It seems like I get at least one call a week from a former student or business associate who tells me that their local entrepreneurship incubator is failing. I think that I know why. About a year ago I visited the Cambridge Innovation Center, otherwise known as CIC, which houses over 600 entrepreneurs, most feverishly working at their computers on their new startups. As a simple exercise, I would pick one at random and sit down and ask: "So, what's your new business?"

The answers were mostly the same. It generally sounded like "it's a new ap or website for blah, blah, blah." I would then ask: "who's the customer?" And invariably it would be aimed at a person "just like me" (from the perspective of the entrepreneur). My interpretation: They were all making "web caca" aimed at the worst possible customer (one with no money) and most were going to fail. In fact, most do fail. Their business models are no better than lottery tickets, but with a lot more work.

As part of my teaching, I started to emphasize that "Customer Selection" and identifying a B2B as opposed to a B2C business opportunity was a much better place to start.

Yet, most incubators are filled with B2C businesses that are likely to fail. So what's there to do about it?

I've created a new concept in incubation which I call an "Innovation Partnership." It has three constituents: The entrepreneurs (who are likely to be technically oriented or at least competent), real mid sized businesses that are non-tech based, and a group of entrepreneurial mentors (and educators) whose job is to guide the entrepreneurs through a process.

The process is: 1. Learn the business model of one of the "sponsor" companies. 2. Apply your technical knowledge to creating a business model innovation that can help the company survive or grow. 3. Pitch it to the company and see if they will buy it. 4. If they buy, then, and only then, make it.

I am working with MAUA, an engineering school in Sao Paulo Brazil on building an Innovation Partnership Center, the first of its kind and if it's successful, I'll let you know.