value. Hence, metaphorically, releasing bottlenecks at any point should help with
the flow. The Core Metrics for pipes, naturally, then, measure smoothness of flow and/or removal of
bottlenecks. Inventory turnover is one such metric to check how often the flow of goods/services moves
through the pipe. All forms of Output/Input ratios for intermediary teams on the Pipe are, again, checks
to understand the rate of flow and identify creation of bottlenecks.
PLATFORM METRICS
But this tends to be much more complex in the case of platforms where flows are multi-directional. Moreover,
they are interdependent because of network effects. e.g. optimizing activity on the producer side may have
unexpected implications on the consumer side. On a dating network, allowing over-access to men may be
unattractive for women. Hence, even if you have two different teams optimizing for two different metrics on the
producer and consumer side, the activities of one team may adversely impact the pursuits of the other team.

How then does one go about deciding on platform metrics?

THE BUSINESS OF ENABLING INTERACTIONS

This takes us back to a theme I repeatedly talk about. If I had to condense the essence of Platform Thinking in
one line, here’s what it would be:

We are in the business of enabling interactions.

This is much like the Commander’s Intent I mentioned earlier and has important implications. Irrespective of
how big your firm is, how complex the operations are, the goal should always be to optimize the core
interaction.

1. Identify the Core Interaction that your platform enables

2. Remove all bottlenecks in the Core Interaction to ensure that it gets completed across Creation, Curation,
and Consumption

3. Ensure that the Core Interaction is repeatable and repeats often

From a metrics perspective, this essentially means that the Core Metric that rules everything should measure
interactions.

If you’re running a platform business, you need to start measuring and optimizing your core interaction.
METRICS DESIGN AROUND THE CORE INTERACTION
So we get the fact that we need to measure interactions. However, we still need a measure, a number
that shows the Core Interaction is working well. As with all metric design, it is still possible to choose the
wrong metric despite understanding the importance of measuring the Core Interaction.

To design the right metric, let’s revisit what the Core Interaction on a platform actually entails.

From earlier essays in this series, we note the following:

1. A platform enables an exchange of information, goods/services, money, attention, etc. between the
producer and consumer. For a visual guide to how this works, check the article here.

2. The exchange of information always occurs on the platform. The other exchanges may or may not
occur on it. The exchange of information enables every other exchange to take place. To understand
the mechanics of this, refer this article.

3. The exchange of information is the key source of value creation across all platforms and can be
visualized as the Core Interaction of the platform. To understand the structure of the Core Interaction in
detail, check the article here.

4. The Core Interaction has three parts: Creation, Curation, and Consumption of the Core Value Unit
Let’s now look at the different types of platforms and tease out relevant key metrics.

TRANSACTION CAPTURE
Some platforms capture the transaction between producers and consumers. These platforms typically
track actual transactions. Platforms like the Amazon marketplace may measure gross value of
transactions. Those like Fiverr (which have fixed value per transaction) may simply measure number of
transactions. Airbnb tracks number of nights booked. This is a better indicator of value creation than
simply tracking number of transactions. At the same time, it doesn’t care about value of transactions
(spare mattress being booked vs. castle) as the goal is simply more value created irrespective of type of
customer.
TRANSACTION TRACKING
Some platforms can track the exchange of goods and services in addition to capturing the exchange of money.
Upwork, for example, can track the number of hours of work delivered by the freelancer (producer), a key
measure of value creation. Clarity.fm can track the duration of the consulting call between an expert and the
information seeker.

MARKET ACCESS
Some platforms are unable to capture the transaction and the exchange of money. They create value by
allowing producers access to consumers. In these cases, one of the common metrics tracked is the platform’s
ability to generate leads. OpenTable specifically tracks reservations. These are not the actual transactions at
the dinner table but serve as a proxy for the value created. Some platforms may track overall/relevant market
access. Dating and Matrimonial sites often talk about the number of women registered as that determines the
value that a user can expect to get.

CO-CREATION
One of the key properties of platforms is the fact that external producers can add value- Whether it is new
apps on an app store, new videos on YouTube or new pictures on Flickr. In these cases, one is tempted to
solely track these co-created Value Units. However, Creation forms only one-third of the Core Interaction. The
proof of the pudding, in such cases, lies in repeat Consumption. Some platforms may track the total
consumption, some may track the percentage of Value Units that cross a minimum threshold of Consumption.
I tend to favor the latter as measuring and increasing the percentage of units that get minimum consumption
ensures that the platform focuses on getting more producers who create relevant units that will be consumed.
It also ensures that, over time, the feedback loops (in the forms of notifications to producers) will encourage
the creation of the kinds of units that get greater consumption.

QUALITY AS VALUE
Some platforms may create value largely by signaling quality. Reddit is one such example where Curation is
more important than Creation or Consumption. Such platforms may track the reputation of users and create
feedback loops that encourage users to participate often, gain karma and use that to participate further in the
curation process.
MARKET ATTENTION
Platforms where the Core Value Units are content e.g. YouTube, Medium, Quora, etc., the engagement
of Consumer Attention serves as a key metric. Measuring number of videos or articles uploaded or
number of videos viewed or articles read is often not enough. These give indications of Creation and
Consumption but not of Curation. We need some indicator of quality as well. This is why many such
platforms track the percentage of content which gets a minimum engagement. Medium tracks views and
reads separately, indicating that it requires a minimum commitment from the Consumer to determine
the quality of the content.

THE EASY METRIC FALLACY

While working with companies on this, I’ve often noted the following:

1. Creation is the most common metric tracked. Number of apps, number of videos, number of sellers
etc. This is misleading.

2. In the case of Market Attention category platforms, Consumption is the most common metric tracked.
This is an improvement but still not a measure of quality.

3. Curation is rarely tracked and is often the most important metric that determines the health of the
platform.

4. The measure of transactions that should be tracked varies with type of platform. In some cases,
number of transactions may suffice, in other cases, volume of transactions may matter.

5. The metric that best explains interactions will change over the life cycle of the platform, and it’s
critical to identify points at which these transitions occur. Companies often make the mistake of sticking
on with an older metric when their business has scaled. Identifying and vetting the Core Metric at every
point is very important.
COUNTER METRICS
While measuring the platform’s ability to create interactions is important, it is equally important to measure its
failure to close interactions. I will explore this further in a subsequent post.

THE WAY FORWARD

The discussion on metrics is deep and cannot be done justice in one post. I’ll cover this more as we move
further in the series. The key point, though, remains:

On a platform, the Core Metric that rules them all must measure and optimize the Core Interaction.

TWEETABLE TAKEAWAYS
For platforms, the Core Metric to be tracked must measure and optimize the Core Interaction.
The goal of a platform is to repeat and optimize the Core Interaction that creates value.
MARKETPLACE METRICS: THE
THREE SUCCESS FACTORS
The three key metrics to track marketplaces success.

Sangeet Paul Choudary

Marketplaces are difficult businesses to run. Like all multi-sided platform businesses, they suffer
from the classic chicken and egg problem: the technology has no value unless buyers and sellers are
present and you can’t get the buyers on board unless you have sellers and you can’t bring in sellers
without having buyers. Hence, building a marketplace is a lot like building two separate companies
simultaneously, each dependent on the other.

There are three factors that determine success for a marketplace business

Liquidity or critical mass

The lifeline of a marketplace (and any platform business for that matter) is liquidity. Liquidity is a state
where there are a minimum number of producers and consumers on the marketplace and there is a
high expectation of transactions taking place. This is similar to the critical mass of users that is needed
on a social network for users to find value in the network. Critical mass is a state where there is enough
volume of supply and demand, for transactions to start sparking.

The first and most important metric to watch out for is the percentage of listings that lead to transactions
within a certain time period. This serves as a proxy for the efficiency of the marketplace. Merely
increasing the number of buyer and seller sign-ups doesn’t serve a purpose unless this metric starts
rising. The time period would depend on the category. AirBnB listings would find transactions sooner
than listings on a buy-and-sell real estate marketplace. This could also depend on ticket sizes within
the same category. Fiverr and oDesk are both services marketplaces but the turnover on Fiverr is most
likely higher, owing to the much smaller ticket sizes.

To get to liquidity, the marketplace also needs to solve the chicken and egg problem and get both
buyers and sellers on board. Marketplaces leverage a variety of tactics for circumventing this problem
including building single user utility, stealing traction and piggybacking other platforms.
Matching: Curation of products/service
Users visit a marketplace with a highly transactional intent and want to find what they’re looking for at the
earliest. In this aspect, transaction businesses are remarkably different from engagement businesses. A user
visiting AirBnB or Yelp has a specific intent in mind. Hence, the quality of the search algorithm and the
intuitiveness of the navigation are critical to delivering value. In contrast, a user visiting Pinterest often wants to
spend some time and consume content on the platform. Hence, the infinite scroll!

The efficiency of discovery and matching is critical to a marketplace’s success. Percentage of searches that
lead to listing profile visits within the first page of results is one such metric. When listings are served instead,
as a feed, the clickthrough per session can serve as a proxy as well. The best metric to track matching
efficiency varies with the business model of the marketplace as well as the category.

Trust: Curation of participants

Building trust is central to marketplaces where transactions carry risk. AirBnB is an example of a player in a
high-risk category, that succeeded because of its ability to curate its participants. AirBnB allows hosts and
travelers to review each other and has one of the highest review rates among marketplaces. It also takes
additional measures to build trust, including having photographers certify a host’s listing.

This was one of the factors that helped AirBnB challenge CraigsListbecause CraigsList never built a strong
curation system for participants.

Focus on the trust metric is very important to move from appealing to an early adopter audience to appealing to
a mainstream audience. While early adopters use new marketplaces because of the novelty, opening up to a
larger market requires the trust and reputation management systems to be alive and kicking.
What’s not as important:

User interface and design are less important with transactional businesses as compared to engagement
businesses.

On a marketplace, the ability to search and transact/interact should be as intuitive as possible. Beyond
that, the look-and-feel and design are purely hygiene factors. Unlike social networks, marketplaces are
transactional and users typically don’t have long visit lengths engaging with the product. Hence, UI is
not as important a criterion as the other three mentioned above.

In summary, if you’re building a marketplace:

1. Focus on liquidity, not just user growth

2. At critical mass and beyond, closely track matching efficiency

3. When moving from an early adopter to a mainstream audience, ensure that the trust systems are in
place and functioning well.
ENGAGE FURTHER

C-level Executive Education Platform Architecture and Strategy

C-level and business leadership-level exec ed towards a platform Engagement on a specific platform strategy and implementation. Includes:
implementation at a client organization. It may also include workshops for platform business design, layout of feedback loops and network effects,
execution and implementation teams. For larger teams, this may be done as monetization scenarios, management of curation and governance of the
webinars remotely. ecosystem, data strategy, roadmap creation and metrics definition, among
other things. This may be done remotely or in-person or through a
combination.

Commissioned Research Retained Advisory

In-depth research, commissioned by the client, to create thought leadership Retained advisory relationships with a specific project (or multiple projects)
material, layout future industry scenarios or study business model at a company, or advisory boards, typically structured as 6-12 month
transformation. retainers.

To engage further, please write in at the

Corporate Speaking following:
Keynote speaking at sales events, executive briefings for C-level execs, [email protected]
and speaking and briefings at executive planning sessions and offsites. [email protected]