Mercury's Blog

I recently found out about a new competitor in corporate prediction markets: CrowdClarity.

I’m a little partial to these guys, mainly because they come from my alma mater, the University of Michigan. The key people look to be a mix of entrepreneurial students and professionals. The company itself was started a year and a half ago, but has had success with early pilot projects.

In fact, three slides included in their “Learn more” online slideshow are quite powerful statements as to why prediction markets can be useful. They were predicting car sales in the winter of 2008/2009, during what was one of the most volatile months the industry has seen in many, many years. And the prediction market beat the internal forecast made at the beginning of the month, and the expert forecast made at the end of the month.

Here are the slides:

To recap, the prediction market beat the official GM forecast (made at the beginning of the month) easily, which isn’t hugely surprising considering the myopic nature of internal forecasting. But the prediction market also beat the Edmunds.com forecast. This is particularly interesting, as Edmunds would have had the opportunity to review almost the entire month’s news and data before making their forecast at the end of the month.

Examining the numbers

Let’s quickly quantify this error. Assume an average Chevrolet sells for $18,000. After dealer markup, assume that GM/Chevrolet receives $16,000 per vehicle.

Within the first week of November 2008, the prediction market would have warned Chevrolet that they were going to miss their revenue targets by $800 million in the Chevrolet division alone. And depending upon the exact product mix, this could have easily exceeded $1billion.

Now Chris can blather on about corporate prediction markets, but he’s simply wrong. Assume that even with three weeks’ early warning Chevrolet was only able to save 10% of that gap, it’s still $80million in savings. Even if a corporate prediction market for a giant company like GM cost $200,000 a year, that would still be a return on investment of 40,000 %. And again, that’s in the Chevrolet division alone. (Note: It would be a rare prediction market that cost $200k/year to run.)

Now not every problem should be solved by a prediction market. This is where management expertise comes in: are the errors large enough to warrant the cost of reducing those errors? But big problems with big numbers are often very suitable to address with a prediction market.

Summary

I’d like to wish good luck to the CrowdClarity team. It’s great to see Wolverine entrepreneurs working on prediction markets. There are more and more players each year in the corporate prediction market scene, but with case studies like this behind their belts, they’ll be well-placed to pick up some business.

I know it’s been a bit quiet here on Mercury’s Blog for a few months. Over the past year I’ve been studying at Cambridge University for my MBA, and this summer I undertook a research project around Y Combinator.

Y Combinator is a very unique seed accelerator program. Startup companies get ~$15-20k in funding in return for ~6% equity, three months of intensive business and product advice, connections to mentors and potential advisors. Inkling Markets was in the second cohort of Y Combinator funded companies, for example.

My hypothesis was that a lot of people/organizations are starting seed accelerators without really examining the full scope of innovations they need to think about in order to achieve their goals. I wanted to take the opportunity to look into why entrepreneurs choose to go into a seed accelerator, why individuals choose to start a seed accelerator, and then propose a framework for designing new programs.

You can click here to view the post on my personal blog that has a bit longer executive summary. But I’ve embedded the paper and data sources below. I hope you find it interesting.

Copying Y Combinator

Appendix A – List of Seed Accelerators

Click here to view the list of seed accelerators. Only seed accelerator programs are listed; see main paper for details.

Appendix B – Example Seed Accelerator financial model

Appendix C – List of all companies founded by Seed Accelerators

OR

Click here to view the list via Google Docs.