Tech startups are always an experiment

Last fall, I taught entrepreneurship for only the second time since I got to SJSU. (I’d previously taught it as an adjunct at UCI, building more on my experience as an entrepreneur than any formal preparation).

About 2/3 of the way through the semester, as part of making a point about the lack of sufifcient information in any startup situation, I wrote on the board

Business is an experiment

which will become the mantra for all my future entrepreneurship courses.

To explain my thinking further:

• if you’re doing something new whether new to the firm or new to the world — the answer doesn’t exist;
• in a startup, for many questions you won’t have the time (or money) to get a definitive answer; so
• you need to make and implement a decision with the recognition that the experiment may fail — and thus both look for signs of failure and find an expedient way to mitigate against (rather than prevent) such failure.

That leads me to an an interesting quote from Carol Bartz (newly appointed Yahoo CEO). It came from her 2001 talk at Stanford about encouraging entrepreneurship in established companies: at that time she was CEO of Autodesk, the CAD software company.

As an aside, I’d disagree with her use of the term “entrepreneurship”: if you use the term to refer to any form of innovation or business initiative, then the term doesn’t have any real meaning. While “corporate entrepreneurship” is a bit of an oxymoron, at least it demarks a form of initiative distinct from starting a new company.

However, I couldn’t agree more with her on the philosophy she brought to Autodesk to deal with the turbulent dot-com era, a philosophy she called “Fast Fail Forward.” As transcribed by Scott Fulton of Beta News:

I had to do this during the dot-com time, where everybody panicked and decided that you guys [Stanford students] were going to rule the world. … [P]eople got even more cemented in and scared to take risks, because what did it mean in an established company?

So we started this thing called 'fail-fast-forward,' and the whole idea is, listen, failure is very acceptable. When it happens, make sure you identify it quickly, and hopefully it's in a forward motion. And then start going again.

In other words, nothing new happens without risk, and the only way to deal with risk is to accept it rather than try to completely prevent it.

Hat tip: to Scott M. Fulton, III of BetaNews.