For the past 30 years, “creative destruction” has been a source of
fascination at top-tier business schools and in magazines like this one.
The almost obsessive interest in this topic is unsurprising, given the
ever-changing, never-ending list of transformative threats—which today
include the internet of things, 3-D printing, cloud computing,
personalized medicine, alternative energy, and virtual reality.
Our understanding of the shifts that disrupt businesses, industries,
and sectors has profoundly improved over the past 20 years: We know far
more about how to identify those shifts and what dangers they pose to
incumbent firms. But the timing of technological change remains
a mystery. Even as some technologies and enterprises seem to take off
overnight (ride sharing and Uber; social networking and Twitter), others
take decades to unfold (high-definition TV, cloud computing). For firms
and their managers, this creates a problem: Although we have become
quite savvy about determining whether a new innovation poses a threat, we have very poor tools for knowing when such a transition will happen.
The number-one fear is being ready too late and missing the
revolution (consider Blockbuster, which failed because it ignored the
shift from video rentals to streaming). But the number-two fear should
probably be getting ready too soon and exhausting resources before the
revolution begins (think of any dot-com firm that died in the 2001
technology crash, only to see its ideas reborn later as a profitable Web
2.0 venture). This fear of acting prematurely applies both to
established incumbents being threatened by disruptive change and to
innovating start-ups carrying the flag of disruption.
To understand why some new technologies quickly supplant their
predecessors while others catch on only gradually, we need to think
about two things differently. First, we must look not just at the
technology itself but also at the broader ecosystem that supports it. Second, we need to understand that competition may take place between the new and the old ecosystems,
rather than between the technologies themselves. This perspective can
help managers better predict the timing of transitions, craft
more-coherent strategies for prioritizing threats and opportunities, and
ultimately make wiser decisions about when and where to allocate
organizational resources.
You’re Only as Good as Your Ecosystem
Both established and disruptive initiatives depend on an array of
complementary elements—technologies, services, standards, regulations—to
deliver on their value propositions. The strength and maturity of the
elements that make up the ecosystem play a key role in the success of
new technologies—and the continued relevance of old ones.
The new technology’s ecosystem.
In assessing an emerging technology’s potential, the paramount
concern is whether it can satisfy customer needs and deliver value in a
better way. To answer that question, investors and executives tend to
drill down to specifics: How much additional development will be
required before the technology is ready for commercial prime time? What
will its production economics look like? Will it be price-competitive?
If the answers suggest that the new technology can really deliver on
its promise, the natural expectation is that it will take over the
market. Crucially, however, this expectation will hold only if the new
technology’s dependence on other innovations is low. For example, a new
lightbulb technology that can plug into an existing socket can deliver
its promised performance right out of the box. In such cases, where the
value proposition does not hinge on external factors, great product
execution translates into great results.
However, many technologies do not fall into this plug-and-play mold.
Rather, their ability to create value depends on the development and
commercial deployment of other critical parts of the ecosystem. Consider
HDTV, which could not gain traction until high-definition cameras, new
broadcast standards, and updated production and postproduction processes
also became commercially available. Until the entire ecosystem was
ready, the technology revolution promised by HDTV was bound to be
delayed, no matter how great its potential for a better viewing
experience. For the pioneers who developed HDTV technology in the 1980s,
being right about the vision brought little comfort during the 30 years
it took for the rest of the ecosystem to emerge.
An improved lightbulb and an HDTV both depend on ecosystems of
complementary elements. The difference is that the lightbulb plugs into
an existing ecosystem (established power generation and distribution
networks; wired homes), whereas the television requires the successful
development of co-innovations. Improvements in the lightbulb will thus
create immediate value for customers, but the TV’s ability to create
value is limited by the availability and progress of other elements in
its ecosystem.
About the Research
We developed and explored the ideas described in this article
during a five-year research project on the pace of substitution in the
semiconductor-manufacturing ecosystem.
The semiconductor industry’s remarkably robust progress over the
past 60 years was made possible by innovations in the lithography
technology that semiconductor manufacturers use. We studied the
successive generations of lithography equipment and noticed a pattern:
In some cases, the new technology dominated the market in a matter of
two to five years, whereas in other cases it faced prolonged, unexpected
delays in achieving market dominance—and sometimes never did. This was
true despite the fact that each generation offered superior performance,
even on a price-adjusted performance basis.
To test our hypotheses about how ecosystem emergence challenges and
extension opportunities affect the pace of substitution, we first
collected and analyzed detailed data on every product and firm involved
in every generation of the technology. We supplemented that information
with extensive interviews with executives from firms throughout the
ecosystem.
Our statistical analysis showed that 48% of the variation in the
pace of substitution was attributable to traditional factors:
price-adjusted performance differences, the number of rivals in the
market, and the tenure of the old technology. When we added
consideration of the ecosystem dynamics discussed in the article, we
were able to account for a remarkable 82% of the variance.
For more details on the research, see “Innovation Ecosystems and the Pace of Substitution: Re-examining Technology S-Curves,” by Ron Adner and Rahul Kapoor, Strategic Management Journal (March 2015).
Read more
The old technology’s ecosystem.
Successful, established technologies—by definition—have overcome
their emergence challenges and are embedded within successful,
established ecosystems. Whereas new technologies can be held back by
their ecosystems, incumbent technologies can be accelerated by
improvements in theirs, even in the absence of progress in the core
technology itself. For example, although the basic technology behind bar
codes has not changed in decades, their utility improves every year as
the IT infrastructure supporting them allows ever-more information to be
extracted. Hence in the 1980s, bar codes allowed prices to be
automatically scanned into cash registers; in the 1990s, aggregating the
bar code data from daily or weekly transactions provided insight into
general inventory; in the modern era, bar code data is used for
real-time inventory management and supply chain restocking. Similarly,
improvements in DSL (digital subscriber line) technology have extended
the life of copper telephone lines, which can now offer download speeds
of 15 megabytes per second, making copper-wire services competitive with
newer cable and fiber networks.
The War Between Ecosystems
When a new technology isn’t a simple plug-and-play substitution—when
it requires significant developments in the ecosystem in order to be
useful—then a race between the new- and the old-technology ecosystems
begins.
What determines who wins? For the new technology, the key
factor is how quickly its ecosystem becomes sufficiently developed for
users to realize the technology’s potential. In the case of cloud-based
applications and storage, for example, success depended not just on
figuring out how to manage data in server farms, but also on ensuring
the satisfactory performance of critical complements such as broadband
and online security. For the old technology, what’s important
is how its competitiveness can be increased by improvement in the
established ecosystem. In the case of desktop storage systems (the
technology that cloud-based applications would replace), extension
opportunities have historically included faster interfaces and
more-robust components. As these opportunities become exhausted, we can
expect substitution to accelerate.
Thus the pace of substitution is determined by the rate at which the
new technology’s ecosystem can overcome its emergence challenges
relative to the rate at which the old technology’s ecosystem can exploit
its extension opportunities. To consider the interplay between these
forces, we have developed a framework to help managers assess how
quickly disruptive change is coming to their industry.
There are four possible scenarios: creative destruction, robust resilience, robust coexistence, and the illusion of resilience.
Creative destruction.
When the ecosystem emergence challenge for the new technology is low
and the ecosystem extension opportunity for the old technology is also
low (quadrant 1 in the framework), the new technology can be expected to
achieve market dominance in short order (see the video “How Fast Does
New Technology Replace the Old?”). The new technology’s ability to
create value is not held back by bottlenecks elsewhere in the ecosystem,
and the old technology has limited potential to improve in response to
the threat. This quadrant aligns with concept of creative
destruction—the idea that an innovative upstart can swiftly cause the
demise of established competitors. While the old technology can continue
serving niches for a long time (see “Bold Retreat,”
by Ron Adner and Daniel C. Snow, HBR, March 2010), the bulk of the
market will abandon it relatively quickly in favor of the new
technology. As an example, consider the rapid replacement of dot matrix
printers by inkjet printers.
Robust resilience.
When the balance is reversed—when the new technology’s ecosystem
confronts serious emergence challenges and the old technology’s
ecosystem has strong opportunities to improve (quadrant 4)—the pace of
substitution will be very slow. The old technology can be expected to
maintain a prosperous leadership position for an extended period. This
quadrant is most consistent with technologies that seem revolutionary
when they’re first touted but appear overhyped in retrospect.
Bar codes and radio frequency identification (RFID) chips provide a
good example. RFID chips hold the promise of storing far richer data
than bar codes ever could, but their adoption has lagged because of the
slow deployment of suitable IT infrastructure and nonuniform industry
standards. Meanwhile, IT improvements have extended the usability of bar
code data, as we’ve already discussed, relegating RFID to niche
applications and keeping the RFID revolution at bay for the past two
decades. It may well be that RFID does eventually overcome its
challenges and that ecosystem extension opportunities dry up for bar
codes. If this happens, the dynamics will shift from quadrant 4 to
another quadrant, and the pace of substitution will quicken. But that
will be small consolation to the firms and investors that committed to
RFID decades ago. The opportunity cost of waiting for the rest of the
system to catch up can mean that being in the right place 10 years too
soon is more costly than missing the revolution completely.
When substitution is slow, there are also implications for the new
technology’s required performance levels. Every time IT improvements
make bar codes more useful, for example, the quality threshold for the
RFID technology is raised. Thus performance expectations for the
innovation keep ratcheting upward, even as its wide adoption is held
back by the underdeveloped state of its ecosystem.
Robust coexistence.
When the ecosystem emergence challenge for the new technology is low
and the ecosystem extension opportunity for the old technology is high
(quadrant 2), competition will be robust. The new technology will make
inroads into the market, but improvements in the old-technology
ecosystem will allow the incumbent to defend its market share. There
will be a prolonged period of coexistence. Although extension
opportunities are unlikely to reverse the rise of the new technology,
they will materially delay its dominance.
An instructive example is the competition between hybrid
(gas-electric) automobile engines and traditional internal-combustion
engines. Unlike fully electric engines, which need a supporting network
of charging stations, hybrids were not held back by ecosystem emergence
challenges. At the same time, however, traditional gas engines have
become more fuel-efficient, and the ecosystem for the traditional
technology has improved, too, as gas engines have become better
integrated with other elements in the vehicle, such as heating and
cooling systems.
A period of robust coexistence can be quite attractive from a
consumer perspective. Performance of both ecosystems is improving—and
the better the old technology’s ecosystem becomes, the higher the
performance bar is for the new technology’s ecosystem.
The illusion of resilience.
When the ecosystem emergence challenge is high for the new technology
and the ecosystem extension opportunity is low for the old technology
(quadrant 3), not much will change until the emergence challenge is
resolved—but then substitution will be rapid. Examples here are HDTVs
versus traditional TVs, and e-books versus printed books. Both of those
revolutions were delayed not by advances in the old technology’s
ecosystem but by ecosystem-emergence challenges in the new technology.
In scenarios in this quadrant, an industry analysis will most likely
show that the old technology maintains high market share, but growth has
stalled. Because rapid market-share inversion is to be expected once
the new technology fulfills its value creation potential, the dominance
of the old technology is fragile. It is maintained not by continued
progress in the old technology but by setbacks for the new competitor.
Implications for Action
Once you understand that in the race to dominance, ecosystems are
just as important as technologies, you will be better at thinking
through how quickly change is going to occur—and deciding what level of
performance you need to aim for in the meantime. We will consider how to
tackle these questions shortly, but first let’s review a few general
truths that emerge from this perspective.
- If your company is introducing a potentially transformative
innovation, the full value will not be realized until all bottlenecks in
the ecosystem are resolved. It may pay to focus a little less on
perfecting the technology itself and a little more on resolving the most
pressing problems in the ecosystem.
- If you are a threatened incumbent, it pays to analyze not just the
emerging technology itself but also the ecosystem that supports it. The
greater the ecosystem-emergence challenge for the new technology, the
more time you have to strengthen your own performance.
- Strengthening incumbent performance may mean improving the old
technology—but it can just as easily mean improving aspects of the
ecosystem that supports it.
- Every time the old technology’s performance gets better, the performance threshold for the new technology goes up.
With that overview in mind, let’s look at how to use this framework
to analyze your own technology strategy. We recommend having executive
conversations focused on two questions: Which quadrant is our industry
in? and What are the implications for our resource allocation and other
strategic choices?
Which quadrant are we in?
Without the benefit of hindsight, your response to this question is
clearly a matter of judgment. Some people would look at electric
vehicles in 2016 and say they are still stuck in quadrant 4 (where we
have placed them in our framework), pointing out that the charging
infrastructure and battery performance are insufficient for mainstream
adoption. Other people would position EVs on the cusp of quadrant 2,
claiming that acceptance is growing and that better batteries make it
possible to drive longer distances before recharging. Still others would
place EVs solidly in quadrant 2, arguing that Tesla’s success in
selling its vehicles and populating its waiting lists is a sure sign
that commercial potential is no longer constrained.
How Big a Threat Is the New Technology?
Predicting the pace of substitution requires
analyzing the competition between the new- and the old-technology
ecosystems. Six questions can help innovators and incumbents assess
their positions and strategies.
New-Technology Questions
These questions (drawn from The Wide Lens, by
coauthor Ron Adner) address the emergence challenges that confront the
new technology. The answers should help innovators decide how to adjust
their strategies.
- What is the execution risk—the level of difficulty in delivering the focal innovation to the market on time and to spec?
- What is the co-innovation risk—the extent to which the success of the new technology depends on the successful commercialization of other innovations?
- What is the adoption-chain risk—the extent to which
other partners need to adopt and adapt to the new technology before end
consumers can fully assess its value proposition?
The greater the extent to which the new technology is
facing any of these risks, the greater the challenge to be overcome, and
the longer the expected delay in adoption of the technology.
Old-Technology Questions
These questions address the prospects for improving the
competitiveness of the incumbent technology. The answers should help
incumbents identify opportunities they might exploit.
- Can the competitiveness of the old technology be extended by further improvements to the technology itself?
- Can it be extended by improvements to complementary elements in its ecosystem?
- Can it be extended by borrowing from innovations in the new technology and its ecosystem?
The more positive the reply to each of these questions, the greater the extension opportunity for the old technology.
Read more
The sidebar “How Big a Threat Is the New Technology?” suggests issues
to think through as you debate which quadrant you’re in. Some questions
pertain to the new technology and some to the old—but you will want to
consider them all, regardless of whether you are an incumbent or a
start-up. Don’t expect all individual team members to agree on the
answers to these questions. It is precisely by going through the process
of articulating different views that teams can make the most of their
collective insights.
What are the implications for resource allocation and other strategic choices?
Each quadrant in the framework carries different implications for
resource allocation decisions. And since markets are not transformed all
at once, the quadrant also suggests possible ways to position yourself
during the transition.
In quadrant 1 (creative destruction), with the old technology
stagnant and the new technology unhampered, innovators should
aggressively invest in the new technology. Incumbents should follow the
familiar prescriptions for embracing change to withstand the winds of
creative destruction. Part of that is looking for niche positions where
they can survive in the long term with the old technology. For example,
pagers were largely replaced by cell phones, but they are still used by
emergency-service providers.
In quadrant 2 (robust coexistence), incumbent firms can continue to
invest in the old technology and aggressively invest in improvements to
the ecosystem, knowing that the new and the old technologies will
coexist for an extended period. As in quadrant 1, they should also seek
niche positions for the old technology for the long term, but there is
less urgency to do so. New-technology innovators should move full speed
ahead on perfecting the new technology along with its complements. That
includes testing and refining the offering with early adopters and
segments that are potentially receptive.
In quadrant 3 (the illusion of resilience), new-technology champions
should direct resources toward resolving their ecosystem challenges and
developing complementary elements, and resist overprioritizing further
development of the technology itself. When the bottleneck to adoption is
the ecosystem, not the technology, pushing technology progress is
pushing the wrong lever. Incumbents, for their part, must guard against
the false assumption that they are maintaining their market position
because of the merits of their own technology. As publishers of road
atlases will attest, this is probably a time to harvest and make only
incremental improvements, with an eye toward sunset; it is not the time
to redouble innovation efforts in the old technology.
One final note about the dynamics of change. Every innovator wants to
end up in quadrant 1 so that it can play the classic
creative-destruction game. But there are different paths for getting
there. A hypothesis that predicts a transition path from Q4 to Q3 to Q1
is a bet on the exhaustion of the old technology. For an innovator, that
would mean focusing on aligning the new-technology ecosystem without
great concern for extending a performance advantage. In contrast, a
predicted path of Q4 to Q2 to Q1 would mean competing against an
improving incumbent-technology ecosystem. Here the innovator needs to
continually elevate its performance while it simultaneously perfects the
ecosystem.
