Sunday, July 22, 2012

Data, Speak to Me!

Working with one team, a project was progressing into the definition stage. The team was trying to create an objective for the administrative process. They were concerned that they weren't meeting a 14-day deadline very often. The demand for this process was expected to triple as the company grew and they might have to add staff.

The first step of the process seemed to take the longest. Some thought it was averaging a week, leaving only a week for the other three departments to finish the process. They started to outline the steps and talking about the issues and the problems that created the delay.The team started to brainstorm some solutions.  They had some technology answers to the problems. They had some suggestions for suppliers without any ability to influence them.

The Lean experts looked at cycle times and wanted to cut the waste. The Six Sigma experts wanted to look at errors and causes for rework.

No one had taken time to look at the data. The Lean guys were concerned about a cycle time that in worst case would account for one-fourth of a person's time, even at triple the demand. The Six Sigma guys hadn't looked at the histogram, or paid attention to the frequency in processing times.

Once the data was analyzed, people figured out that half of the items were processed in the same day. Only 1 in 4 exceeded a three day processing time. Instead of averaging a week, the first step was averaging a day and a half. 19 of 20 were completed faster than a week. With a bit more data, the team learned that those that took more than 1 day were due to supplier issues, but were easily addressed. Even with problems, the delays were not that big. 

The team did succeed in shrinking the overall range through some creative solutions. But they hadn't started out with quite the mountain to turn into a mole hill. The data showed them that the solutions could be very focused, a relatively inexpensive compared to the ones they brainstormed in the beginning.

No comments:

Post a Comment