Iterations

Mike Bria (mbria) had an tweet the other day about iterations and why we prefer them. Iterations provide two primary benefits over non-iterative development. In no particular order:

• Planning adjustment
• Customer feedback

It's important not to forget the latter. We iterate because we think giving the customer access to the software every week or two weeks or month provides the customer the opportunity to give the developers feedback on value. The more access to the software the customer has, the better the customer is able to steer the software towards greatest value.

Mike, I believe since tweets are naturally short, believes that perhaps teams are focusing too much on the former benefit. Perhaps even getting a little draconian with tracking velocity and stories and tasks and everything we can think of to make sure the plan is on track. The thing about plans is that they are never perfect. George Patton is said to have said "A good plan violently executed today is far and away better than the perfect plan executed tomorrow." That's because tomorrow, the plan won't be perfect anyway.

The point is you should not try and build a perfect schedule. The more time you spend tracking story hours and task minutes, the more time you are not communicating, developing, or doing any other the other things you need to do to pull value through the value chain.

So iterate, but iterate to increase value, not make the perfect schedule.