Lessons from Agility

Agile development - and now I mean agility in general, not just a specific agile software development method - received lots of attention lately and while many are promoting and practicing it some can simply not accept that we can't plan as well as we could do before. Whenever I talk to the opponents of agile software development I always start with admitting that agility is not an ideal software development method. At least it isn't idea for the developer who has to suffer from uncertainties all the time, but it is the best for end users who mainly enjoy its benefits, meaning quick improvements of features and usability.

Facebook founder Zuckerberg's recent letter to users about changes in access control teach us at least two important lessons agility.

1. The initial access control policy did not scale, it had to be changed: especially in social media we have no way of knowing how a feature or a policy that works well with 50 000 users will work with 10 million. We have to try it and then adapt.

2. Discussion of the new policy started back in July. Being agile - and quick - with features doesn't mean that you can be very quick with your community. In a way agility is capped at the top speed of your community.