@singhns – Narinder Singh
Recently I met up with Professor Kevin Boudreau of London Business School to discuss his research on topcoder. Along with Professor Karim Lakhani (Harvard Business School) and their students they have produced dozens of papers analyzing what happens on topcoder and I wanted to understand why. An excellent video exists of how they first encountered topcoder, but Professor Boudreau’s comments on how topcoder was changing research possibilities in innovation because it was “like the fruit fly” sparked my interest.
Fruit flies (or specifically Drosophila melanogaster) transformed genetic research. In 1910 future Nobel Prize winner Thomas Hunt Morgan began using them in genetics research at Columbia University. One of the key attributes of the fruit fly was that a generation was approximately ten days. This allowed rapid studies of genetic inheritance and gene mapping in unique ways and led to many breakthroughs in biology.
In attempting to study technology, software and innovation in business today we struggle first with the relative newness of the field. Its not like we can go back to Adam Smith to get anything but the most general of principles. In addition, looking at what companies are doing in the “real world” is nearly impossible – as its difficult to isolate the attributes that drive outcomes, comparisons are challenging, failures are seldom openly shared, and the time horizon between action and results is too long. So we see much of the ‘research’ consists of anecdotes and is often done within industry.
The most academically (i.e. rigorous, peer reviewed, etc.) studied area of innovation in software is open source. Because everything is public you have real and objective data based on contributors, community size, rate of change, growth of usage, forking, etc.
As a research vehicle, topcoder provides the best of both of these worlds and more. With thousands of challenges running each year there is a natural data set for how work is setup, pricing, level of participation, characteristics of participants, etc. In addition, the platform of topcoder allows the setup of experiments that fit within the overall context but test various aspects of how innovation works. This has allowed for experiment driven research through topcoder on collaborative versus competitive markets, behaviors of individuals in development, impacts of disclosing work on innovation, tournament design and much much more (over 700 citations). Topcoder, like fruit flies, allows generations of innovation studies to be done rapidly in a controlled environment – and it works on real world problems.
While we are most excited by what the community creates and how our members build their own skills, we believe in the power of research and value of sharing. So we welcome the partnership with academia and will continue to do more to open up our data for our members and the world to do interesting things with. Its nearly impossible to get to the level of impact of the fruit fly, but we’ll do our best!
PS the challenges are the fruit flies, not the members of the topcoder community!