TQMI


QUALITY WAY

   
 
Through holistic and systemic approach . . .
 
trans

Newsletter from TQM International

Vol. 9 : No. 2 - Sep 2009
18th Year of Service to Society
trans
Six Sigma Project Case Banking-Customer Care-Call Center

By Mr. S.P.Chokkalingam

Six sigma Head. TQMI
Background:

Customer satisfaction survey conducted by multinational bank in India showed one the low score item was for Interactive Voice Response System (IVR).

The bank deployed Six Sigma project to improve the score. When the team started with the project, they got into a great dilemma. They did not understand what is the metric that they will improve as customer satisfaction score is lag measure and even if they improve the IVR processes, it would be difficult to validate the results as the next scheduled customer satisfaction survey was four months away. It may be a costly affair to conduct an additional customer satisfaction survey for this alone.

Problem defition:

They decide to interview some customers informally to capture the customer voice and understand the problem. Some customers said that they are not able to get the information they require. They often disconnect without getting the information they want. Based on this the team decided on a surrogate measure as CTQ, i.e. call abandonment rate. They found out of 70,000 calls attempted 14,000 calls were abandoned in one week and sigma level was 2.3.

Goal:

After some of discussions and benchmarking against competition, it was agreed that the goal should be no more than 5% abandoned calls.

Analysis:

Team members randomly talked to some customers to identify about 50 probable causes. They prioritized the causes and arrived at following 5 high impact causes:

  • IVR does not respond to inputs some times
  • Incompatibility between switch and some land line exchanges
  • Call gets disconnected due to poor coverage at cells
  • Customer gets confused
  • Customers not familiar with use of IVR

“Quality depends on good data. It also depends on executive leadership in using that data.”

Team stratified the data by land line and mobile callers and found statistically significant difference between land line and mobile through Chi-square test. Further stratification revealed there is no significant difference between exchanges or between cells. But the significant cause for land line is ‘No response’ and mobile is ‘Confused’.

The team discussed what could be the reason for “No response”. One of the team members who participated in the telephonic survey said that most of the “No response’ cases were from PCOs. They decided to visit some PCOs and test themselves. They also took some technical persons with them. The interesting thing that they found in the PCOs was that most PCOs were still using dial type instruments. The technical expert pointed out that these phones dial in the pulse mode and the IVR system does not respond to pulse as it needs tone dialing. They have now validated one of the root causes of the problem.

The second highest problem in both cases was “Customer confused’. They again talked to some more customers why they are getting confused. Is the language not clear? What is the root cause of the customer getting confused? During the interviews, a large number of customers said that some of the menus have too many options. By the time IVR completes all the options, the customer forgets the options and get confused. The team decided to test this does not need validation.

Solutions:

‘No response’ could be solved by providing a pulse to tone converter. But it is costly option. Pulse phones will be replaced over a period of time in any case. It was decided to select another possible solution. The system does not recognize pulses. So if the system does not receive any tone dialing for specified time, the call is automatically transferred to the Call Centre executive.

The second cause of too many menu options was tackled by analyzing the frequency of usage of different options by the customers and preparing Pareto diagram and retaining only highly used options and every menu was restricted to three options.

For the purpose of unfamiliar customers, a booklet explaining in a simple language with pictures was planned.

Results:

The abandon rate came down to less than the target of 5% after implementation of the solutions. Next customer satisfaction survey the scores for IVR moved up appreciably.

“Excellent things are rear.”