How Do You "Design" A Business Ecosystem?: by Ulrich Pidun, Martin Reeves, and Maximilian Schüssler
How Do You "Design" A Business Ecosystem?: by Ulrich Pidun, Martin Reeves, and Maximilian Schüssler
BUSINESS ECOSYSTEM?
By Ulrich Pidun, Martin Reeves, and Maximilian Schüssler
This article is the second in a series of publica- ecosystems exhibit strategic challenges not
tions offering practical guidance on business found in other governance models, such as
ecosystems. The first article addressed the how to solve the chicken-or-egg problem of
question “Do you need a business ecosystem?”, creating a critical mass of partners and cus-
this article deals with ecosystem design, and tomers during launch and how to build a
subsequent articles will address how to man- scalable and defendable model.
age a business ecosystem and how to measure
its success over time. Moreover, business ecosystems, similar to
residential districts, cannot be entirely
If designing a traditional business model is planned and designed—they also emerge.
like planning and building a house, design- This adaptability is actually one of their
ing an ecosystem is more like developing a major strengths. However, there are some
whole residential district: more complex, key design choices you need to get right in
more players to coordinate, more layers order to increase the odds of success. These
of interaction and unintended emergent design choices are not independent; they
outcomes. must be consistent with one another and
offer a coherent overall configuration.
What makes ecosystem design distinctive is
that it requires a true system perspective. It Based on an analysis of more than 100 suc-
is not sufficient to design the value cre- cessful and failed ecosystems across sectors
ation and delivery model; the design must and geographic markets, we find that the
also explicitly consider value distribution ecosystem design challenge can be ad-
among ecosystem members. This is further dressed by working through six sequential
complicated by the limited hierarchical questions (see Exhibit 1):
control in an ecosystem and the need to
convince partners to participate, which •• What is the problem that you want to
poses specific governance challenges. And solve?
Exhibit 1 | The Six-Step Journey of Business Ecosystem Design
1 What is the problem that 2 Who needs to be part 3 What should be the initial
governance model of your
you want to solve? of your ecosystem?
ecosystem?
• Is the problem big enough? • What are the players and roles? • How open should the ecosystem be?
• Is an ecosystem the right choice? • Who should be the orchestrator? • What should the orchestrator control?
• What kind of ecosystem do you need? • How can the orchestrator motivate the
other players?
4 How can you capture 5 How can you solve the 6 How can you ensure the
the value of your ecosystem? chicken-or-egg problem? evolvability of your ecosystem?
• What should you charge? • What does it take to achieve critical mass? • How can you scale the ecosystem?
• Whom should you charge? • What is the minimum viable ecosystem? • How can you defend the ecosystem?
• Which side of the market should you • How can you expand the ecosystem?
focus on? • How can you protect against backlash?
•• Who needs to be part of your The value potential from addressing an un-
ecosystem? met need is difficult to predict but poten-
tially very rewarding, because there is ini-
•• What should be the initial governance tially no offering to compete with. Who
model of your ecosystem? would have guessed 20 years ago that post-
ing selfies, photos of your food, and cat vid-
•• How can you capture the value of your eos is such a deep human need that multi-
ecosystem? billion-dollar businesses like Instagram and
YouTube could be built on them? By con-
•• How can you solve the chicken-or-egg trast, removing an existing friction is more
problem during launch? predictable. The ecosystem value proposi-
tion is a function of the size of the friction,
•• How can you ensure evolvability the share of the friction that can be elimi-
and the long-term viability of your nated by the ecosystem solution, and the
ecosystem? willingness of customers to pay for it.
Design
Battery manufacturer Car manufacturer
Car
Software operating
Battery system
Battery + charging
(rental fee)
Better Place Consumer
Battery
Power charging
load balancing
Battery charging/
Utilities
exchange infrastructure
Power
On the other hand, open ecosystems are Good ecosystem strategy may be to identi-
difficult to control and are thus best suited fy and occupy potential innovation or ca-
for products and services with limited pacity bottlenecks that can become an im-
downside and relatively low cost of failure. portant source of value. Successful
In case of high failure costs, and a corre- orchestrators claim important system con-
sponding need to limit the downside, a trol points that allow them to capture their
closed ecosystem may be the better solu- fair share of value. For example, Nest de-
tion. It allows for a more deliberate design cided to engage in alarm and monitoring
of the ecosystem and for closer control of itself because these are essential function-
partners and of the quality of the offering. alities for controlling the home. Apple
Moreover, a more closed ecosystem helps pre-installs Apple Maps on the iPhone in
the orchestrator capture value by, for ex- an attempt to oust Google Maps. And Goo-
ample, charging for access. gle uses its Google Play store to control the
otherwise very open Android ecosystem.
The right level of openness will depend on
the relative importance of the individual There are, of course, many other initial
factors, such as growth versus quality, de- governance questions. For example, when
centralized versus coordinated innovation, designing a transaction ecosystem, the plat-
and speed versus consistency of co-evolu- form orchestrator must decide whether the
tion. Competition with other existing or matching of producers and consumers
emerging ecosystems in the same sector should be done by algorithm (Uber) or by
can also play a role, because a new ecosys- users (Facebook); whether pricing should
tem needs to find a differentiated position- be based on rules and algorithms (Lending-
ing, such as the degree of openness. Club) or on offer and negotiation (eBay);
and whether curation should be done by
We have seen many ecosystems start with platform editors (Wikipedia), user feedback
a rather closed governance model in order (Airbnb), or algorithms (Google Search).
to establish high quality and open up later. These decisions depend on the specific
For example, the Q&A platform Quora context and get at the heart of the ecosys-
started as an invite-only ecosystem that tar- tem’s operating model, value creation
geted prominent technology entrepreneurs. mechanism, and differentiation.
By building this dense and exclusive net-
work of experts, Quora was able to develop We will address the question of ecosystem
an inventory of high-quality content that governance in a future article in this series.
then made it easy to attract a broader audi-
ence when the platform later opened up.
However, there are also examples of eco- Step 4: How Can You Capture
systems that start as open to gain traction the Value of Your Ecosystem?
and become more closed later, such as the
knowledge ecosystems investigated by Järvi What Should You Charge?
and her colleagues.3 When the basic setup of the business eco-
system is defined, the next big design step
What Should the Orchestrator is to find a way to translate the benefits
Control? that the ecosystem creates for its customers
As an orchestrator, you face an additional into value for its participants. Monetization
design question: What do you want to do is one of the biggest challenges of the eco-
Airbnb is an example of an ecosystem with And then there are examples of ecosystem
both substantial demand-side economies of plays that have neither substantial supply-
scale (indirect network effects) as well as nor demand-side economies of scale. We
supply-side economies of scale (from call them dead ends. An example is the orig-
spreading the high fixed-cost for technolo- inal Yahoo internet portal and search en-
gy and marketing). We characterize this gine, which started as an employee-edited
model as a flywheel, with winner-takes-all- hierarchical database that classified web-
or-most characteristics. pages using a tree structure of categories.
This model worked well for some time, but
Some ecosystems have supply-side econo- as the internet grew exponentially, it be-
mies of scale but only limited demand-side came apparent that it was not scalable, and
economies of scale, such as additive manu- Yahoo was overtaken by Google with its au-
facturing and many other solution ecosys- tomatic and easily scaling page-rank algo-
tems with small network effects. They need rithm.
to follow a path of patience when it comes
to growth, but they also have a good It is important to understand the scalabili-
chance of achieving a profitable and de- ty position of your emerging ecosystem
fendable position. and to adapt your ecosystem design and
ecosystem strategy accordingly. However,
More critical are ecosystem models that scalability is only the first step toward
have high demand-side economies of scale long-term viability. To thrive in the long
PATH OF
DEAD END
PATIENCE
Low
Low High
Supply-side economies of scale
However, our analysis of successful and And finally, ecosystem carryover happens
failed ecosystems showed that many eco- when a successful business ecosystem ex-
systems found it easier to achieve scale pands into a neighboring domain. It is an
than to sustain it. Defending a strong posi- important route for ecosystem growth and
tion as an ecosystem is challenging because expansion, as we will discuss in the next
an attack can target either the demand or section, but also a significant threat for es-
the supply side of the market. We identi- tablished ecosystems. A special case of eco-
fied four main mechanisms of attack that system carryover is nested ecosystems. For
ecosystems need to be aware of. example, we could imagine that ride-hail-
ing platforms at some point may come un-
First, multihoming happens when suppliers der pressure from broader mobili-
or consumers participate in multiple com- ty-as-a-service ecosystems that include
peting ecosystems at the same time, or eas- multimodal transport solutions, which may
ily switch between ecosystems. Restaurants in turn be attacked by even broader smart-
may find it attractive to offer their dishes city ecosystems.
on multiple food-delivery platforms, for in-
stance, and consumers use different ho- Digital technologies, while offering new
tel-booking platforms to chase the best of- ways of exploiting network effects and sup-
fering. Multihoming is a particular risk for ply-side economies of scale, also make it
an ecosystem if switching costs are low. For more difficult to defend an established po-
example, because credit cards tend to have sition. Digital business models have much
low annual fees, many people carry multi- lower entry barriers than traditional brick-
ple cards in their wallet. By contrast, only and-mortar businesses. It is so easy to cre-
few people can afford to carry both an An- ate a digital platform today that you do not
droid and Apple phone, so they tend to even need to program your own software
choose one model and stay with it for at but can build it from components available
least a couple of years. on cloud-services platforms. Digital net-
work effects as a barrier to entry are much
Second, disintermediation happens when weaker than the physical network effects
partners from two sides of a transaction of a railway or telephone network. More-
ecosystem bypass the matching platform over, they can quickly be reversed once a
and connect directly. For example, Home- network starts losing users and gets into a
joy, an online platform that connected cus- downward spiral, as once-dominant plat-
tomers with home service providers, in- forms like Second Life and BlackBerry have
cluding house cleaners and handymen, painfully experienced. And finally, the high
suffered from disintermediation as custom- rate of innovation and technological dis-
The application of the four defensibility The most effective defense, however, is to
tests to ride-hailing platforms such as Uber, ensure that you offer the best available
Lyft, and Didi uncovers the fundamental product or platform and the best overall
challenges of their business models. On the ecosystem solution. Having only a superior
positive side, there is only limited threat product or platform is no longer sufficient
from niche specialization in ride-hailing, in an ecosystem world, where competition
and the platforms can clearly benefit from happens at the system level. In many ways,
substantial indirect network effects. How- the BlackBerry was superior to the iP-
ever, these effects are only local and diffi- hone—in terms of data security, keypad,
cult to transfer to new cities. More impor- and battery life—but Apple offered the
tantly, ride-hailing suffers from far-better overall solution with its ecosys-
multihoming and low switching costs for tem of app developers.
riders, who have a high incentive to use
multiple ride-hailing services, as well as At the same time, you must not neglect
drivers, who can easily switch between your core product. In 2004, Microsoft’s In-
platforms or even serve multiple platforms ternet Explorer captured close to 95% mar-
at the same time. As indicated above, the ket share and was widely considered to
resulting high recruiting and retention have won the browser war. However, with
costs for drivers lead to limited supply-side no serious competitor left, Microsoft under-
economies of scale, further reducing barri- invested in the further development of the
ers to entry. browser and its underlying ecosystem. Infe-
rior product execution and product innova-
By contrast, video games score better in tion from 2004 to 2015 allowed Firefox and
the defensibility test. They exhibit moder- Chrome to enter and eventually dominate
ate, but global, indirect network effects, the market.
Martin Reeves is a managing director and senior partner in BCG’s San Francisco office and the chair-
man of the BCG Henderson Institute. You may follow him on Twitter @martinKReeves and contact him by
email at reeves.martin@bcg.com.
Maximilian Schüssler was a project leader in BCG’s Munich office and an ambassador at the BCG Hen-
derson Institute.
To succeed, organizations must blend digital and human capabilities. Our diverse, global teams bring deep
industry and functional expertise and a range of perspectives to spark change. BCG delivers solutions
through leading-edge management consulting along with technology and design, corporate and digital
ventures—and business purpose. We work in a uniquely collaborative model across the firm and through-
out all levels of the client organization, generating results that allow our clients to thrive.
For information or permission to reprint, please contact BCG at permissions@bcg.com. To find the latest
BCG content and register to receive e-alerts on this topic or others, please visit bcg.com. Follow Boston
Consulting Group on Facebook and Twitter.