Six Sigma Case Study
Statistical Survey Analysis for Improvement
This Six Sigma case study looks at how a call center used survey results to find out what actions they should take.
Most companies use surveys to tell them how they are performing on certain customer satisfaction metrics. This call center, however, used statistical techniques to analyze the results to find conclusive answers on the actions that needed to be taken to increase customer satisfaction.
In this six sigma case study, we look at how a call center that wanted to get more knowledge out of their customer satisfaction surveys.
For years, customer satisfaction surveys were sent out to thousands of customers after they had called in to the center. For the most part, the results from the surveys were averaged out and monthly customer satisfaction levels were broadcast to all departments.
The client wanted a survey from which they could get critical information that would point to optimizing customer satisfaction.
A team was formed to revamp the customer satisfaction survey in order to attain conclusive information on how to increase customer satisfaction.
First of all, the team created a Level 0 process map and identified the possible inputs that may affect the output of customer satisfaction. By the time the map was finished, there were over 90 inputs which were identified!
To include questions and ratings on all the identified inputs would make the survey too long and possibly discourage customers from completing the survey. So the team decided to use a C&E matrix to prioritize the inputs. It was decided that the top 20 inputs from the C&E matrix would be used in the survey.
At the end of the survey, there were two main Yes / No questions. The first was, "Did your call have a proper resolution?" The second was, "Were you satisfied with your call?"
The answers to these questions would tell us definitively what the customer thought of the call on two fronts - First Call Resolution and Customer Satisfaction. All the ratings on the characteristics and inputs in the first 20 questions of the survey could tell us how they affected the outcome of the two main outputs.
The team went ahead and conducted the surveys in their new format. The results were collected over three months.
The team then used Logistic Regression to analyze the results and find the factors that have a significant impact on the outputs.
Some conclusions were expected. For example, length of call and issue resolution were the biggest contributors to customer satisfaction.
But some conclusions were unexpected. For example, the agent's accent / voice clarity did not have any significant effect on customer satisfaction for over 90% of call types. However, there were 2 call types where this was a major contributor to customer satisfaction.
The analysis from the new surveys allowed us to prioritize where our focus should be in order to increase customer satisfaction. This project gave us the direction to create at least 4 more focused projects that would have a great impact in improving customer satisfaction.
The analysis also helped the HR department tremendously in deciding what funds to allocate to what type of training according to its impact on customer satisfaction.
As we all know, customer satisfaction is a metric that takes time for actions to have an impact on. For the first six months after preliminary improvement actions were taken, customer satisfaction consistently averaged approximately 5% higher than before.
Once the spawned projects are completed, this should increase further.
--End of Six Sigma Case Study--
Leave this six sigma case study and go back to Main Case Studies page.