Carnegie Mellon students to exploit social networks to predict success of Internet startups

Opinion
Mar 23, 20101 min

Venture firm teams with CMU to ferret out the next Google

Carnegie Mellon University has joined forces with Silicon Valley VC Charles Moldow on an interesting student business competition that had as its aim coming up with the best way to predict an Internet startup’s success.

The winning team, announced on March 20, earned $2,000 for their idea of blending social networking with the Delphi Method of forecasting to spot the next Google or Facebook. The contest drew students from various CMU schools, including business and computer science.

Also read: Want a job? Get a computer science degree

“We thought it would really be great if we could create an XYZ that would help us decide where to make our investments,” Moldow said.

The Delphi Method is an interactive forecasting technique involving panels anonymously answering questionnaires in an effort by researchers to find a consensus of opinion. Social networks would enable formation of a very large virtual panel, and an algorithm would do the dirty work of sorting through the results and spotting the most reliable forecasters from the bunch.

“Thus, the most successful forecasters would emerge organically from the crowd and could be ‘harvested’ for their insight for a number of potential applications,” the team wrote in its proposal.