Enhance Productivity and Efficiency with Stephen’s Innovation Insights

Innovation Insights by Stephen Shapiro

A magazine asked me to write a book review of Innovation Tournaments by Christian Terwiesch and Karl Ulrich. The book arrived in the mail yesterday and I immediately turned to the index to see if InnoCentive was listed. Sure enough, we are mentioned in several places in the book.

This got me thinking: Is InnoCentive a tournament?

The word tournament is derived from the French word for “medieval sport” and is now used to describe a wide variety of competitions.

Most competitions/tournaments are quite entertaining. And by their very nature, there is always a winner. One could argue that tournaments are “spectacles designed to find a champion.”

Given this widely held point-of-view, using the word tournament as a descriptor of InnoCentive seems to be inaccurate.

The NCAA basketball championships are a tournament. The “World Series of Poker” is a tournament. American Idol is a tournament. With each of these, there is always a winner. The purpose of the tournament is to find that winner while (usually) providing entertainment value.

InnoCentive is not interested in finding a winner for the sake of naming the champion. The objective is to find workable solutions to real business problems. Their approach is one I call a “contingency-based, value-driven pricing model.” Admittedly, that does not sound as sexy as calling it an innovation tournament.

Here’s how it works. A company has a problem they want solved. They decide the “value” of finding a workable solution and they offer a “bounty” to anyone who can provide one. The bounty is only paid when they get what they need. This “pay for solution” model outsources the risk associated with complex problem solving.

Here are other examples that illustrate the key difference between the bounty-based approach with the tournament-based approach.

The NetFlix Prize was not a tournament.  They only paid the team that improved the recommendation engine by 10%.  This makes is a bounty-based approach. You only pay the bounty when you get a successful solution.

In contrast, The Cisco iPrize, can be thought of as a tournament. According to their website, they will “select up to 32 semifinalist teams that will work with Cisco experts to build a business plan and presentation… Up to eight finalist teams will present their business ideas to a judging panel to compete for the grand prize: a $250,000 award shared equally by members of the winning team.”  The LG Electronics competition (read my article on it here) was also a tournament-based approach.

The key difference is the way the challenge is articulated.  With the bounty-based approach, the success criteria is clearly defined and you know if someone provided a successful solution:  Did you improve the recommendation engine by 10%?  Did you find a chemical compound that has specific properties?  Did you develop a mathematical model that optimizes solves a specific problem?  The “winner” of the bounty is determined by this success criteria.  If the criteria is not met, the bounty is not paid.

With the tournament-based approach, the success criteria is not defined.  The winner is the “best” of the submissions.  Although these types of competitions can yield excellent solutions, I know from inside-information that the results are often less than stellar.  One company that uses this type of tournament described the results as a “PR success yet a commercial failure.”

Both approaches can provide value to any organization.  It’s just important to recognize that they are useful in different ways.  Tournaments can be great to get a broad set of ideas for an undefined space.  Bounties are great for when you are hunting down usable solutions.

  1. Hi Steve

    Interesting post. I have read the book and I am currently putting a tournament together inside O2. I think that the “tournament” moniker is just a snazzy title, but the method has some merit. Essentially, it seeks to set up the “best” selection process in a statistical sense. This is the best that one can do when the initial spread of ideas is wide and the search lacks enough specificity to run a bounty-based challenge. I see this as complimentary to the Innocentive model, which the authors appear to have misunderstood.

    The tournament does not produce winners, but a shortlist that has a high statistical chance of containing winners, given various unknowns. Users of the method should be clear that this statistical selection is only statistically valid within the confines of the tournament, not within the wider context of actual execution. As you say, the ideas might not be champion ideas at all once executed.

    Therefore, it is critical to apply further filters to the process, such as the application of business modelling and various other tools to evaluate the ideas shortlist. I would say that a bounty-based challenge would be a follow-on tool.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Bring Stephen’s innovation insights to your next event!