Saturday, May 26, 2007

How Do you Know if You are Improving or Getting Worse?

It is critical to measure, analyze, and summarize the performance of processes that create your service and/or product.

A cultural issue, it can't be one person or one department responsible for measurement. It is a shared responsibility from the CEO/Owner/Plant Manager to the front line operator.


There are two ways to look at this: You can decide in advance what will and will not be measured. Or, you can measure "everything" and, over time, cut back to the vital few. Let the data tell you what should and should not be measured.

Measuring performance applies to any business. Great example: If you watch Saturday Night Live, you know Darrell Hammond as the guy who impersonates seemingly everyone from Bill Clinton to Jesse Jackson. He has been on the late night icon for twelve years, the longest stint of any one performer in the show's history.

Hammond also performs in comedy clubs. In a recent article, he talked about setting a goal of ten laughs per minute in his stand up routine. A tape recorder measures his performance and allows him to grade his process for delivery comedy. The analysis allows him to:
1. Measure the number of laughs per minute versus the goal of ten.
2. Learn which jokes, impersonations, etc. garner the most laughs.


Both help him continuously improve his process for delivering comedy.

3 comments:

Anonymous said...

Stephen,

This may be one of the best examples I have seen for illustrating the quality process. It goes directly to the central, very human, issues involved. There is much less chance that the listener or reader will get the impression that things are too technical or too mathematical to apply to the everyday world that they live in.

Great post!

Anonymous said...

Yes at some level too much measurements can only cause more confusion. Most people are not "number" oriented. So when providing data, it must be kept at a simple yet to the point of why it must be measured.
I have one question: which is more significant- measuring scrap dollars or rework dollars?

Stephen said...

I appreciate these great comments.

Both touch on a very important point: The involvement of humans. Speaking with data is very important but the language must be clearly understood by all involved. A prime example is knowing there is a difference between practical and statistical significance. Those obsessed with collecting data often throw a blind eye to the practicalities of managing a process.

To the point "too much measurement can cause confusion": I agree to a point but feel that good and well managed measurement actually eliminates confusion. Speaking objectively (versus subjectively) must be the language of the business. I don't think you need a graph for every point you make but do think objective evidence is vital for making critical decisions.

I agree that some people are not number oriented. Some people prefer pictures. So graphing data might be a better approach. I strongly urge companies to lead their people down a path to be more objective and less subjective.

Scrap or rework: First, I'll state the obvious; both are forms of waste and must be eliminated (or significantly reduced) to become lean. To me, scrap is the target. Scrap is lost $ where rework does not involve a total loss.

I'll go one step more and say don't separate the two. I see it as one category: Waste due to Defects. You should have a pareto diagram of your top defect categories in terms of dollars or quantities rejected. This pareto analysis should drive your efforts.