Just as the Human Genome Project aimed to crack the human code, the Startup Genome Project hopes to be able to drill down into some of the details of what it's calling the "innovation code."

The project has launched a survey to investigate "the science of startups," lessons from which it plans to incorporate into blackbox, a new seed accelerator that was created with the merger of techVenture, Cofounder Network, Founders First and Startup School.

"Despite the supreme economic importance of scalable startups, we still don't understand the patterns of successful creation," say the project organizers. "More than 90% of startups fail, due predominantly to self-destruction rather than competition. For the less than 10% of startups that do succeed, most encounter a handful of near death experiences along the way."

Linking this discovery to the way in which Pandora broke down music into small parts in order to dissect people's musical tastes, the Startup Genome Project hopes to "organize principles, tactics, and methodologies in relation to specific factors. When we define the factors these things depend on we can begin creating a system to manage the chaos and complexity of organizing the information about what makes startups successful. While the common wisdom seems to be that the number of factors is infinite or indefinable, our initial set of variables has brought a lot of clarity."

The Startup Genome Project aims to measure a number of factors that may contribute to startups' success or failure, including design, customer development, market. It's a daunting project, and it's hard to imagine that something like "success" can be quantified, but the results from survey participants should be interesting nonetheless.