Satisfied customers are the best benchmark for the success of a support center, right? Customer satisfaction is certainly a better gauge of effectiveness than pure operational metrics such as average speed of answer or even support rep’s productivity, but it’s also harder to measure. The gold standard for customer satisfaction is to use a so-called transactional survey, one that is done right after a support case is resolved. Is your survey working? Let’s find out….

Test #1: do you get a high return rate?
Regardless of any other considerations, surveys that don’t get answered are worthless. Small return rates (say, less than 20%) mean that you are not capturing a true picture of the customer base. Typically very unhappy and very happy customers respond to surveys. If you’re hitting 20% and above, you’re capturing the all-important middle ground.

To increase your return rate, make the survey as short and to the point as you can. In most circumstances, a handful of ratings are all you need: ask the customer to rate the professionalism of the support rep, the speed of resolution, and the overall experience. Asking more questions won’t give you more data; it will actually turn customers off.

Test #2: do you sample properly?
Many support centers do not survey each and every customer coming through the support center because the volume would be huge (and repeat customers may object). Fine, just make sure that your sampling techniques are fair and adequately random.

Along the same lines, ensure that each support rep gets enough survey results (here’s your response rate dilemma again) to make evaluations possible.

Test #3: do you move fast?
Transaction-based surveys work when customers have a chance to remember the transaction in the first place. Send the survey immediately after the event.

Avoid snail-mail surveys. Email or phone is more immediate.

Test #4: Do you use the survey to do service recovery?
When using yearly surveys, I’ve always been frustrated that, by the time I got the feedback, I couldn’t do anything to turn the customer around on a bad experience. With a transactional survey, you get feedback quickly enough to address issues on the spot, so take advantage of it. Yes, it does require that you quickly attend to survey results rather than letting them pile up until it’s time to compute this month’s score.

Responding personally to very good and very poor surveys also has the nice side effect of increasing the response rate for the survey. Once customers know you read the surveys, they are more likely to respond to them.

Test #5: do you use the survey to coach?
Transactional surveys give you specific feedback on specific cases, hence specific individuals. While one great survey doesn’t mean that the support rep is perfect, and one bad survey doesn’t mean he’s terrible, coaching from survey results is very effective in showing reps where good habits – and bad – translate into specific customer experiences.

Like surveys, coaching is best served hot: don’t wait too long between receiving the customer feedback and using it with the rep.

Test #6: do you spend reasonable amounts of money on the survey?
Surveys don’t need to break the bank. In the technology field at least, email surveys are very effective and remarkably inexpensive. If you are using phone surveys, which are very effective at getting answers but very expensive, reassess your priorities.

Customer satisfaction surveys are an essential component of support metrics, both operationally and strategically. Make sure you can rely on yours.

About the author
Francoise Tourniaire is the founder and principal of FT Works, a consulting firm that helps technology companies create and grow their support operations. She is the author of “Best Practices for Support Metrics”, a practical guide to selecting and using metrics in support centers. For more information, visit www.ftworks.com or call 650 559 9826.