RESEARCH SUPPORT UNIVERSITY OF MICHIGAN BUSINESS SCHOOL MAY 1996 CUSTOMERS ARE DIFFERENT: SATISFYING ALL TYPES IS A CHALLENGE WORKING PAPER #9690-04 BARBARA EVERITT BRYANT UNIVERSITY OP MICHIGAN BUSINESS SCHOOL

I Customers Are Different: Satisfying All Types is a Challenge Barbara Everitt Bryant University of Michigan Business School Annual Quality Congress Chicago, Illinois May 13, 1996 Whatever kind of organization you come from; whatever type of product or service you offer; if you are customer-driven and quality-conscious you are probably making some kind of measure of what your customers think about the quality of what they buy from you. Customer evaluations are a valuable tool for planning market strategies and product improvements. Successful companies retain their customers because those customers like the quality of what they deliver and feel they are getting value for the price paid.

I JUN 1 7 996 You may know a lot or a little about who your customers are and how they regard the products and/or services of your organization. You may even know something about how they regard your competitors —and, therefore, in developing marketing and improvement strategies where you have an edge or need to make changes. You know customers come in all sizes, shapes, and ability to spend. What you may not realize —because I can't find that there has been anything written about this in the literature in our fields —is that some of these differences are independent of what you do as a company or organization to provide quality, value, and customer service. What we are finding, using the database of the American Customer Satisfaction Index, is that there are some basic differences in customers which cut across the whole population of U.S. household consumers and thus affect nearly all of us working in quality, satisfaction, marketing, and planning. The American Customer Satisfaction Index, or ACSI as its cosponsors the University of Michigan Business School and ASQC call it, is the first crossindustry measure of the satisfaction of household customers in the United 2

I States. You may know about the customers in your industry, but ACSI provides us with measures of satisfaction with quality among customers of 200 companies in 34 industries in 7 sectors of the economy. That's why we've been able to discover differences which affect customer satisfaction independent of what the customer is buying. I'm presenting these findings today somewhat in the spirit of "quality and satisfaction provider beware or rejoice!" In this era when manager's and employees' compensation are often tied to the results of customer satisfaction measurement, some of you have a head start or a handicap in satisfying your customers just because of the demographic and socioeconomic characteristics of who your customers are. But, let me digress for a moment to explain how we measure satisfaction in the American Customer Satisfaction Index. Our methodology differs because it uses a model. ACSI models customer satisfaction using an econometric model designed at the University of Michigan Business School to maximize the 3

I relationship between customer satisfaction, or ACSI, and customer retention. Customer retention is a major determinant of profitability for most companies. The input to the model is a customer survey, not a single all purpose survey, but a survey which interviews separate samples of recent customers of each of the 200 companies. Customers are selected by screening a large number of replicate national, random-digit-dial samples of households. ACSI measurement starts from the universe of U.S. consumers in households with telephones, not from company customer lists. All customers in the 200 samples are asked the same 15 questions. We also collect 10 of their demographic characteristics. Thus ACSI uses uniform, cross-company, cross-industry measurement. There are no differences in the questions but different examples are given respondents to explain quality attributes. These make the interview as relevant for the purchaser of a personal computer as for the purchaser of a soft drink. Multiple questions are used to model each of the variables in the model: Customer Expectations, Perceived Quality, and Perceived Value —the 4

drivers of satisfaction. The ACSI itself is modeled using three separate questions which reflect the three schools of thought on how satisfaction should be measured. The questions: Did what the customer receive exceed or fall short of expectations? What is the customer's overall assessment of the quality of what he or she purchased? How does this quality compare with that of an ideal product or service of the type purchased? Finally, the modeling determines the consequences of satisfaction —Customer Complaints, and Customer Loyalty in the form of Customer Retention and Price Tolerance. I'm going to talk only about one of these variables today: Customer satisfaction, ACSI, which is at the heart of the model and measured with three questions answered on 1-10 scales. These three questions are weighted in the modeling and converted to a single ACSI index on a 0-100 scale. In analyzing 42,000 customers interviewed in the first year of ACSI, 1994, we found some important differences among these customer which prove to be common whether they were customers for automobiles or the gasoline that goes in them; whether they ate at fast food restaurants or were 5

I users of electric or telephone service; whether they were customers of General Foods, General Motors, or General Electric. Before I describe these differences, I need to explain two things. In 1994 the national ACSI was 74.5 on a 0-100 scale. That index was based on averaging the individual ACSIs of the customers of each company, then weighting that company average by company sales, industry sales, and the percent contribution of the economic sector in which the industry operates to the Gross Domestic Product (GDP). That was so that the national index reflects company and industry sizes. Today, when we are looking at individual customers, however, I am not using company size weights. Instead, we will be considering only averages for individual customers grouped by their demographic characteristics. On that basis the 1994 average ACSI is 77.7. Rounded to 78, that is the national number to which we will compare groups. The second thing I want to explain is that average customer satisfaction is very different in very explainable ways for various foarsectors of the economy. The ACSI demonstrates these differences, and confirms those found earlier in 6

I the Swedish Customer Satisfaction Barometer, established in 1989, and the German Customer Barometer, established in 1992. The U.S. is the third country to have a national customer satisfaction index. Customers are more satisfied with the quality of manufactured products than with services. Quality control is easier to maintain for manufacturing than for services which have huge people variations and supervision differences. Customers are more satisfied with manufactured non-durables more than durables. For durables such as automobiles, household appliances and personal computers the service component as well as product quality affect satisfaction. This is not so for a can of pineapple, a pair of jeans, or a bottle of beer in the non-durables sector. Finally, when it comes to services, customers are more satisfied by services delivered by the private sector, where companies which do not deliver good service cannot survive long, than by services delivered by the public sector. However, the public sector today is showing increasing customer orientation and is working on continuos improvement to meet the complaints of disgruntled taxpayers and voter. 7

I So now, with these two facts in mind: The national standard for comparison among consumers is an ACSI score of 78, and those consumers are more satisfied with manufactured goods than with services, let's look at some cross-cutting customer differences. The first basic difference in customers which we find is the difference by sex. Across all seven sectors of the economy(l-digit SIC) measured in the ACSI, women are more satisfied with the goods and services they purchase than men are. The difference is that women give about a 5-point better evaluation on the 0-100 scale than men do. We can ask ourselves why. Is it that women are kinder and gentler in their evaluations? Personally, I don't think so. I think it could be that women have had more shopping experience. They are better shoppers who are then more pleased with what they buy. Dropping from the sector to the industry level, the difference holds. Whether looking at men's and women's ACSI for such manufactured products as automobiles, apparel, processed foods, or soaps and detergents, or at service industries such as discount stores, fast food restaurants, airlines, 8

I or local police service, women give ratings which are 2-5 points higher than those given by men. There are age as well as sex differences in satisfaction. Recognizing the higher ratings given by women, we've looked at age groups separately for the two sexes. Now we find that satisfaction doesn't vary much within each sex among customers 18-54, but those 55 and over are a lot more satisfied than those under that age. I thought a lot about why this might be. Then I realized that these persons who were 55 and older in 1994 were born before 1939. Their first shopping experiences were during the depression, during the war, and in the boom period after World War II when consumer demand was so great that companies could sell anything they could produce. This was long before the quality movement brought on by a highly competitive economy. When an older person buys today's goods, he or she may be — either consciously or unconsciously —comparing these with what he or she bought in the "NOT such good old days." So now we know that sex and age make a difference in customer satisfaction. What else? 9

I What else is socioeconomic level. Here the data go counter to what one might expect. Shouldn't those who can afford to buy the highest quality merchandise be the most satisfied? The answer is a resounding, "no." Across all seven sectors measured in the ACSI, satisfaction declines as socioeconomic status goes up. Apparently those who can afford more are also more critical. The socioeconomic status variable here is a combination of education and income levels. Satisfaction goes down as education and income go up, but the differences are much more marked when we do a combined socioeconomic status comparison. The U.S. consumer is a rational human being who is looking for quality at a price he or she can afford. She may think Steuben glass is beautiful, but she'll be happy with a glass object from K-Mart or Wal-mart. He may fantacize about driving a BMW, but he's very satisfied with the quality/value tradeoff he made for under $ 10,000 when he purchased a Chrysler Neon. 10

I That's why discount stores or growing and department stores declining —the Wal-marts are delivering more customer satisfaction than most. department stores. The final demographic difference I'll talk about is race and ethnicity. Ours is a nation of diversity in which one-fourth of us are African Americans, Hispanics, Asians, American Indians or Eskimo Natives. At first glance, it appears that Asian customers are less satisfied than other U.S. customers with the quality of the goods and services they buy — giving average ratings of 72 on the 0-100 scale, compared to 78-79 for others. However, it is well to keep in mind that Asians are also our most educated segment. Forty percent have some college education compared to about 22 percent of the population as a whole. When we control for education level, however, the racial and ethnic differences wash out. While Asian college graduates are the least satisfied, a score, Asians with only a high school education or less have the highest satisfaction for manufactured goods, and whether for manufacturing or for 11

I retail, satisfaction is higher among the less educated for all racial and ethnic groups. Now, you know why I said at the start that some of you start with handicaps and some start with advantages in delivering customer satisfaction. And in looking at results among your market segments, it is well to keep in mind that what you want to do is track change among your different customers. Is your company improving or declining, keeping in mind that your measures may start from different base levels because: Women customers are more satisfied than men. Satisfaction is higher among those 55 and over. Satisfaction declines as socioeconomic status rises. When education levels are controlled, satisfaction levels are similar for racial arid ethnic groups. But I'll close now with a warning: Despite all of your efforts, customer satisfaction has declined among household consumers in the three 12

I years of measurement in the ACSI The decline is mostly for service industries rather than manufacturing. I'm shifting now to the national index in which individual customer scores are weighted by company and industry sales and to the sector proportions of GDP. In the baseline ACSI the national index was 74.5. Today it is 73.0, a decline of two percent. That is not good news and means we've all got a lot of work ahead of us. In the long run, customer satisfaction is the key to customer retention. 13