Tuesday, August 5, 2008

Prediction Markets Speed Innovation

When I wrote the post on Web 2.0 tools last week I noted with some satisfaction that I was familiar--through use or writing or both--with all the tools listed in the McKinsey chart except predictive markets. I thought that concept had something to do with finance and stock markets, and that was the initial focus. This paper seems to do a good job of summarizing the economic foundations of the concept. However, the article made it clear that the applications now extend beyond financial markets. I thought I ought to find out what the management and marketing implications were.

Still looking for the roots of the concept, I found a paper on an experiment at Google in what the economists called “information aggregation and revelation.” That was getting closer. This paper gives a number of interesting applications of prediction markets. One I thought was particularly interesting was the Iowa Electronic Markets, a University of Iowa project that conducts markets on elections as well as business topics like interest rates and corporate earnings. There’s lots of information on the site and you can register to participate if you like. Quoting another study, the Google paper comments that, “In the political domain, Berg, Forsythe, Nelson and Reitz (2001)summarize the evidence from the Iowa Electronic Markets, documenting that the market has both yielded very accurate predictions and also outperformed largescale polling organizations.” See pages 6 through 10 for their discussion of the accuracy of prediction markets. It’s impressive and certainly argues for “the wisdom of crowds.”

The authoritative article in the general media is a New York Times piece from this April. It discusses a variety of uses by major corporations from GE to Best Buy. I was especially taken by the InterContinental Hotels example. I know why it attracted me. All my work in customer service and database marketing makes clear that most companies do not tap the knowledge of their lower-level personnel in ways that can improve their business. The IHG project tapped their 1,000 technology staff in order to develop and prioritize ideas. They used software called IdeaPagent from the consultancy NewsFutures. Two hundred employees participated anonymously, submitting ideas that ranged from spectacularly awful to the basis for new corporate projects. They seem pleased with their experience and interested in other applications.

There are other fascinating examples in the NYT article, worth reading in its entirety. Prediction markets is not a new idea. However, it is one that seems to have been facilitated by Web 2.0 software tools. It’s not news to any of us that the pace of change has speeded up and we need to tap all possible sources of innovation. Prediction markets represent one way to speed successful innovations to market.

No comments: