Outsourcing research

InnoCentive has an intriguing business model.  As described on their website:
"InnoCentive® is an exciting web-based community matching top scientists to relevant R&D challenges facing leading companies from around the globe. We provide a powerful online forum enabling major companies to reward scientific innovation through financial incentives."
If your company cannot solve an important technical problem, you register as a seeker; if you are a scientist, engineer, etc  with a bit of free time and an itch to solve some interesting problems, you register as a solver. InnoCentive gets the seekers and solvers together, with financial prizes from the seekers to the solver or solvers who produce the useful solutions.  The Rockefeller Foundation has recently partnered with InnoCentive to apply the same platform to global development problems.  In this case, the seekers will be non-profit entities chosen by the Foundation that serve poor or vulnerable peoples.

One might be tempted to think that this “dating service” approach to problem solving could be useful only to small start-ups that cannot afford their own research.  Turns out, if you think that, you are really wrong.  A recent Harvard Business School working paper by K.R. Lakhani, L.B.Jeppesen, P.A.Lohse and J.A.Panetta has analyzed InnoCentive results for 166 scientific problems that the research laboratories of “large and well-known R&D-intensive firms had been unsuccessful in solving internally.” Several of the problems reflected several years of unsuccessful effort in the company’s research labs.  The results are fascinating.

First, the bottom line: 29.5% of the problems posed by the seekers were solved by the solvers. For the 166 problems studied, 240 individual scientists on average examined each detailed statement of the problem, and 10 individuals submitted solutions for evaluation.   Winning solvers spent on average 74.1 hours developing the solution - not bad for problems that had frustrated entire research labs, sometimes for extended periods.

The most interesting results came from the question “Who becomes a successful solver?.” The answer -“ the further the focal problem was from the solver’s field of expertise, the more likely they were to solve it.” And an equally fascinating corollary, “ being more specialized resulted in a higher probability of creating a winning solution.” Thus the most likely problem solver was  was not a “renaissance” figure, but an expert in a relatively narrow domain completely outside the domain of the problem. 

And how did the solvers do it so quickly? Winning solvers almost always (72.5%) built on pre-existing solutions to other problems - 55% depending on their own previous work, and 60% on previous work of others.   The authors do not discuss the overlap between the results of the “Who?” and the “How?”, but clearly the two results indicate that many of the problems were solved by taking common knowledge and/or techniques across domain boundaries.

The importance of this kind of transfer of knowledge from one domain to another is well known  in studies of creativity.  For example, Csikszentmihalyi notes (p.88) “An intellectual problem is not restricted to a particular domain.  Indeed, some of the most creative breakthroughs occur when an idea that works well in one domain gets grafted to another and revitalizes it.”  Thus InnoCentive has found an interesting, and apparently quite effective, way to stimulate that important form of creativity that depend on transfer of information across domains.

From the description of the research project, it sounds like the authors could ask - and answer- several other interesting questions. I mention the first above: for those winning solvers who came from a domain outside of the problem domain, what percentage built their solutions on pre-existing solutions, and the related question of what percentage of solvers from the domain of the problem used pre-existing solutions?  The geographic locations of the solvers would also be interesting to know - are they, for example, predominantly from locations where there is oversupply and consequently underutilization of trained scientific personnel?  If so, this could provide interesting data on ways to outsource research to such areas.

In any case, if we view knowledge and creativity from the perspective of a network, it seems that InnoCentive has found a way to link previously unlinked elements.  It certainly provides useful links across knowledge domain boundaries- and might be a means to provide links across unlinked geographic boundaries as well.  Pretty exciting stuff!