The Index of Consumer Satisfaction and Corporate Marketing Policy


Robert W. Pratt, Jr. (1972) ,"The Index of Consumer Satisfaction and Corporate Marketing Policy", in SV - Proceedings of the Third Annual Conference of the Association for Consumer Research, eds. M. Venkatesan, Chicago, IL : Association for Consumer Research, Pages: 742-745.

Proceedings of the Third Annual Conference of the Association for Consumer Research, 1972      Pages 742-745


Robert W. Pratt, Jr., General Electric Company

[Manager, Business Research & Forecasting Operation, Major Appliance Business Group, General Electric Company.]

As I understand it, the Index of Consumer Satisfaction (ICS) is being developed to supplement "cost measures" provided by the Consumer Price Index with a "benefit measure"--that is, a "subjective" indicator of consumer satisfaction. To quote Dr. Pfaff, the ICS "tries to determine the benefit that consumers receive--or the degree of satisfaction they experience--from the operation of the market."

Consumer satisfaction has been a basic concern of the business community for a long time. In a real sense, it's what we sell. And I believe business will be warmly receptive to both methodological and substantive insights generated by this work. For at least the last twenty years, there has been a wide recognition among researchers and planners in industry, government and the academic that we must move beyond economic and demographic variables and develop the capability to identify and measure sociological and psychological variables relevant to a particular issue. Witness the recent proliferation of work in the general area of social indicators.

My assigned role this morning is to comment briefly on potential implications of the ICS for formulation of corporate marketing policy--not public policy, not broad business policy, but marketing policy

While the ICS, in the form just described, may have immediate implications in the public policy area, I foresee difficulty in using it as an input to marketing policy. As I proceed to explain why, it will be useful if you picture yourselves as members of a marketing policy group within a single company. In general, you have available to you information about the consumer environment, the industry or industries in which you compete, competition, and, of course, your own products. The extent of this information varies greatly by product and industry. Your task is to make decisions that will optimize the market position of goods and services offered by your Company. Now, why do I feel that the ICS would have limited value in accomplishing this task?

First, consider the difference between the ICS and the Consumer Price Index. The distinction between an index structured using dollars as a common base and an index structured using measures of "satisfaction" as a base is an important one. In general, when making either tactical or strategic decisions, management will have available a great deal of extremely detailed information about costs and prices, and companies can and do take a wide variety of actions based on this information.

But I just don't believe that the Index of Consumer Satisfaction, in its present form, will offer comparable inputs for decision making. Although there may be a few exceptions to this, basic input data used to construct the ICS, regardless of how it is massaged and analyzed, is not sufficiently specific for use by any one company in the formulation of marketing policy. Let me use the appliance area as an example. The question asked on the ICS pilot study regarding satisfaction with household appliances was this:

"And what about the appliances you have, that is such things as your refrigerator, stove, and toaster, on the whole, how satisfied are you with your appliances?"

Answers were recorded on a seven point scale ranging from "very satisfied" to "not at all satisfied."

Now, wearing your marketing manager's hat, suppose the index is declining, indicating that consumers are growing less satisfied with their household appliances. This knowledge alone does not provide a basis for adjusting policies or programs. The management of a single company must have much greater detail before making decisions to change products or marketing programs. Perhaps I can clarify this point by describing one satisfaction/dissatisfaction study recently completed by General Electric. This will give you an idea of the level of detail required by management for decision purposes.

Over the past six years, we have conducted personal interviews with a national probability sample of 1,000 households each month. For four of those years, respondents have been asked the following question about each appliance available to them from among a list of thirty.

"Thinking about your (NAME OF PRODUCT), considering everything about it, which statement best describes your overall satisfaction or dissatisfaction with it?"

After considerable experimentation, we decided to use a five-point scale ranging from "extremely satisfied" to "extremely dissatisfied." For the first two years of this study, an open-ended "why" question was asked following every response. For reasons discussed below, the "why" question was later asked only for dissatisfactions. Over the four years, 48,000 respondents were questioned about 665,000 individual appliances. Here are some observations based on findings from this research.

1.  The mean level of reported satisfaction across all appliances and all brands was 94 percent.

2.  There was considerable variation in level of satisfaction across products and brands; also, level of satisfaction varied by such characteristics as age of appliance, time since most recent repair experience, satisfaction with that repair experience, and so on. In addition, as you might expect, the satisfaction level for a particular product or model frequently varies over time.

3.  We found that respondents had a very difficult time articulating their reasons for satisfaction. Since, when acquiring a product, the buyer expects that he will be satisfied, the "why are you satisfied" question elicited such responses as: "It works," "It does what it's suppose to do," and so on. We eventually stopped asking reasons for satisfaction simply because the answers had no value for either operational or strategic decisions. Nothing was being learned that could be translated into marketing policy.

4.  For an individual company to make decisions that will increase the benefits provided by its products to end consumers - and, hence, presumably increase their satisfaction- the company must have information specific to its products and models. Similar information about competitive offerings as well as a national Index of Consumer Satisfaction would, of course, represent potentially valuable norms. To realize this potential, marketing research would have to insure comparability between its measures and external measures, such as the ICS--for example, by using identical questions.

5.  The information we have found useful for decision purposes is that associated with reasons for dissatisfaction, rather than reasons for satisfaction. For example, statements associated with breakdowns or repairs accounted for a high percentage of reasons given for dissatisfaction with all brands of television sets just a few years ago. Breakdowns and/or repairs accounted for a much lower percentage of all dissatisfactions for major appliances and housewares. By far the dominant reason given for dissatisfaction with TV repairs was the fact that the repairman was not able to complete the job on one visit, whether because he was not qualified, did not have the right parts, or whatever. (Contrary to what you might think, for all appliances, cost of repair accounted for only 14 percent of the reasons for dissatisfaction.) Within GE, actions were initiated to improve both training of servicemen and availability of parts on repair trucks. One result is that the overall level of dissatisfaction with our repair performance has declined and the mix of reasons for dissatisfaction is changing. Incidentally, reasons for dissatisfaction are quite different for new as compared to older appliances.

6.  In a number of instances, expressed dissatisfaction had nothing to do with the appliance asked about; rather, reasons referred to features and/or performance characteristics that the respondent would like to have. For example: "I'm dissatisfied with my refrigerator because it is not 'frost free'." This type of answer apparently reflects a form of "psychological obsolescence" with the present unit.

7.  We learned that answer patterns are extremely sensitive to question wording. For example, if respondents are asked if there is anything at all disliked about an appliance, the recorded level of dissatisfaction jumps. Presumably one could design questions that would virtually eliminate expressed dissatisfaction. Our questions, of course, were designed to provide a level of detail needed for in-Company decision purposes.

8.  Finally, in view of Martin Pfaff's earlier reference to studies that indicate a low correlation between job satisfaction and job performance, another finding that may be of interest to you is this: Respondents who said that they were dissatisfied with a specific GE appliance and also that they had no intention of replacing that appliance at the time of the interview, and who were found on a reinterview to have replaced the appliance, were just as likely to have purchased GE as similar respondents who initially said they were satisfied with the brand. While I believe we have explained this to our own satisfaction, the explanation is too lengthy to review here.

The essential point I have tried to make is that inputs to marketing decisions, if the results of these decisions are to prove truly beneficial to end consumers, must be relatively specific. This information will not be forthcoming from the ICS as presently constituted, if my understanding of the index is correct.

Let me make a few additional observations regarding the ICS that might serve as a basis for discussion.

1.  First, only 342 respondents of the 574 sampled respondents completed the full set of questions, a loss of 40 percent. It would seem that any set of questions developed for use with a national sample, for purposes of developing a representative national index, would have to be simplified so that they could be answered by a larger percentage of designated respondents.

2.  Second, the effort so far has paralleled as closely as possible the Consumer Price Index in both concept and execution. Given the difference in the nature of what is being measured, I would ask whether or not this is necessary or desirable?

3.  Third, for some products, respondents were asked about the importance of various product attributes in their purchase decision. A great deal of research has been done that suggests that interpretation of answers to questions of this type must be made with great care. For understanding purchase decisions, at least for consumer durables, I have become increasingly convinced that we come closest to actionable results when we have available panel data that can be analyzed using longitudinal techniques.

To conclude. In these brief remarks, I have attempted to focus on the potential for application of the Consumer Satisfaction Index to the formulation of marketing policy. Extensive reference to one GE study has been made simply to convey the level of detail management wants as an input to policy formulation. I have intentionally avoided reference to what I think are interesting and provocative issues associated with the ICS, such as the question of the extent to which government should be held responsible, or assume responsibility, for the "happiness" of citizens. While interesting, this type of question didn't seem to fit my assignment here. I have also avoided talking about specific methodological detail. Such discussion could go on all day and probably much longer, and I see no particular purpose to be served by making scattered observation.

I believe results from empirical research of the type described this morning will eventually become necessary inputs to policy making in both the public and the private sectors. The present effort is a well conceived and scholarly piece of research, and I hope the project will be continued.

As I stated earlier, measures of satisfaction/dissatisfaction have been of paramount importance to marketers for a long time. If they have not already done so, in the process of refining the design of the Index of Consumer Satisfaction, I would urge that its architects consider availing themselves of some of the knowledge that has been gained in this area over a considerable period of years.



Robert W. Pratt, Jr., General Electric Company


SV - Proceedings of the Third Annual Conference of the Association for Consumer Research | 1972

Share Proceeding

Featured papers

See More


Semantic Processes in Memory-Based Consumer Decision Making

Sudeep Bhatia, University of Pennsylvania, USA

Read More


Feature A Benefactor or A Victim? How Charity Appeals with Different Protagonist Foci Affect Donation Behavior

Bingqing (Miranda) Yin, University of Kansas, USA
Jin Seok Pyone, University of Kansas, USA

Read More


Emotion, Scientific Reasoning, and Judgments of Scientific Evidence

Caitlin Drummond, University of Michigan, USA
Baruch Fischhoff, Carnegie Mellon University, USA

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.