How should we measure the "success" of social innovations?

This week’s readings showed a number of ways to view Negroponte’s work with the One Laptop Per Child project. In the Forbes article “The Prophet of Cheap”, Negroponte describes two camps of how people see his work; “those who will say we promised to build a $100 laptop and didn’t” and “those who will say we put a stake in the ground and created a bloody netbook industry”. Two additional views were described in the Scientific American article “Frugal Innovation: India Plans to Distribute Low-Cost Handheld Computers to Students”; there are those who believe that “even if Negroponte had been successful in meeting the $100 target, the Indian government could not have afforded OLPC laptops for the huge student population”, and others believe “a new device by itself will not help overcome more fundamental challenges such as lack of resources and inadequate teaching”.

With such varied opinions about the true impact of the OLPC project, I was led to question what would make this project successful, and how we should measure the amount of “social good” created by social innovations. Should we measure the success of social innovations by whether or not they meet their pre-stated goals, the extent to which they create something radically different and challenge the status quo, or the real impact it has on those in need?