Anne Truitt Zelenka

Marshall Kirkpatrick interviewed Paul Graham of Y Combinator who notes a move toward smaller scale startups:

There’s definitely a trend toward smaller investments, because it costs so much less to start a startup now. And if you take less money initially, you keep more options open. Once you’ve taken a VC-scale investment of two or three million dollars, you give up the option of an early acquisition.

Paul further advises would-be entrepreneurs not to try to figure out everything about their business in advance:

What I tell founders is not to sweat the business model too much at first.

Paul’s approach may be an easy target for ridicule, but it makes good sense, and not just from the standpoint of keeping the early acquisition option open. Trying to figure out everything about a business up front is akin to the waterfall model for software development. It likely works about as well for business creation as it does for building large enterprise systems, that is, not very well.

One of the best ways to learn is by doing, not by thinking. In software development, the waterfall approach has been largely displaced by more experimental approaches like incremental prototyping and agile development. Both those approaches decrease the scale of successive releases so as to find out more about what really works in less time.

In other domains, out-of-the-box thinkers are seeing how small-scale, fast feedback experiments can quickly test many promising ideas. I wrote last week about psychologist Seth Robert’s outrageous weight control approach, developed via self-experimentation. Experimenting on one’s self is about as small scale as human psychological experiments can get.

You don’t have to have all the answers before you try something. Trying is how you get answers.