One of the really interesting problems is scaling up pilot programs to large scale. There's no end of successful pilot programs. What's been found however, is that when you try to scale a pilot program, who's members are usually made up of highly motivated/qualified specialists, and try to incorporate people who aren't invested in the program/who aren't specialists (often because there simply aren't enough to populate large-scale programs), many pilot programs turn into failures.

It makes policy decisions, even at the state level, very difficult.

In addition, the best "do-ers" to execute pilot programs aren't always (perhaps even are rarely) good educators for the next round of specialists to scale up the program. They'd rather be on the ground than writing documents about best practices. So a lot of institutional knowledge from the pilot programs are lost.

There's probably a good opportunity for an app/service/startup that makes recording this institutional knowledge more seamless for the on-the-ground "do-ers," but I'm not sure what the business model would be.

> There's probably a good opportunity for an app/service/startup that makes recording this institutional knowledge more seamless for the on-the-ground "do-ers," but I'm not sure what the business model would be.

Most government social programs take bids for outside companies to do technical assistance or manage the grant's data. Get an office on K Street in Washington D.C. and bid on that part of the grant.