Copyright Society of Management Accountants of Canada Sep 1993Customer satisfaction studies are increasingly being used as the basis for performance evaluations, bonuses, and awards. As a result, the temptation to manipulate and influence study results has never been greater. And because human opinions are of such a delicate and indistinct nature, even a few creative adjustments here and there can have a strong impact on the final results of a customer opinion survey.
What are some of these creative ways that managers have found to influence their customer opinion results?
1. Let the employees who are being evaluated collect the survey results.
This is a sure way to dramatically influence the results. In one case, account representatives being held accountable for survey scores visited the customer's office and asked the survey questions directly, filling in the customer's answers on the survey form. When someone in the company finally questioned this approach, a parallel telephone survey conducted b an independent firm uncovered dramatically lower results. In another example, a hotel chain had its employees slip a survey under the doors of guest rooms. Employees quickly figured out that they were only hurting themselves by giving surveys to guests who had problems at the hotel.
2. Re-contact customers who have been sued and encourage them to alter their previous answers.
One of the more insidious forms of cheating, this approach involves looking through customers' answers, contacting those who gave a less-than-desirable rating, and gently badgering them into giving a revised evaluation. During the course of follow-up conversations, customers are encouraged to state that they were "satisfied" after all.
3. Exclude potentially unhappy customers from the list of those to be surveyed.
If it is known beforehand that some customers will give low survey scores, a reasonable explanation can sometimes be constructed for excluding them from the survey. The technique was used to its fullest extent in a program where detailed listings of customer answers were relayed to company managers in the field each month. For the following week, these managers--the same ones held accountable for customer scores--were allowed to specify exclusions of individual customers from the database before the results were ever tabulated.
4. Avoid asking questions that might make you look bad.
If you know that customer answers to certain questions will put you in a bad light, you can probably dream up a few reasons why those questions should not be included in the survey. In one example of this form of cheating, a senior manager lobbied for excluding a question that asked if customers thought the firm was "a well-run company." This question clearly reflected directly on his own performance, which was not sterling. In another example, an electric utility wanted to exclude any questions about price. Of course, it so happens this company's prices were relatively high.
5. Ask leading questions.
Almost anyone who has been surveyed before--which nowadays includes almost all of us--has encountered leading questions. A common example is the use of an answer rating scale that makes it fairly easy to get high scores. By asking for a yes or no answer to a question such as, "Were you satisfied with the service," you can almost guarantee very high results. Most customers will have to be very unhappy before they will intentionally get someone in trouble by saying they were not satisfied.
6. When all else fails, shoot the messenger.
If you dig deeply enough into any research survey project, you will eventually find some nuance of the program that can be called into question. This is almost inevitable, since any researcher will tell you that, alongside the statistical marvels and mathematical wizardry they perform, much of what researchers do is highly subjective and based on personal opinion.
How can cheating on customer satisfaction surveys be allowed to happen? The answers are many, but most often they involve some form of organizational politics. Whoever has the management muscle to get his or her way in the company can usually construct reasonable sounding justifications for changes in the survey program.
In any case, if your company has or is considering implementing a customer satisfaction measurement program, the previous examples should give you some idea of how these programs can be manipulated. While the vast majority of customer survey programs are honest by design, there will always be a few individuals who will go to extraordinary lengths to attain higher survey results. And in the never-ending corporate struggle for truth, justice, and higher customer scores, this temptation to cheat is sometimes overwhelming.
Joseph E. Toole is director, client services with Walker: Customer Satisfaction Measurements, Indianapolis, Indiana. The editor of this column is Sarah Kennedy, president, the Canadian Productivity Network, Mississauga, Ontario.
My Research