I asked for a topic today on Twitter and friend Christian suggested I post on managing failure. I've been thinking about it and I think it comes down to degrees. Failure relies on two things to determine magnatude: how much it costs to fail and how much time it takes to fail. In managing failure the key is to define success and define the mini-successes if you fail to meet the overall goal.
For instance, if you make a social game and set the bar of success at a million daily active users (DAU, a that would be a smashing success). On the way to the release you: do it quickly & ahead of schedule, learn about scaling servers to accommodate players. The game comes out and after the perscribed time period you're only at 200,000 DAU - Did you fail?
One the one had yes, you didn't reach the initial goal. But if you managed the process and mark all the successes of the journey (scheduling, scaling) there are huges win to point to. Additionally, if you completed the project at half the budget would you consider it a success? Was the work done quickly ahead of schedule? Is that success?
One of the best things we can do to manage these systems is to create parameters within our work. Avoid a rolling deadline - set it! We have two weeks.... that's it! Or set budgets that will cost this much - not a dollar more, not a dollar less. By setting rules we can manage expectations for our failures. It's similar to when you go to a casino. We say to ourselves, I'm spending only $100, its spent. We will not gamble with any additional funds. Using this priciple for work, we control for the degree of failure.
The key in my view, is to make sure you measure your success in terms of degrees within the two dimensions of speed and cost. If you can fail fast that's ok. If you fail cheap, that's even better. The ultimate - fast and cheap! Be beware of slow and expensive! That will destroy you.
How do you manage failure or account for it?
No comments:
Post a Comment