Are Scientists Deluded (about innovation)?

iStock_000003455087Small

[This article was originally posted on LinkedIn October 20, 2014]

Peter Thiel recently gave a lecture as part of a Stanford course on How to Start a Startup. As part of this talk, in characteristically provocative fashion, he characterized scientists as deluded. His comments were in the context of showing that it was the combination of the amount of value created and the amount of value captured that dictated the rewards to innovators. Here is his comment (from ~29 minutes into the video):

Certainly if you go back to you need to create X dollars in value and you capture Y percent of X, I would suggest that the history of science has generally been one where Y is zero percent across the board, the scientists never make any money. They’re always deluded into thinking that they live in a just universe that will reward them for their work and for their inventions.”

Is there any merit to this criticism? I would say that there is some, in that when many (far from all!) scientists attempt to become technical entrepreneurs they can make some serious errors and fail to capture the value they might. On the other hand, I think that this statement represents a fundamental misunderstanding of the scientific enterprise.

It may be unwise to disagree with Thiel, but it seems to me that he errs in that much of what he calls scientific innovations were really advances in knowledge and understanding (discovery). If you look at slide 19 of his deck you will see that he refers to things like the ideal gas law, understanding the atom, discovery of antibiotic activity, and genome mapping as ‘scientific innovations’. In doing this he is failing to differentiate between three slightly distinct processes: (a) the discovery of knowledge (discovery, or basic science), (b) the application of that knowledge to yield a new or improved outcome (invention, or applied science), and (c) the adaptation of that invention by innovators to create value for users who then adopt it (innovation). It is certainly true that all of these steps (discovery, invention, innovation) contribute to the creation of value, but the distance of discovery from the creation of value is one of the main rationales behind the public funding of basic science. So, to me, Theil fails to differentiate between discovery and innovation, and it is unfair to call scientists working at the discovery stage deluded.

That being said, the merit in Theil’s criticism is that it can be argued that in many (but far from all!) cases, scientists can do a poor job of capturing value from their contributions, even when they move into the role of innovator. A broad discussion of the need, at least as felt by scientists, to capture more value from their work is far too much to attempt here. However, it is my opinion that many scientists would benefit from a better understanding of innovation and the process of innovation. Here are three of several areas where scientists need to better understand innovation:

(1) Innovation is more than just invention. An invention (defined above as the application of knowledge to yield a new or improved way of achieving an outcome) does not truly become an innovation—the point at which value capture becomes most possible—unless and until that invention is adopted and put into practice by users willing to pay for the privilege. In other words, someone needs to want to use the invention badly enough to pay for the use of the invention. An invention that is never used is clearly not an innovation.

This aspect of innovation is both obvious (when stated) yet frequently neglected or misunderstood by would-be technical entrepreneurs. The issue is that it is easy for technical entrepreneurs to assume that a group of customers will want to use their invention, based on the technologist’s own experiences. Sometimes this is a correct assumption, particularly when the technologist represents a lead user within the target market segment. Frequently, however, the technologist is extrapolating; in this case they are subconsciously expecting prospective users to have experience and knowledge similar to theirs, and to have values, and make judgments the same way that they do. In particular, technical entrepreneurs tend to focus heavily upon technical merit.

But, extensive work on the diffusion and adoption of innovations—for instance as summarized by Everett Rogers in his 2003 book, The Diffusion of Innovations—shows that the intended users of technology based innovations frequently have unique values and biases, and do not make decisions strictly on the basis of technical merit. The fact that a technologist’s intuition about an invention is insufficient to achieve adoption is also behind things like Geoffrey Moore’s ideas inCrossing the Chasm; the values, perceptions, needs and behavior of various user segments can differ radically from those of would-be innovators, and from that of other segments.

(2) Assumptions must be tested. Given the preceding point, would-be technical entrepreneurs need to know how to test their assumptions about their users or customers (the market). These then need to be tested and refined until they have a sufficient understanding of actual needs, and their customer’s actual behavior relative to adopting solutions to meet those needs, so that they can successfully introduce their invention in ways that lead to its adoption.

This need has been recognized by people like Steve Blank and Eric Ries, leading to the ideas behind Lean Startup. The implementation of strategies like A-B testing on the internet also reflect this need to test assumptions, or at least to test. Ironically, this is not so widely recognized (in my experience) among scientists. The irony is that this approach is just an application of hypothesis testing, a part of the scientific method that all scientists should already understand well. The trap for scientists is that they are just as likely as anyone else to fall in love with their own idea and to proceed solely on the basis of assumptions.

(3) All innovations are not created equal. Finally, given the fact that most scientists involved in applied research operate as employees of a company, it is important for scientists to understand the differences (outlined by Clayton Christensen and Michael Raynor in their treatments of Disruption Theory) between different types of innovations (i.e. disruptive vs. sustaining). Different types of innovations need to be developed and introduced by different types of organizations.

Many large technical organizations are, of course, incumbents in mainstream markets and are actually quite good at creating sustaining innovations for their markets. And, in that venue industrial scientists can shine. However, the very fact that this is true can handicap the scientists within those companies when they mistakenly assume that all innovations proceed in that fashion. A scientist within a mainstream incumbent working on a technology that would generate value via disruption may be in for a rough ride.

Conversely, individual scientists (or small teams) wishing to become technical entrepreneurs via a startup (for instance) need to recognize the challenges they will face if they attempt (as a lone-entrant) to bring a sustaining innovation to an existing mainstream market.

Conclusion

So, I would argue that scientists are not actually deluded in that they do not actually conduct discovery work in the belief that a just universe will amply reward them for their efforts. But, I would argue that when scientists attempt to become innovators, they do frequently need to work to better understand innovation as a process. They need to recognize when they are making assumptions, and recognize that even with objective merit, inventions are adopted by real users in quirky ways that depend upon sociology and psychology as much as technical assessments. Consequently, they need to test and refine their assumptions about how users would perceive their inventions in the same way they would test and refine an experimental hypothesis. Finally, they need to correctly understand the landscape for innovations. While some innovation exercises are aimed at maintaining a competitive advantage in established markets, other innovation exercises are aimed at finding and satisfying an entirely new (and hopefully rapid growth) market in a unique way. These exercises require different organizations and approaches.