I am working with a client who has a whole continent full of different “tribes”, all of whom are going agile and all of whom have huge change agendas on at the moment.

So I am working on a change management plan to help the organisation roll out consistent approaches to IT development and to demonstrate continued improvement in speed and quality across multiple development teams for multiple internal and external clients (the tribes I mentioned).

I think I have a cunning plan but it will be interesting to see how it plays out in reality.  Rather than rolling out a structured change program I am recommending we let all the cowboys run wild and do whatever change they want.

Then instead of driving change, the central group responsible will define the rules for experiments so that:

• When a group want to try a new idea they must first report in that they are doing so.  They do not need permission but need to state the experiment they are performing, why (which can be as simple as “make life easier for ourselves), when the first phase will happen and most importantly – what they predict will happen.
• Then at the end of each phase we will check whether the outcome was as expected.  If it is then we will ask why and then see how we can share the knowledge more widely.  If the result was better or worse than expected then we will ask why and see what we can change or roll out more widely.

Doing this will help our central team focus on some core activities which we can also treat as experiments.  It will also allow individual tribes to move at a pace (and in a direction) that suits them rather than waiting for us to catch up.  Of course it might also lead to chaos, lack of focus and inertia.

Give me a call in a month or so (or add a comment here) if you want to learn how it turns out.  Or give me a call now if you think you know how it will turn out and want to warn me.