How to learn what your customers want, Part II

Written by Deena Waisberg

You’ve created a customer survey, sent it out and the respondents have sent in their answers. Now you’re eager to see the results. But to make sure your analysis is meaningful, you have to take certain steps.

Richard Pridham, president of The Agili-T Group, a Montreal-based consultancy specializing in managing customer relationships, offers tips for analyzing customer survey data.

Kill bad data

Discard incomplete surveys and duplication, which you’ll get when a respondent presses the send button more than once on a web-based or e-mail survey.

Use sophisticated software

Most low-end survey software offers reporting with graphs and cross-tabulation. However, Pridham advises against using these analysis tools. To really analyze data, he says, you’ll need a statistics program like SPSS ( or Statistica (

Find an interpreter

Assign someone within your organization who understands fundamental statistical analysis to interpret the results. If no one in your firm has these skills, then hire a company that offers analytics services.

Bias the feedback

It sounds counter-intuitive, but you shouldn’t treat each customer’s feedback equally. Why? Some customers are more important to the profitability of your business than others. In fact, it’s often 20% of your customers that are responsible for 80% of your business. Take the financial contributions of each customer into account when looking at the feedback and deciding what criticisms need to be addressed.

Commit to action

Once you’ve analyzed the data, use what you’ve learned to make improvements. Identify the root cause of problems, not just the surface issues. “No survey is worth anything unless action comes from it,” says Pridham. Finally, tell your customers — by letter, e-mail or through the corporate website — what you’ve learned and what improvements you are implementing.

Read other pointers on How To contribute to your business success!

© 2003 Deena Waisberg

Originally appeared on