Six Sigma


Kristen Huber

Kristen is a Tokyo-based budding analyst with an interest in project management

Let's use DMAIC to solve a problem concerning the quality of calls in a call center.


Define: We want to raise the customer satisfaction rating to 90 percent.

We will establish a team to use Six Sigma methodology to reach this goal.

Six Sigma


Measure: Using descriptive and exploratory data analysis to collect data on the problem. Above is a fishbone diagram which may be created during a brainstorming session. I did some analysis of call center data downloaded from this website. I found that the mean number of calls is 19.2, the mean number of answered calls is 9.58 and abandoned calls is 9.63 for ever 15 minute interval.

Analyze: Hypothesis testing is used to test the hypothesis that low call quality is related to the high number of abandoned calls. 100 customers that have called the call center recently are randomly selected for a survey. The results show a positive corrolation between customers that have had their phone call recently abandoned and low customer service rating.

Improve: If we want to improve the situation at the call center we need to look at what times the call center is flooded with calls. 21 percent calls are made within the hours of 8 and 10 (likely people taking a mid-morning break) and the largest percentage (almost a quarter!) are taken from 12 to 14 (lunch break of course.)


To improve, let's increase the number of call center staff from 10 to 14, and check back in a few weeks to see how the situation has changed.


Control: A survey a few weeks later shows the customers satisfaction with the call center has greatly improved. To be careful though, a survey should continue to be held every few weeks to be sure no future problems arise.