Skip to content
Customer Satisfaction Measurement | Part 3 of 6

A Questionnaire That Gets Answered

Thirteen questions, five minutes, and a fieldwork plan that brings the response rate to 50 percent in industrial business.

Every department wants its own question. The marketing department wants brand perception, engineering wants a rating of the drawing exchange, and the managing director wants a free text field for ideas. The result is forty questions that take twenty minutes, and a response rate of 15 percent from the friendliest contacts only.

The example company allowed itself thirteen questions and five minutes: eight criteria rated from 1 to 5, one overall satisfaction question, one recommendation question from 0 to 10, one choice of the three most important topics, and two open questions.

Find your eight criteria

The criteria come from what your customers already talk about. The team read the complaints, the supplier scorecards, the visit reports and the lost quotations of the last two years, grouped the topics along the order process and kept what a customer can judge from experience: delivery reliability, product quality, complaint handling, lead time, information and communication, price for performance, technical support and quotation speed.

Eight is a limit, not a target. A ninth criterion costs more answers than it brings insight. Three of the eight were offered with a no experience option, because not every contact has had a complaint, a design question or a quotation in the last year. Without that option you get polite guesses instead of data: complaint handling in the example was rated by 30 of 50 contacts, and those 30 are the ones who know.

Five rules for the wording

  • One topic per question. Quality and delivery in one sentence produce an answer you cannot use.
  • Ask about experience, not about opinions on you. “How satisfied are you with our delivery on the confirmed date?” works. “Do you think we are a reliable partner?” does not.
  • No leading words. Drop “fast”, “excellent” and “reliable” from your own questions.
  • Offer a no experience option wherever a contact may have none.
  • Let the customer set the weight.

The last rule deserves a sentence of its own. When customers rate the importance of every topic from 1 to 5, almost everything lands between 4 and 5 and you learn nothing. Asking which three of the eight topics matter most forces a choice. The share of contacts who name a topic becomes its importance, and because everybody names three of eight, the natural dividing line sits at 37.5 percent.

The recommendation question runs from 0, not at all, to 10, certainly. Answers of 9 and 10 count as recommenders, 0 to 6 as critics, and the score is the share of recommenders minus the share of critics, in points. It is easy to follow over the waves and it explains nothing on its own, so it is never reported without the criteria.

Use the same scale for all ratings and name every point: 1 very dissatisfied, 2 dissatisfied, 3 neither, 4 satisfied, 5 very satisfied. Then test the questionnaire with two colleagues and one friendly customer contact, with a stopwatch. More than seven minutes, or one question that needs explaining, means another round of editing.

The fieldwork makes the response rate

The tool does not bring answers. People do.

Day What happens
-5 Account managers announce the invitation to their class A contacts
0 Personal invitation from the managing director with purpose, length and closing date
7 First reminder, only to those who have not answered
10 Response checked by class, calls to class A customers without an answer
14 Second and last reminder
21 Survey closes, thank-you note with the date of the feedback

Fifty of 100 contacts answered in the first wave and 60 in the second, because by then customers had seen that the first one led to action. The invitation names the month in which they will hear the results, and that promise is the one part of the plan you cannot miss.

Newsletter

One Lean idea.
Every two weeks.