Innovation Systems: Aligning the Incentives

By Philip Evans

One of the most striking elements of this year’s list of up-and-coming innovators is the number of companies that depend—in whole or in part—on platforms and standards created by two longstanding most-innovative companies: Apple (number one) and Google (number three). WhatsApp and Line are both messaging apps, and Pinterest, Square, Netflix, and Spotify all leverage Apple’s iOS and Google’s Android platforms to bring a mobile dimension to their products.

These interrelationships highlight the role that system architecture increasingly plays in innovation. For an innovation system to work, it needs to be more than the sum of its parts. In the mobile world, for example, the iOS and Android platforms radically simplify the technical and marketing challenges for app makers, lowering the cost of innovation. And by enabling the app makers, Apple and Google help shape the relative competitive advantage of iOS versus Android—and the device and advertising revenues they drive. Everybody wins because standards allow different parts of the overall system to evolve independently. The system enjoys the benefits of a portfolio of independent bets; in a world of open recombination, anyone can try his or her hand, and no innovator has to ask permission.

The plain fact, as we all know, is that no company can be good at everything. The old AT&T, for example, was brilliant at fundamental innovation (seven Nobel prizes!), but it plodded when it came to making a new handset. (The Trimline took 13 years to bring to market.) Toyota, by contrast, excels at “dimension of merit” innovations—that is, the aggregation of thousands of small improvements in cost or weight or efficiency. Silicon Valley, for all its self-referential mystique, is most talented at recombining existing technologies in order to address unmet user needs. 

There’s a good reason for this. Different kinds of innovation require their own institutional arrangements and capabilities: different talents, cultures, operating scales, funding methods, incentives, time horizons, and appetites for risk. A smartphone, for example, combines fundamental innovations that originated in universities and the labs of Bell or IBM; components are the product of sustained dimension-of-merit improvements by the likes of Corning, Qualcomm, and Nvidia; and the coolest apps typically come from young outfits such as WhatsApp or Groupon—companies that bet big on a hypothesis about what will be fun or useful for users. 

Making all these different pieces work together requires legal interfaces, such as patent licensing, and technical interfaces, such as application programming interfaces. These interfaces are the architecture of an innovation system, often dominated by a single orchestrator such as Apple or Google.

The best architecture—and not the best product—wins in the long run. But architecture is a two-edged sword: it can also constrict innovation. Architecture creates boundaries—and while it liberates innovation within those boundaries, it also inhibits innovation across them. Whether this is a gain or a loss depends on the product and where it stands in its life cycle. When the product is young and performance-constrained, architectural flexibility is typically worth more; as the product matures, the locus of innovation shifts from performance to cost and customization, so modular interoperability tends to gain ground. 

In designing or trying to shape an innovation system, the key is to focus on how to create the technical, economic, and legal conditions that will enable the system to flourish symbiotically by aligning the incentives of diverse participants. Ultimately, it is the end-user experience that matters, and just as in a repertory theater, that depends not on the stars but on the performance of the ensemble.