Improve first. Measure, plan, manage, define and standardise later.
Lately I've been listening in awe to Troy Du Moulin and facilitator Chris Dancy on Pink Practitioner Radio. I wish I had been listening from the start: now I'm digging into the back chapters. (Here's the RSS).
It's fabulous stuff but I found that it's not - for me - perfect. I actually disagreed with Troy! which doesn't happen very often. I was listening to Episode 24 on Problem and Availability Management, where Troy said something like "if its not defined you can't control it and if you can't control it you can't manage it and if you cant manage it you can't improve it", justifying the CMM maturity progression. This is a bit like that old saw about "you can't manage what you can't measure". I've ranted against that before. In this case if you can't define and measure and manage you can't improve, which I can't agree with either.
The very first thing we should do is improve. CSI should come at the start where we identify those things that it is intuitively obvious we can get on with improving right away. Improvement begins now. I've said it before: CMM is back to front.
In the same podcast Troy mentions "good, better, best". We agree there: it doesn't have to be perfect from the start. Let's apply that to principle to how we do improvement:
- start with rough obvious improvement
- then manage and measure the practice to identify better improvement. These will be rough (good) measures.
- then standardise (i.e. define) the practice and make any process repeatable. This gives best planning input for improvement and best measurement.
There, that's CMM backwards.
Waiting to measure anything is a bad idea: get some metrics now. Measurement is like sex: when it is good it is great and when it is bad it is still better than nothing. (There goes another 10% of readers). Bad metrics are good incentives: those people being measured will work hard to improve the data quality and create better metrics.
And waiting to improve is just nuts. Ask the people doing the practice: they'll soon tell you some areas that can be improved right now. I don't need any fancy instrumentation or planning to find that out. ITIL might have misguidedly made CSI the fifth book but actually reading the book will tell you CSI comes at the start. Of course the CSI book launches into all sorts of navel gazing which is just more delay. If you want to do some baselines and assessments and who knows what, go ahead. But get on and fix something in the meantime. Then use all that planning to make your improving better.
No doubt we will - as always - debate this until we end up agreeing it is a disconnect in terminology, a matter of semantics and we are actually saying the same things. But I really really dislike the messages that are sent by saying you can't manage what you can't measure, and you can't improve what you aren't managing. It comes through as well in CMM saying that level 5 is Optimising: back to front.
Get measuring and get improving: start both now.