What does a Google product manager, a law professor, a New Yorker staff writer and a division VP of Best Buy have in common? (Hint: this is not a joke) They all believe in the power of prediction markets. They were the attendees at a McKinsey hosted round table discussion that took place in Dubai, UAE.
Prediction markets are basically markets created for the purpose of making predictions, where a final reward (cash or otherwise) is tied to the occurrence of a particular future event (i.e. likelihood of a company project being delivered on time, or the estimated % sales increase for next quarter).
What makes prediction markets to appealing to companies, is their ability to create a forum for ‘everyone’ to weigh in their true opinions on the probability of a particular event. Historically, opinions/predictions like this were left to experts and division heads. However, they may not always have all of the information; they may not be aware of important influencers; they may be inherently biased, or apprehensive to give unfavorable opinions to senior management. All of these aspects influence their accuracy in predicting an event (future sales, success of a new product’s release, etc).
However, prediction markets give equal voice to everyone, anonymously, regardless of rank or role. The aggregate forecast of all then gives the ‘group opinion’ on the future event. Google and Best Buy are both using this. Best Buy wanted to know what people thought the sales generated for their gift cards would be for Feb 2005. They gave the same data to everyone to consider in their predictions. As it turned out, the aggregate estimate was 99.5% accurate. The paid expert forecasters were off by 5 percentage points. Proving that ‘crowd opinion’ was more accurate.
Giving an equal voice to everyone is also what blogging and social media is about. It is about removing corporate biase and formal publishing filters, and allowing the masses to comment and offer their true, often anonymous opinion.
We can all imagine thousands of uses for this, especially in the middle east, where distance between decision makers and those closest to the end markets can be great. Needless to say, this is catching on as management sees the benefits of accuracy and honesty. A couple interesting takeaways from Google’s experience with prediction markets:
· employees’ geographical vicinity (region, office, and how close your desks are) creates bias
· the closer you are to the CEO, the more disadvantaged you are for accuracy
· crowds get smarter and more accurate over time and experience
· market optimism (i.e. and increasing stock value) result in overly optimistic forecasts. The same works for market pessimism.
see full round table discussion