Big Science only gets you so far. Processes matter, too.

Ted Hewitt and Ed Greenspon

When Canadians think about innovation, they usually conjure up images of caffeine-soaked entrepreneurs launching nifty new products that touch their lives as consumers. Or of grand challenges with the potential to generate big change—“moonshots”, in the parlance of policy, a reference to John F. Kennedy’s 1961 vow to put a person on the moon before the end of the decade.

All that’s quite fine. It’s just that it only captures a fraction of the innovation activity and ignores the plainer but arguably more productive role of innovation—that of linking human ingenuity and advanced technology to new processes for existing applications.

The long march to the autonomous vehicle covers all the bases. First, you link the inner workings of the automobile to a central computer intelligence that oversees such things as navigation or collision avoidance. Next, this “connected” car drives you rather than you driving it, morphing into the autonomous vehicle.

To reach that destination, all kinds of things need to happen along the innovation chain, from idea generation to engineering to assembly.

Then there’s the even less sexy incremental “process” improvements that occur on shop floors every other day, with companies grinding out significant productivity gains.

In the moonshot model, commercial success is generally assumed to flow from the inherent qualities of the widget itself, as measured by its speed, utility, design, portability and other benefits to consumers. In Canada, the BlackBerry is often touted as the model par excellence—except that the narrative usually skips over the unglamorous process gains, punctuated by periodic creative leaps, along the BlackBerry’s journey to an “overnight” sensation.

As policy discussions on innovation once again claim centre stage, it’s time to rethink overly simplistic maps of innovation success. From the archives of the MIT Sloan Management Review comes a reminder of the real stuff of innovation, much of which has little or nothing to do with product development per se. The 2007 Sloan article lists five distinct stages in the innovation process, each one essential.

Stage 1 is idea generation and mobilization, where talented individuals bring forth proposals with the potential for commercialization. Stage 2 involves advocacy and screening; here, innovative ideas must be sold on their merits to potential gatekeepers (senior managers, investors) within and sometimes outside an organization, based on their potential to make (or save) money.

In the real world, light bulbs don’t just switch on in someone’s head. Yet in terms of how innovation is normally understood in policy terms, (research) gets nearly all the oxygen.

In Stage 3, experimentation occurs. This involves background research, prototype design and development, and testing. Stage 4 involves commercialization—the point where funding must be secured and market potential explored. If the tree falls in the forest and nobody notices, you’re likely to end up with a busted innovation.

Stage 5, the final link, involves setting up the mechanisms to effectively deliver the innovation, which may often entail displacing another product or process that has its loyal fans. If the engineers and marketers haven’t stayed on the same page since Stage 2, the absence of the alignment required to ensure a smooth roll out will quickly expose itself.

There is nothing particularly surprising about this five-stage process; in the real world, light bulbs don’t just switch on in someone’s head. Yet in terms of how innovation is normally understood in policy terms, Stage 3 gets nearly all the oxygen.

This is particularly true of government efforts to promote innovation. For example, “innovation” funding channeled to industry through Canada’s granting councils and science agencies—frequently in partnership with post secondary institutions—is directed primarily towards product development research. Similarly, government support for industry research and development through programs such as the SR&ED tax credit focus almost exclusively on scientific experimentation and testing costs linked to tangible materials and devices.

Creative, managerial, commercialization and marketing processes (Stages 1, 2, 4 and 5) are the stuff of great businesses—and they get short shrift. In fact, we know very little about the contribution of R&D investments in these areas as they are currently counted nowhere—not by statistical agencies, not in any statement of national accounts. It makes a true evidence-based approach difficult to achieve. An innovation policy should insist on knowing where the best bang for its buck is generated.

So, what to do?

• We need to embrace the reality that innovation includes not only tangible product development but also the introduction of new services and processes that enhance productivity and create value. This is especially important in an economy 70 per cent driven by the service sector.
• Investments in scientific research and development must be complemented by appropriate investments in activities that support entry into the marketplace. Until we connect the science component with behavioural research, marketing, finance, etc., commercialization will remain elusive. Great discoveries that don’t make it into the order books are an exercise in the theoretical. They don’t build economies and create jobs.
• Finally, with evidence (thankfully) back in fashion, we need to lock down a commitment that investments in each part of the innovation cycle will be fully counted and analyzed, so as to better understand the best pathways to success and where policy can produce a tipping point.

Only then will we have an innovation system whose complexity is sufficiently understood to discern what’s working and not, and why, from the moonshot to the mundane. That’s the table stakes for good policy.