Wednesday, June 10, 2009

Ten Thoughts on what matters at a startup: Failure

Why does failure matter at a startup?

You will fail. What's important to understand is what failure costs and what you learn from it.

You may fail at a small scale by inadvertantly forwarding a stupid email to a customer or by missing an important meeting that you didn't put in your calendar. There's not a lot of cost here, though if you do these things repeatedly it will certainly add up. Aim to not repeat them too often.

You may fail at a larger scale by not delivering a project on time or by spending too much on a major purchase without having researched it. There's more cost here, but it's still not fatal. It does get more important to avoid future repetitions.

You may fail at a fatal scale by blowing through your cash without regard to revenue. Learning from this failure won't help in the short term, but the experience may benefit you at your next startup.

How do you learn from your failures? Start by embracing failure. For example, do a project post-mortem on failed projects and ask hard questions about what went wrong and why. Keep in mind that it's not about blame, but about supporting each other and learning as a team. Of course, you need to be sure that you understand what success looks like too.

Another way to embrace failure is to "Fail fast, fail cheap, fail often." That is, use inexpensive prototypes to quickly try out your product ideas and discover not just the ones that fail, but also the ones that show promise. Keep the prototypes small and focused, and do as many as needed to clarify questions about your products.

None of this is to say that failure is better than success. Success is definitely better. If you're going to fail, though, learn from it and control the cost.

• • •

This is one in a short series of posts called Ten Thoughts on what matters at a startup. The thoughts started life as a presentation I made at VeloCity residence at the University of Waterloo. While they're far from definitive, and aren't a top ten, they've mattered to me in my software startup experience.

No comments:

Post a Comment