Be Careful What You Measure

I was reminded this week of the important lesson to be careful what you measure, measurements drive action and behavior.

The productivity numbers have dropped slightly at the first of the year and not rebounded despite all the continuous improvement efforts. Needless to say, this was not the expected result. Management wanted to know what caused the drop and take the proper countermeasures to reverse the trend.

To add to the complexity, several changes occurred during this time frame regarding the usual suspects of volume, mix, lot size, manpower, new products, methods, etc. There did not appear to be an easy answer to the simple question, “What changed?”

When I observed the processes at gemba, the process flow appeared to be improved yet the numbers told a different story. As one that likes to use data to drive decisions, I dug a little deeper. With a little investigation, I started to analysis the historical data to see if I could determine any patterns or potential causes. Unfortunately, only summary reports were available and the details behind the summary were thrown out every month. Not a bad policy since I was the first to ever request to see this data. The good news is the 5S for this area looks great.

Turning my attention to the spreadsheet used to create the productivity charts, looking for anything abnormal, I noticed that several of the formulas did not make sense to me. Looking closer, I saw that one of the formulas in this year’s spreadsheet pointed to cell in last year’s spreadsheet which turned out to be a blank cell. Apparently, in the process of creating the new 2007 spreadsheet from the old 2006 spreadsheet, the formula did not get properly translated.

A couple of keystrokes later, the formula was corrected. Followed by a weeks worth of clerical effort to restate the year, the productivity measures displayed a more accurate and improved operation.

Although this recalculation of productivity had a positive affect, it is not what I would consider a triumph. Ongoing efforts are still required to truly increase productivity, so it’s back to gemba. However, I am modifying the lesson to “Be careful what and how you measure, measurements drive action and behavior”.