Six Sigma Case Studies : Improving Average Handle Time (AHT) for Online Chat

Industry: Consumer electronics and high tech

Background

The client is a Fortune 500 technology corporation that develops, manufactures, sells, and supports personal computers and other computer-related products. The client has outsourced part of its online chat support services to TELUS International.

The challenge

For the client, longer Average Handle Times (AHT) meant higher chat abandonment rate by customers, increased head count requirements (agents) to serve customers, and overall decreased program profitability due to lost sales opportunities. A Six Sigma Green Belt project was implemented to identify variables leading to higher handle times.

In the Define Phase, the team checked the historical performance using Histogram and Run Charts to check the capability of the process. In the Measure and Analyze Phases, the team used Multi-vari and Pareto Analyses to identify the types of calls with higher handle times. It was discovered that the online sales chat queue was handling several non-sales inquires as well as payment via chat (PVC) issues that were causing high AHT metrics.

Boxplot and Two Sample T-tests were used to verify if the current queuing process using Level 1 agents and Level 2 affected AHT. Test results showed no significant difference in the performance between Level 1 and Level 2 support. Process Value Analysis was also used to identify process steps that are “value adding” versus “not value adding.” A comparison of the top performer process compared to bottom performer process in terms of handle time was also conducted. From this, the team determined that there were processes that could be reduced and/or eliminated completely.

Solution

For the Improve Phase, the team designed and piloted a new process model. Based on the pilot run, the new model was found to be effective and efficient. It also did not affect the client’s financial and CSAT metrics. The model was later presented to the client and was approved and deployed immediately across the floor. The team also revised queuing selection process for Level 1 vs. Level 2 agents.

Results

A significant improvement on the program’s AHT was observed:

• The account’s AHT dropped to 14.92 minutes (beyond project goal of 17 minutes) from baseline of 21 minutes when first deployed
• Weekly AHT stabilized and stayed within the control limits
• Agent headcount for online chat was reduced with agents able to handle multiple chats simultaneously within AHT metrics
• Over $28,000 in savings accumulated in less than 2 months after starting the Six Sigma project