Example Prediction Market for IT Projects

A colleague of mine forwarded me this great research paper on an example internal prediction market for an IT project. The research is not fully complete, but there were a few interesting nuggets that support the usage of internal markets for accurate predictions This is the topic that Bo Cowgill from Google will be presenting at DIG next month.

The research paper highlights 4 key needs for an accurate prediction market

1. Ability to aggregate information and knowledge from individuals
2. Incentives to encourage active participation
3. Feedback to participants based on market prices
4. Anonymous trading

The results from the case study were quite positive. Acxiom Corporation was the test case and used the Inkling Markets software to host the market to predict 26 milestone events of an internal IT project. Two results jumped out at me. The market was 92% accurate on the milestone events (24 for 26) and had an 87% participation rate (33 participants). There was also a higher perceived level of collaboration as a project team, which had positive impact on the outcome.

The authors of the paper are Herbert Remidez, Jr. Ph.D. and Curtis Joslin from the University of Arkansas. Looking forward to seeing further output from the research.

Prediction Markets Explained

Last week, Mark Lorence posted a topic for discussion on prediction markets and questioned if it would have been an effective tool for British Airways in the opening of the T5 terminal in London. I wanted to provide a quick follow up, as the topic of prediction markets was in the NY Times yesterday with references to different organizations that are using them to predict sales forecasts, store openings and project completion dates.

If you have an interest in the software and services companies who provide platforms to facilitate a prediction market, you should check out Inkling Markets, Consensus Point and NewsFutures. Each company provides the technology and infrastructure to setup and manage prediction markets. To give you a sense of the type of problems that a prediction market can be used for take a look at the Inkling Markets home page. They have examples such as "Improve Forecasting of Performance Indicators", "Expose Quality Problems" and "Predict Risk in Your Supply Chain". All real issues that many organizations need to address.

If you want to see how a prediction market may work for something like sports, politics or weather, I would recommend taking a look at TradeSports.com. You could argue this is simply a gambling website since the trading is performed with real money, but if you look a little closer you will find some interesting information. For example, according to the current "2008 Democratic Presidential Nomination", the current trading price for Barrack Obama is 85.5, which means that 85.5% of the market predicts that he will be the democratic nominee. In other words, the "market" strongly believes he will be the nominee. If I were to purchase options at this price and he were nominated, I would earn $14.5 per block of shares ($100 - $85.5) I purchased. If he did not receive the nomination, I would lose my entire investment since the value would be $0 when the market expires, which should be August 28th at the DNC.

Bo Cowgill from Google, who is a speaker at the DIG 2008 conference, runs Google's internal prediction markets. He is also an expert on the topic and operates his own blog on the topic of prediction markets. Also take a look at Midas Oracle. We are looking forward to hear Google's story from Bo at the conference.

One interesting side note is that there is a market on TradeSports.com for Google's Lunar X project and if it will be won on/before 12/31/2012. The current price is 24.3.

R. Todd Stephens, Bo Cowgill and Euan Semple to Speak on Enterprise 2.0 at DIG 2008

I am pleased to announce that R. Todd Stephens, Bo Cowgill and Euan Semple will all be speaking on the topic of Enterprise 2.0 at DIG 2008. Each will be presenting case studies from organizations where they have adopted E2.0 for competitive advantage.

R. Todd Stephens, who is a Senior Technical Architect at AT&T, will discuss AT&T's success with adopting collaborative and social technology. R. Todd's case example will present a 3-tiered program to adopt social platforms and how AT&T has shown business value as adoption grows within the organization. R. Todd has also been actively participating on the Talk DIG blog, which we like to see!

Bo Cowgill, an Economist at Google, will be speaking about Google's use of internal prediction markets. Google has the largest corporate implementation of prediction markets, which they utilize to capture the "wisdom of crowds". If you aren't familiar with the topic of mass collaboration, James Surowiecki wrote the book Wisdom of Crowds, which makes the case that in the proper setting, large groups of people can make better decisions than individual experts. Google is using markets internally to identify winning project ideas based on how employees trade options on the success or failure of a project.

In our final case study Euan Semple, a social computing advisor, will discuss how the BBC has established in an internal social network to improve efficiency and increase innovation. Euan will discuss what tools the BBC leverages to improve collaboration and communication that impacts the corporate culture, individual behaviors and provide the ability to get things done! The BBC has leveraged this platform to even extend the conversation beyond the four walls of the organization and make everyone a marketer for the BBC. Euan also blogs on the topic of social networks and technology at the aptly named "The Obvious?".

We are excited to have R. Todd, Bo and Euan speak about their experiences in knowledge in the area of E2.0.