True Value

Nancy Marie Brown
May 01, 1999

We're not talking ketchup here. You can cut a coupon. Chewing gum's not in this league either, or laundry soap.

The good buys at issue are cars. Washing machines. Refrigerators. Stereo systems. Healthcare plans. Longterm lotta-money purchases. And not necessarily those that make you look thinner, smell sexier, or step more lightly.

How do you market an item like that? What should your strategy be to increase your market share? Who's buying what, and how do you make them switch?

In the 1980s what mattered was customer satisfaction.

But satisfaction, it seems, is passÈ. "It's not clear that satisfaction pays its way," Wayne DeSarbo says. He leans forward, elbows on the table, arms flat. His gaze is focused front, concentrated to a pinpoint. Behind him in the crowded corner office is a solid wall of computer gear: printer, server, monitor, microphone. The window shade is drawn. "Imagine, if you will, you have N competitors," he says slowly. "One of them distinguishes itself with a fullblown customer service plan. Clearly that supplier will benefit. But when everyone starts doing that, the benefit of an expensive customer service program becomes marginal." A distinguished professor of marketing at Penn State, DeSarbo is the kind of guru companies call when they've got a question like, What's the word on satisfaction, Wayne? A recent profile lists among his hobbies golf, tennis, football, power lifting, and chess. He was once the Pennsylvania state champion power lifter in the super heavyweight division.

"What people typically fail to do in the satisfaction game," he continues, "is they fail to become aware of costs on customer satisfaction programs. Customer satisfaction can be both a cost and a benefit. It's a maintenance item. It needs to be optimized, not maximized.

"Track satisfaction levels over time against such things as loyalty, switching, and market share, and you'll find that customer satisfaction may stay stable over time, but these other indices—which are more closely related to profitability—may move dramatically. A prime example is the telecommunications marketplace in the 1980s. The well known market leader lost significant market share over this period of time—even though its satisfaction levels remained quite stable. Satisfaction alone can render a very distorted view of the marketplace.

"What it reduces down to," says DeSarbo, "is value. That's the '90s issue: Value."

In his 1949 Treatise on Values, philosopher Samuel L. Hart wrote: "The concept of value permeates our life at every step. We prefer one thing to another, we shift attention from one event to another, praise one behavior and condemn another, we like and dislike, and whenever we do so, we value."

The most valuable thing to Theophrastus, writing 300 years before Christ, was "time." To Mark Twain, it was "the truth" (to which he added, "Let us economize it.") Other quotables have proclaimed, "Nothing has value" (E. M. Forster) and "What is valuable is not new, and what is new is not valuable" (Lord Brougham).

Webster's defines value variously as "a fair return in goods, services, or money," as the worth or relative worth of something or its "status in a scale of preferences," and as "something intrinsically valuable or desirable."

Leah Gorman, a 41-year-old San Francisco housewife quoted in a 1991 Business Week cover story defines value as "not overpaying for quality."

"Value is becoming the marketer's watchword for the 1990s," Business Week's writers trumpeted. "It's what customers are demanding." Value, they said, "isn't about positioning and image-mongering but about real performance and delivering on promises." Nor is it "just a euphemism for a price cut." No, "value marketing means much more than slashing prices or handing out coupons. It means giving more: an improved product, with added features and enhanced service—all at a better price."

When it comes to the question, How?, the magazine backs off. "For companies that figure out the equation," it not-so-helpfully says, "value marketing could be a way to make a brand name mean something again."

Value is an "obscure and complex construct," DeSarbo admits, but that doesn't make it unmanageable. Business Week was on to something when they said, "Figure out the equation." DeSarbo has. Give him a product—a compact car, a healthcare plan (not ketchup)—and he'll calculate its value.

We start with the voice of the customer. You must listen to the voice of the consumer," DeSarbo insists. "When we say value, we really mean perceived value, value in the eyes of the customer. So we need to consider his or her reference points. Where is the customer now?"

"For instance, you want to know the value of a brand new Jeep Grand Cherokee. Ask someone who has a 1981 Dodge that's beat up and smoking and you'll get a different answer than you will if you ask someone with a year-old Cadillac DeVille.


"Value comes from the consumer's perspective, not the engineer's or the marketer's. What's of high value to one consumer is not high value to another."

Phone surveys aren't going to go away. But as people get less and less open and helpful, DeSarbo has learned ways to make his surveys simpler and shorter. Rating 12 compact cars on various points was easy. Basically the researchers asked customers "to put things in three piles," says DeSarbo: "good deals," "so-so," and "rip-offs." Just like shopping.

"The other issue besides reference points is, What are the buttons they push? For some people, value is a price-quality trade-off. For others it'll mean plain quality. They don't care about price. Others want the cheapest buy around. And then there are others who focus on a single attribute: Mileage. Dependability. Performance."

After the surveys—"typically surveys involving tens of thousands of consumers"—comes the computer work. Organizing all that data into an action plan involves, as DeSarbo puts it, "some non-trivial mathematics. Much of this is CPU intensive." Which is just fine for a power-lifting chessplayer. As he told another magazine, The best part of my whole job is being locked away and thinking of a solution to a real problem, working out the mathematics, doing my own computer programming, and seeing results. It's just an unbelievable high. That's why I'm in this business.

It's not why businesspeople—or MBA students—are in their businesses, though. DeSarbo knows that. He pulls out a paper that he and his former graduate student at the University of Michigan, Indrajit Sinha, published in the Journal of Marketing Research. In it, they describe VALUEMAP as a "latent-structured, ordered-probit multidimensional scaling-based methodology."

DeSarbo flips quickly past all the equations and technical stuff and goes immediately to Figure 5, the "VALUEMAP of 12 Cars."

"No MBA student I know can read this article," he says. "That's not what they come here for. Business-people don't want to look at numbers and equations. They'd rather have pictures, if you will. Here's where things like VALUEMAP come in."

It's two-dimensional, computer-derived. Straight lines, kind of sketchy, not so snazzy in black-and-white. The names of 12 small cars (Eagle Summit, Honda Civic, Geo Metro, Mitsubishi Mirage, Mazda ProtÈgÈ, Subaru Impreza, Ford Escort, Corolla, Saturn, Cavalier, Neon, Sentra) are scattered in the field. Two intersecting arrows, labelled "Segment 1" and "Segment 2," divide them, each arrow bearing a "Lo" and an "Up" marking. DeSarbo takes a pen and draws a dotted line from the Mazda ProtÈgÈ to meet the Segment 1 arrow at right angles: the two lines meet well below the "Lo" cut-off mark. He draws another dotted line from the ProtÈgÈ to the Segment 2 arrow. The meeting spot here is almost exactly between the "Lo" and the "Up." Which means? For Segment 1 of the population surveyed, the ProtÈgÈ is a "rip-off." For Segment 2, it's "so-so," okay, as good as they've got now—they might be driving one.

DeSarbo draws another set of dotted lines. The Honda Civic is off the scale on the Up side for Segment 1 (a "good buy"), while for Segment 2 it's down below "so-so" and heading toward "rip-off."

There's not too much these two groups agree on. Only the Subaru Impreza and the Neon rank above Up on both arrows; only the Mitsubishi Mirage is unanimously below Lo.

Keep in mind that the sample population for this sample VALUEMAP was a bunch of college students in an undergraduate marketing class. They saw data on each car from Consumer Reports: its maker, reliability, mileage, safety features, cost factor (how much of the sticker price is actually the dealer's cost), depreciation, performance, and average price. They lumped each car into "rip-off," "so-so," or "good buy." They were asked about the importance of the eight other attributes. How avidly do you "Buy American"? What's your sex, age, and college major? Do you own a car? Buying one soon? How do you rate price versus quality? Their answers went into the computer and the VALUEMAP came out.

"Now this map is cool in the following ways," DeSarbo says. "First, there's a whole math behind it. We had 95 students, but we could have surveyed hundreds or thousands of customers and VALUEMAP still would have dealt with the problem of heterogeneity. We let the computer program group all those customers who seemed to be using the same basis for comparison, and the best fit to the data was when it split the customers into two segments"—the two arrows. "The program simultaneously finds locations for the various brands of cars we asked the customers to respond to. Two dimensions and two segments recover the data very well. As the consumers make more complex judgments, we need to use more dimensions, but here there were just two: cost and quality.


"It's benchmarking, first. But we can go further than that. For any of the eight attributes"—, reliability, mileage, safety features, cost factor, depreciation, performance, and average price—"we can get the impact on the value of the car if we increase or decrease it. What happens to a car when we increase its MPG? How does that affect its perceived value?" VALUEMAP knows.

"So the so-what' is this: If we know what buttons to push and how that will move cars around in space, we know what things to focus on.

"Say I'm Ford. Suppose I want to come out with a new car. With VALUEMAP I can get a predicted position in car-space. Do I have to go back to the consumer? Do I have to do another survey? Nope. I can just do computer simulations on this map.

"What if GM does some repositioning of its cars? How will that affect the value of my cars? We can do that. What do I do to counter their move? We can do that too.

"Part of our survey—this one took 10 to 15 minutes to fill out—was demographic, psychographic, the characteristics of the customer. We could go back and solve for the demographic characteristics of each segment—here, Segment 1 was mostly female, Segment 2 was mostly male. Women want quality, men worry about the cost." Or, as DeSarbo and Sinha expand it in their paper, Segment 1 is mostly female, price and brand insensitive, safety, performance, and quality conscious, and prefers Japanese cars. In contrast, Segment 2 is mostly male, highly price and depreciation conscious, less interested in the safety, performance, and quality aspects of the automobile, and prefers American cars. "That," DeSarbo says, "gives us the ammo we need to go write ads."

Wayne S. DeSarbo, Ph.D., is the Smeal Distinguished Chaired Professor of Marketing at the Smeal College of Business Administration, 701 Business Administration Bldg., University Park, PA 16802; 814-865-8150; Indrajit Sinha, Ph.D., is assistant professor at Temple University. "An Integrated Approach Toward the Spatial Modeling of Perceived Customer Value" was published in the May 1998 Journal of Marketing Research.

Last Updated May 01, 1999