By Thomas R. Schori, Ph.D., and Michael L. Garee, Principals, Millennium Marketing Research, 808 E. Ironwood, Normal, IL 61761-5239. Tel. 309-532-8466 -
With the strong customer service emphasis today at many companies, selecting a research methodology to accurately identify which, if any, customer service problems to "fix" has become of paramount importance. In this column we would like to present an approach that we’ve used quite successfully over the years to do precisely that.
A common research approach to customer service problem identification¾and one which we feel has inherent design flaws¾is to present a "laundry list" of possible (or even probable) customer service problems and ask respondents to indicate which of the problems they may have encountered when dealing with the client company. While, ostensibly, this approach may seem reasonable, we believe that it also has the tendency to unduly "lead" respondents, which can result in misleading survey results. In other words, we take the position that if one goes "looking for trouble," it’s a pretty sure bet that one will indeed find it!
In our opinions, a better, more reliable approach to identifying customer service problem areas is to have respondents tell us, "top of mind," which specific problems they may have encountered when dealing with the client company. Since this approach utilizes more of an "open-end" questionnaire design, admittedly, it is both more difficult to prepare as well as to complete. Nonetheless, we have found that the little bit of extra effort required¾both on the part of the researcher and the survey respondent¾is certainly well worth it. We think you will find that as well. Here’s how the approach works:
Identify customer ‘interaction’ points
Every business has discrete customer "interaction" points. We therefore first identify these points and then group the survey questions under them. For example, let’s suppose that the company for which we’re conducting the research is an auto repair facility. Typical customer interaction points probably could be expected to include the following: The initial contact (either in person or by telephone) with the repair facility to schedule a time and day for repair; contact with the Customer Service Representative upon arriving at the facility for the actual repairs; contact with the billing department when paying for the repairs; and any other miscellaneous contacts that might have been necessary during the visit, e.g., conversation with the service manager, the mechanic, et al.
Typically, each of these interaction points becomes a major heading on the survey questionnaire and pertinent individual questions are then grouped under each of these headings. For example, under the heading Initial contact with our repair facility, we might pose the following types of questions:
1. Overall, how satisfied were you with how you were treated when you initially contacted us to arrange for your repairs? (The scale might range from "1," Very Dissatisfied, say, to "5," Very Satisfied.)
2. Did you encounter any specific problems during this initial contact that you would like to tell us about? If you did encounter problems, please indicate below, in your own words, the nature of the problem or problems. _________________________________________________________________________________________
3. Again, in your own words, tell us what you think we could have (or should have) done to avoid this problem? _________________________________________________________________________________________
Obviously, the approach we use goes from the general, i.e., level of overall satisfaction, to the specific, i.e., identification of specific customer service problems, if any, as well as what the customer feels could have (or should have) been done to avoid the problem(s) in the first place. However, please note that, if the respondent experienced no problems under any of the interaction points (major category headings) he or she could easily complete the survey in a rather short time. If, on the other hand, problems were encountered, it’s been our experience that the respondent usually is so eager to tell someone about them that the extra time it will take to complete the "open-ended" questions will be insignificant to them. Plus, allowing respondents the opportunity to tell about any problems they may have encountered "in their own words" usually proves far more valuable than forcing them to "fit" their complaints into neat (and usually somewhat vague) categories that are pre-defined by the researcher.
Tabulating, coding results
Now, the sharp-eyed research might point out that tabulating and coding all the open-end data could easily turn in to quite a chore. However, we haven’t found this to be the case at all because the overwhelming majority of survey respondents will not have experienced problems, and will therefore not complete the open-end questions. (If the vast majority of respondents did have complaints, the chore of tabulating and coding open-end survey results would certainly be the least of the client company’s concerns!)
Pinpointing what to "fix"
Unlike the "laundry list" approach mentioned above, which tends to "identify" many, many problems that actually may or may not need to be "fixed," we have found that the approach we’ve outlined here allows the client company to very specifically pinpoint bona fide problem areas. And, with the cost inherent in fixing some customer service problem areas, a company certainly will want to focus their efforts and their dollars only on those areas that are truly problem areas.