Claes Fornell: Customer Science

Swedish-bred Claes Fornell, director of the National Quality Research Center at the University of Michigan, exhibits little of the reserve often attributed to his countrymen. His voice rises with excitement when he speaks of the research that led to development in 1994 of the American Customer Satisfaction Index, a collaboration of the UM Business School, the American Society for Quality and Fornell’s international consulting firm, CFI Group.

Fornell’s sampling and survey techniques — influenced by the university’s long history in measuring consumer confidence — have already been adopted by many researchers who measure customer satisfaction in the housing industry, including Paul Cardis of NRS Corp. “He’s the guru we all watch,” Cardis says.

The trouble is, builders who have their own internal satisfaction surveys are not paying as much attention. Those simplistic surveys might be doing more harm than good.

“ACSI is a national measure that really counts the value of products to the people who buy them,” Fornell says. “It’s that consumer point of view that’s unusual. Traditional accounting leaves customer relationships off the balance sheet. It produces financial reports that give a distorted view of how a company makes money. In ACSI, we are able to capture what affects the quality perceptions of consumers, and that creates a much better understanding of a company’s current condition and its future capacity to produce wealth.

“We felt from the beginning that this measure would be able to predict consumer spending, which makes up two-thirds of our national economy, and also predict corporate earnings and stock-market performance of individual companies. And it has turned out just that way.”

The ACSI reports scores on a 0-100 scale in seven economic sectors, 35 industries (including e-commerce), 190 companies and a number of federal and local government agencies. It does not measure the housing industry ... yet.

What makes the ACSI so superior? It’s the model as well as the methodology. From random-digit-dial telephone samples, more than 70,000 consumers are identified and interviewed annually. Scores for a company’s customers are aggregated to produce company-level results. The score for an industry consists of the average of its company scores weighted by the revenue of each company. Sector scores are compiled by weighting industry revenues. The national ACSI is determined by weighting each sector’s contribution to the gross domestic product.

Fornell’s econometric model was first used in Sweden, in 1989. It is now used in Asia and South America as well as Europe. The ACSI model is light-years beyond counting yes and no responses to the famous “would recommend” question most builders rely on as a measure of customer satisfaction. Fornell explains the difference: “The ACSI model is a set of causal equations that link customer expectations, perceived quality and perceived value to satisfaction. In turn, satisfaction is linked to consequences as defined by customer complaints and customer loyalty measured by price tolerance and customer retention. For most companies, repeat customers are major contributors to profit.”

The crucial work of the ACSI model is filtering out what Fornell calls “noise” from raw data. “When we ask people today how they voted in the last presidential election, Bush wins by 12 percentage points,” he says. “Obviously, many of those people are lying, because Gore actually won the popular vote. That’s typical in survey research. People lie. They misunderstand questions. There are all kinds of reasons why raw survey data is misleading. But we know from experience that about 30% of it is inaccurate.

“We have to employ certain mathematical tools to purify the data and turn it into information you can use. The answers to certain questions are interrelated in such a way that if you know enough math, you can sort out the noise. It’s like solving a puzzle.”

Fornell says he has stayed away from housing because he perceived it to be an industry that is not dependent on repeat business and one in which customer satisfaction does not affect re-entry into the market after an initial purchase. “I always thought re-entry was more a function of life changes,” he says, “a new job, a bigger family...” He is intrigued, however, by Pulte Homes’ contention that every satisfied customer generates 5.5 new customers for the builder.

“Satisfied customers are very good for business, but dissatisfied customers have even more impact in the opposite direction,” Fornell says. “To put it another way, a bad customer experience does more harm than a good one does good.”

Fornell also takes issue with builders who measure customer satisfaction at closing. “How can you measure at closing? You’ve bought this thing, but you don’t know how it’s going to work for you. You have no consumption experience at closing. To survey then and take the responses at face value might create a false sense of security that would have no relationship to reality. What if I ask you how you like your new car the day you drive it off the lot? Would the response be the same six months or a year later? Which one has the most validity in predicting the manufacturer’s future economic performance?”

Cardis believes such errors are widespread among home builders. He’d love to see housing included in the ACSI. “What Professor Fornell calls noise is the reason so many builders are scratching their heads trying to figure out why 99% of their buyers say they would recommend them to friends, but when J.D. Power surveys their market, they rank at the bottom.”

Cardis sees another reason builders should push to get home building into the ACSI. “It would give us a connection to other industries,” he says. “We’d be able to see where we rank in comparison to automakers and airlines in meeting customer expectations.”

Fornell promises to take another look at housing. In the meantime, he suggests that builders visit his Web site to see the surprising changes in customer satisfaction happening in those other industries. “Just hit www.bus.umich.edu,” he says, “and click your way to our site. Scores and comments are updated quarterly.”

What you’ll see is that the airlines are doing miserably, which is probably no surprise, but Fornell attributes that as much to lack of competition as to Sept. 11. “Most airports are dominated by a single carrier,” he says. “There’s little incentive for the airline to provide good service and reach high levels of customer satisfaction when the customer has no alternatives.”

The auto industry is another story. Competition is fierce, and while it would seem that customer satisfaction and perceptions of quality would change slowly, they actually occur very fast. “In the mid-1990s, the Japanese manufacturers were strong, Americans weak, and the Europeans were in the middle,” Fornell says. “Today, it’s Europeans on top, with the Americans and Japanese going back and forth for second place. Nobody predicted the leap the Europeans made.”

Builders who get a handle on what drives customer satisfaction can make the same leap. Paying attention to the guru of satisfaction measurement is a good idea. Lobbying to get the ACSI to measure the housing industry is not a bad one.