Entrepreneurship

The Overlooked Key to a Successful

Scale-Up
by Jeffrey F. Rayport, Davide Sola, and Martin Kupp

From the Magazine (January–February 2023)

Sean Lemoine

Summary. Many start-ups experience enormous popularity and runaway growth,

but only a few go on to become stable giants. What separates them from the pack?
They all go through a developmental stage called extrapolation, say three business
school professors. This... more

Consider the tales of three start-ups that seemed poised for

success.
 00:00 / 21:53
Listen to this article

To hear more, download the Noa app

In 2012, King Digital Entertainment had established itself as a

developer of popular free games on smartphones. Its user base
was growing exponentially, driven by the hit game Candy Crush
Saga. From mid-2012 to mid-2013 the company experienced a 12-
fold increase in revenue but only a sixfold increase in costs. The
result was a nearly 70-fold increase in operating income, from
€10.5 million to €716 million.

SoundCloud was an online audio-sharing platform and a rival of

Spotify and Apple Music. From 2012 to 2013 its user base grew 15-
fold, from 10 million to 150 million registered users. However, its
revenues increased less than 50%, from €8 million to €11 million,
while its operating costs grew 75%, from €16.5 million to €28.5
million.

In 2017, WeWork, a celebrated coworking venture, had raised $10

billion in equity and debt. Its top-line revenues had doubled for
five consecutive years, and its membership had grown 10-fold.
But over the same period, operating costs rose from $400 million
to almost $2 billion, leading to significant and deepening losses.
Sean Lemoine documented the joy of setting off high-powered rockets at LDRS (Large Dangerous Rocket Ships), an
event for rocket enthusiasts held in the Mojave Desert.

Of these three high-flying start-ups, only King Digital

Entertainment became a stable, highly profitable business. What
explains their diverging fortunes?

Drawing on an examination of dozens of rapidly growing ventures

and our experience teaching courses on scaling up enterprises at
our respective business schools, we’ve concluded that what made
the difference was that King Digital Entertainment engaged in a
developmental stage we call extrapolation, in which a company
explores profitable growth options while exploiting economies of
scale and scope. This stage isn’t part of traditional organization
theory, which says that businesses are in either exploration mode
or exploitation mode.

Exploration involves the search for product-market fit. The

company’s hypothesis about how it will deliver value is tested to
determine whether customers have a problem to be solved or a
pain point to be addressed—and are willing to pay for the
company’s solution.
Exploitation begins when the fast revenue and profit growth
enjoyed in the start-up stage slows and reverts to market norms.
In this phase the company aims to strengthen its competitive
advantage by fine-tuning the business model and strives to
achieve incremental long-term growth and stable profits.

These two stages are well-known—start-ups often begin with a

bang, and a few seem to emerge as stable giants. But in our view
extrapolation is the often-overlooked but critical phase between
exploring many opportunities and exploiting one.

During this stage start-ups pursue two goals. The first is to

confirm the extent to which product-market fit shows that there is
demand for the company’s offering. The second is to achieve what
we call profit-market fit—to demonstrate not only that the
venture can ramp up revenue rapidly but that every new
customer brings in additional revenue and incurs only marginal
cost—the key to profitable growth.

The company must construct a

business model that boosts revenue
while reducing variable unit costs and
containing fixed costs.

King Digital Entertainment, SoundCloud, and WeWork all proved

the value of their offerings by achieving impressive growth in
numbers of customers and theoretically were positioned for
market dominance. But King alone was able to turn its top-line
growth into comparable profit growth during the extrapolation
phase. Each new smartphone user who downloaded the Candy
Crush Saga app brought in revenue that went almost directly to
the bottom line. SoundCloud, in contrast, never managed to
develop a scalable and profitable way to monetize the enormous
consumer audience it had built. WeWork’s problem, aside from
the well-known controversies surrounding its ill-fated initial
public offering, was failing to establish an increasingly profitable
business model to support its global network of coworking spaces.

What is the key to successful extrapolation? It demands new ways

of thinking about strategy, operations, financing, and speed. It
also requires approaches to organizational structure, culture, and
talent that are distinct from those of the other two phases. Start-
up and enterprise leaders alike must consciously treat
extrapolation as a specific stage in the development of any new
venture or new-to-market offer.

Principles of Extrapolation
In the companies we studied the extrapolation phase spanned as
little as a year and as much as three years. Eventually, competitive
responses, market saturation, or shifts in the external business
environment brought this phase of dramatic growth in revenue
and operating income to an end.

Our research shows that ventures that succeed at extrapolation

have three characteristics:

1. They understand and leverage the conditions that are

critical for success.

2. They follow a rigorous extrapolation process.

3. They have ambidextrous organizations that can manage

strategic experimentation and disciplined execution
simultaneously.

Let’s look at each of these in turn.

Critical Conditions
Extrapolation requires two types of conditions to be in effect:
necessary conditions, which don’t by themselves create
extrapolation but must be present for it to occur, and sufficient
conditions, which can produce it.

There are two necessary conditions:

A robust market. Extrapolation requires a large number of
customers who have similar needs and will pay for a product that
meets them. (It may seem obvious that the market must be big
enough to support scale, but we are often struck by how many
early-stage start-up teams miss this point or misconstrue the
relevant data.)

Solution repeatability and distinctiveness. The product the

company offers must be the same for each customer but
differentiated from competitors’. Homogeneity simplifies the
business model and makes it easier to scale up.

In addition, we have identified five sufficient conditions. Not all

are essential to success in every case, but several are always
present when extrapolation works.

An effective go-to-market strategy. The venture must have a

clear plan to reach users through direct or indirect channels, turn
them into loyal customers, and persuade them to promote the
product. Consider the pre-owned-apparel business ThredUp,
which appears to be in a successful extrapolation phase. To get
there leadership had to scale up both sides of its platform,
recruiting enough sellers of used clothing to attract buyers. By
focusing on achieving high engagement and satisfaction among
both buyers and sellers, it activated powerful word of mouth,
which propelled the growth it needed to pull off its recent IPO and
generate a 39% annual increase in profits.

A proven monetization approach. The offer ultimately must

generate revenue directly or indirectly through payments or
advertising. Without healthy revenue sources, scale is hard to
justify or support. That was one of the problems that dogged
SoundCloud: It derived limited revenues from listeners and only
modest revenues from the musicians whose songs it hosted. In
comparison, Spotify, while still challenged in its profit model,
derives huge revenues from a mass audience of paying listeners.
Network and density effects. While network effects kick in when
a platform or a product attracts enough users to make it more
valuable to other users, density effects happen when the
concentration of users in one geography or market segment
intensifies substantially, leading to virality or word of mouth, or
the average number of new-user referrals each existing user in the
network makes is greater than one (a concept known as the viral
coefficient). Effective extrapolation usually (but not always)
requires strong network and density effects that enable
economies of scale and limit defections to other offers.

Increasing returns. The company must construct a business

model that boosts revenue while reducing variable unit costs and
containing fixed costs. Tech platforms (like Facebook, Nextdoor,
and Slack) are famous for achieving extreme economies of scale,
given that each incremental unit of service delivered often incurs
zero variable cost. Such platforms also grow fixed costs more
slowly than top-line revenue—often by a factor of three to five—
an obvious formula for success.

Substantial capital resources. Without question, there are

founders and ventures that can bootstrap their way to scale. The
direct-to-consumer home-goods retailer Resident was able to
largely self-fund its early growth because it sold big-ticket, high-
margin products (mattresses) from day one. But in reality, most
successful ventures must raise significant outside capital in order
to achieve the rapid growth in scale seen in extrapolation.

A Rigorous Process
Successful extrapolation requires a focused, systematic approach
to identifying and removing internal business-model constraints
on growth. The theory of constraints process, first codified by
Eliyahu Goldratt in his classic book, The Goal, suggests that
companies can do so by following these five steps:

1. Articulate the growth goals (for example, “achieve a fivefold

increase in revenues and a 10-fold increase in operating
margin”) and examine whether the necessary and sufficient
 conditions for achieving them are present. The goals
selected will be influenced in part by market and business-
model realities and in part by the founding team’s level of
ambition.

2. Define the critical assumptions underpinning your

business model. These should be based on an assessment of
what factors must be in place to produce growth. In other
words, ask “What needs to be true?” for you to realize your
growth goals. For example, to increase revenue 10-fold, you
might need five times as many customers who make
purchases three times as frequently as your customers
currently do.

3. Identify the business model constraints—the barriers to

achieving your growth goals—and the right sequence in
which to tackle them. For example, managers may discover
that the cost of a certain input is an enormous constraint or
that the market is just too small. The theory of constraints
tells us that no chain can be stronger than its weakest link,
and that thinking applies to the business model. Every
model will have one or more constraints that limit its
output.

4. Develop a way to remediate the most significant constraint.

You may need to either examine how other companies have
adapted their models to address the same type of constraint
or develop an innovative new business model to find a
work-around. The key is to apply what we call “strategic
experimentation,” in which you validate adaptations or
innovations on a small scale before applying them to the
whole business.
5. Once the first constraint is no longer a barrier to growth,
select the next one and remediate it. Continue this iterative
process until all significant constraints have been
addressed.

To understand this process in action, we studied how Niraj Shah

and Steve Conine, the founders of the home-goods retailer
Wayfair, approached extrapolation. Initially known as CSN Stores,
the venture began as a collection of niche e-commerce sites
focused on narrow product categories, each accessed through a
generic web address in the form of “product.com” (such as
RacksAndStands.com or EveryGrandfatherClock.com). As
revenues grew, CSN wound up with more than 200 such sites and
encountered a first significant constraint: Few customers who
bought from one site had any idea that the same company
operated other similar sites. That meant CSN lost sales from
satisfied customers who otherwise would have come back to buy
again. Here the constraint was the dispersion of channels. There
were too many channels and no network effects.

The retailer’s first remediation effort focused on increasing repeat

purchases. The founders rebranded the business as Wayfair and
then consolidated all the “product.com” sites under the new
name, combining millions of SKUs on one platform. Once that
integration was complete, cross-selling and repeat purchase rates
took off.
Sean Lemoine

The second remediation effort focused on increasing customers’

lifetime value by improving fulfillment. The business was drop-
shipping more than 85% of orders, meaning they were sent
directly from suppliers, which made it difficult to ensure that they
were filled in an accurate and timely fashion. Customers who had
bad experiences with orders were unlikely to return. To address
that problem, Wayfair established a network of distribution
centers to handle its fastest-moving SKUs. That way it could
“forward position” inventory for vendors (without taking title to
merchandise, sustaining its asset-light business model), enabling
faster fulfillment and more-consistent service to consumers. That
helped cement loyalty to the Wayfair brand.

The third remediation effort focused on lack of product

differentiation and defensibility. Management realized that the
business was constrained by its generic and commoditized
offerings. As a result it had to compete on price. The goal was to
start selling products that were exclusive to the site and more
distinctive in the eyes of consumers. That was achieved by
implementing several branding initiatives. Wayfair worked with
suppliers to develop private label lines, capturing additional
points of margin with an array of proprietary offerings. Shoppers
could no longer directly compare Wayfair’s prices against
competitors’. In addition, Wayfair created “lifestyle” brands,
which presented otherwise unrelated SKUs (say, a sofa and a
dining room table) in highly styled groupings. Because many of
these were also private label, they boosted gross margins.
Meanwhile the groupings increased average order value by
encouraging shoppers to buy combinations of items. All those
changes led to higher lifetime values.

An Ambidextrous Organization
In their 2004 Harvard Business Review article “The Ambidextrous
Organization,” Charles O’Reilly and Michael Tushman described
companies that could simultaneously explore new businesses
while exploiting their existing core businesses. The ability to do
that is crucial to extrapolation success.

Almost all growing ventures, after they move beyond the early
start-up stage, routinely need to reinvent themselves and refine
their core business. But such flexibility is especially important
during extrapolation. “When we were growing the user base, we
had to adjust and change our monetization mechanism several
times,” King Digital Entertainment’s cofounder Riccardo Zacconi
told us. “Initially it was geared toward advertising, but then it
became almost entirely reliant on selling virtual goods.” In fact,
the adaptation process at King went further than a pivot in
monetization. Relentless experimentation brought about changes
in the revenue model, management processes, staffing levels, and
the organization of the company’s studios and teams.

King managed to scale up its infrastructure while reducing unit

costs. It did that by rigorously analyzing the drivers of its
customer acquisition cost and then lowering it through iterative
in-market testing. That business model transformation was
carried out at the same time that strong product-market fit
unlocked rapid revenue growth. During the extrapolation phase,
King’s leadership was able to simultaneously explore and exploit.
Rather than canceling each other out, those seemingly opposed
approaches came together to make the organization stronger.
 See more HBR charts in Data & Visuals 

In our research we have found that the following elements are

crucial for rapidly growing start-ups and corporate ventures that
aim to achieve ambidexterity:
Modular organizations and autonomous teams. Ventures that
successfully navigate the high-stakes transition from start-up to
scaled-up keep their working units small. King, for example,
doesn’t have one creative studio, as many entertainment
companies do; it operates five studios, in parallel, that are part of
creative clusters around the world. Voi Technology, a European
app-based electric-scooter company, is organized into units for
different metro markets. Such modular approaches allow
ventures to expand without losing their agility. During
extrapolation, businesses need to replicate the success of proven
business models while maintaining the flexibility to invent new
ones. Without modular forms of organization, pivots become
challenging or nearly impossible to pull off.

Extrapolation is also most effective at companies where authority

is distributed to teams rather than held within a tight
management hierarchy. For example, at King’s successful
competitor Supercell, Ilkka Paananen aspires to be the “least
powerful CEO in the world,” which means that his teams can
make all key decisions about game franchises without consulting
him or others in top management.

Swift reallocation of talent. When a business enters the

extrapolation phase, management must begin assigning its most
valuable human capital to its highest-potential opportunities.
Both King and Supercell move game developers off unpromising
or maturing game franchises with ruthless speed and discipline.
Resident, the bed-in-a-box mattress company, laid the
groundwork for rapid talent reallocation by launching itself as a
virtual organization long before the pandemic. Because most of
its staff works remotely, Resident can tap the best employees from
anywhere around the world and continually reassign them to
projects with the highest prospects.

Cultural management. Because workforces tend to expand

dramatically during extrapolation, culture is an essential tool for
maintaining a firm’s focus, mission, and direction. King invests in
large-scale “infomarket” events, where it brings all its talent
together to reinforce cultural norms and “create energy.” At
Zoom, the teleconference software company, the founder and
CEO, Eric Yuan, champions what he has called a “culture of
happiness,” predicated on the idea that a happy organization is
more likely to develop products that will delight users.

Expansion of opportunities. Many of the successful ventures we

examined didn’t regard their initial market opportunity as fixed.
Rather, they pursued two goals simultaneously: first, increasing
the total potential market, and second, unlocking higher-quality
revenue growth (that is, revenues with higher margins). Chegg,
for instance, started as an online marketplace where college and
university students could rent or buy used textbooks. Under new
leadership, it reframed its mission of serving students with a
holistic offer—a web portal that the company originally called the
Student Hub—which was designed to address an array of
students’ needs (with course schedules, study guides, information
on internships, and more). That change dramatically expanded
the potential market while also shifting the model from a seasonal
sales and rental business with narrow gross margins to a platform
business with robust gross margins, higher and more-predictable
revenue, and reduced seasonality. Chegg’s story demonstrates
how extrapolation entails the creation of a better, more
sustainable business model too.

The embrace of inorganic growth. While early-stage start-ups are

relentlessly focused on refining their offers in search of product-
market fit, ventures in the extrapolation phase often consider
acquisitions as a way to expand geographic footprints (and thus
their markets), address talent gaps, add features or functionality,
or augment reach in terms of audience, users, or customers.
Chegg went on an acquisition spree (buying six companies in 15
months) in order to add features to its Student Hub portal.
Though not all those purchases were successful, together they
brought in critical capabilities during a period when time was of
the essence.

...

Our research demonstrates the wisdom of approaching

extrapolation as a distinct phase of business development with its
own principles, processes, and management requirements. More
than half the ventures we studied did so, and they appeared to
have a higher likelihood of success.

Managing the dynamics of rapid growth is an enormous challenge

for young businesses—start-ups and initiatives within enterprises
alike. Once executives understand the fundamental differences
between exploration, extrapolation, and exploitation—especially
the requirement of not only product-market fit but also profit-
market fit—they can confidently navigate the difficult transitions
between the phases, make the necessary changes in focus, and
adopt the right approach at the right time.
A version of this article appeared in the January–February 2023 issue of Harvard
Business Review.

JR
Jeffrey F. Rayport is a senior lecturer in the
Entrepreneurial Management Unit at Harvard
Business School.

DS
Davide Sola is a professor of entrepreneurship
and strategy at ESCP Business School.

MK
Martin Kupp is an associate professor of
entrepreneurship and strategy at ESCP
Business School.
Recommended For You

Why Multibusiness Strategies Fail and How to Make Them Succeed

PODCAST
How to Scale a Start-Up

The 3 Things That Keep Companies Growing

More of Us Are Working in Big Bureaucratic Organizations than Ever Before