Tuesday, September 01, 2009

Simplify, or the problem with Six Sigma

I was on a conference call last night with the other members of the PMI Agile Community's steering committee. Last week we had a successful launch event at Agile 2009 in Chicago and we were discussing how we continue to move forward.

We were discussing vision and values and principles and finally one of the members called "in the grass" meaning our discussion was getting to detailed. As another member stated, we should be focusing on short term plans, not some long term vision when it's hard to figure out what any of us will be doing a year from now.

At this point, I recall a conversation I had with David Allen, author of Getting Things Done. He said we tend to want to focus on where we feel the pain. If our boat just hit the rocks, we don't think about our ship's vision, we figure out how to fix the hole.

To me, this is one of the challenges with Six Sigma. Taking a traditional DMAIC approach takes some time, 6 months is not uncommon for a DMAIC project. There's a lot of analysis that happens before any changes are made. I prefer a Lean approach. Go in, make a quick assessment, improve on the process, and then take another quick look at where to go next. I've implemented new business processes in as little as 3 months with only 1-2 weeks focused on analysis.

With my Agile Community it's even simpler since we don't have any processes in place yet. We need to try some stuff out and see what works. That's not to say a vision isn't important, it helps us make the right decisions. However, now that we're getting members, we need to think about what we have to offer them so they find value in our group. A year from now we can conduct our retrospective, figure out what works and what doesn't, and adjust our vision then.


Calvin.K said...

The times spend on because of the methodology itself but the project. DMAIC can be done in 2 months or less if the project is small, and lean "quick improvement" might also take few months if the project involved is big.

Actually it is not say that lean or six sigma which to be used but which is suitable for your project.

Basically, lean doesnt help much on quality related improve, and six sigma dont really care about the cost invested.

Lean six sigma run in parallel is better actually

Bob Tarne said...

Calvin - You're right. I guess what I've seen in some cases is people taking a project that could be done in 2 months and stretching it out to 6 months by doing way to much analysis.