A Degree of Satisfaction
ARE ANALYSTS OVERLOOKING CUSTOMER RELATIONSHIPS?
By Rhea Wessel
Claes Fornell, founder of the American Customer Satisfaction Index (ACSI) has been measuring the “happiness” of custom- ers of listed companies for the past 18 years, and now, he uses that information to invest via a hedge fund he operates.
A professor at the University of Michigan, Fornell con- ducts market surveys of customer satisfaction, compiles the ACSI, and releases data to the public through the company he chairs, ACSI LLC.
The ACSI, established in 1994 and based on a model For- nell developed in his native Sweden, is widely recognized as the industry standard for measuring customer satisfaction, and ACSI data have been the basis of numerous studies showing a link between customer satisfaction and financial performance of a company.
The ACSI was developed from an idea similar to the concepts behind happiness indices for countries. Where- as a national happiness index reports how “end users” of the economy feel about its overall impact on their lives, the ACSI measures the sentiment of the end users of specific products or services.
“Especially in a service economy, we need an index that does not focus only on prices and productivity,” says Fornell. “We need one that measures the quality of the economic output. Who deter- mines the quality in a market economy? The consumers. If the quality is poor, they won’t purchase the product.”
A portfolio that Fornell manages via CSat Investment Advisory, LP, a hedge fund, goes long on shares of large- cap consumer goods companies with high customer satis- faction scores and shorts those with low scores. According to performance results reported by the ACSI, the CSat Classic Fund, based on ACSI data, beat the S&P 500 Index by nearly 400% (before fees) from April 2000 to April 2012.
Still, many financial analysts question the value of customer satisfaction and are reluctant to incorporate customer satisfaction data into their models in a systematic way, says Paul-Valentin Ngobo, a professor of marketing at the University of Orleans and a co-author of a paper called “Is Customer Satisfaction a Relevant Metric for Financial Analysts?” which was published in the March 2012 issue of the Journal of the Academy of Marketing Science . The upshot: analysts who ignore customer satisfaction information may be depriving themselves of important nonfinancial information that can affect a company’s future income.
“We know that customer satisfaction is highly related to cash flow,” he says. “When customer satisfaction increases, companies generally experience higher future cash flows, reflecting the fact that satisfied customers are more loyal, buy more often, recommend the product and services to friends, and are less price sensitive—all of which reduce the volatility of the company’s cash flows.”
The impact of customer satisfaction on stock return risk was the subject of a 2009 paper published in the Journal of Marketing by co-authors Sundar Bharadwaj, now a professor of marketing at the University of Georgia, and Kapil Tuli, an assistant professor at Singapore Management University. Companies that have increases in satisfaction over time experience decreases in risk, they found. “One reason is because higher satisfaction leads to fewer returns of a com- pany’s product and, therefore, less risk,” says Bharadwaj.
Tuli and Bharadwaj examined how customer satisfaction affects share volatility. “What we found is that it has a robust effect on the volatility of the stock-i.e., on the idiosyncratic risk and also on the beta, the systematic risk,” says Tuli. “Firms that have higher customer satisfaction tend to have lower systematic and lower idiosyncratic risk. You can use multiple alternative methodologies, and the result is very strong and very consistent.”
The paper challenges what Bharadwaj calls the “usual financial paradigm” (i.e., to increase return, one has to increase risk). “What we found was the opposite-that you can get increases in return with reductions in risk,” says Bharadwaj. “My sense is that financial analysts primarily look at fundamentals. And for them, in many cases, the fundamentals are reflected in cash flows. Our idea is that satis- faction data can actually add new information above and beyond fundamentals.”
CUSTOMER SATISFACTION SURVEY METHODOLOGIES
Customer satisfaction surveys in the business-to-consumer sector became widespread in the U.S. and Europe in the 1990s. They were implemented as companies faced slow- ing growth, increasing competition from Asian companies, and changing cost structures that made price competition difficult for companies in developed economies. Most surveys were—and still are—conducted for internal use, such as segmenting customers, developing internal change pro- grams, and setting executive pay. Volkswagen, for instance, has made customer satisfaction scores a part of variable remuneration for executives since 2010.
The results of customer satisfaction surveys tend to be board-level matters of corporate development and strategy that are discussed briefly in annual reports but seldom in the public venue. Reportedly, they may represent the larg- est single item in most marketing budgets.
Typically, surveys are specific to an industry and questions are designed around qualitative research on how people make their purchase decisions. A list of attributes and crite- ria is created and then turned into a questionnaire. Depend- ing on the purpose of the survey and the budget available, a sample size is selected.
Data collected are analyzed in various ways (for instance, with a quadrant analysis method, through linear or non- linear regression, or with a latent-class analysis). Typical results consist of scores for satisfaction with various facets of the product and service as well as for overall customer satisfaction, which is considered the most relevant metric for financial analysts. Some analyses calculate percentages of satisfied and dissatisfied customers or create scores for customer loyalty.
The quadrant analysis method (or the “importance-performance analysis”) plots the satisfaction scores for each product and service attribute versus the importance of the attribute for the customer, such as a product’s price or the company’s service. By identifying areas of high importance with low satisfaction scores, according to Ngobo, these analyses offer valuable insights into what things a company can improve to have the biggest impact on customer satisfaction in the shortest period of time.
Regression analysis shows which attributes have the largest impact on satisfaction—for example, the impact of prices on satisfaction. And latent-class regression analyses examine how variables may affect satisfaction differently across segments of customers—for example, how price may be more important for some customers while quality may be more important for others.
The American Customer Satisfaction Index, which uses a multiple-indicator approach to measure overall customer satisfaction as a latent variable, conducts surveys of roughly 225 companies in 45 industries-from energy utilities to hotels. After prescreening to ensure that respondents are customers, it interviews about 80,000 Americans each year. Once the data are collected, modeled, and crunched, an ACSI score between 0 and 100 is assigned to each organization surveyed. Scores can then be benchmarked against other companies in the same industry.
Finally, scores can be projected to create weighted industry and sector scores as well as a national score for customer satisfaction. Projecting the data to a national level enables analysis in conjunction with macroeconomic data.
In the aggregate, customer satisfaction predicts consumer spending well, according to some researchers. “If satisfaction goes up, consumer spending tends to go up,” says For- nell. “Since consumer spending in the U.S. is 70% of GDP, this is very important from a macroeconomic perspective.”
Although ACSI data can give an indication about consumer spending, Fornell concedes that ACSI data essentially provide supplemental information for any analysis.
NET PROMOTER SCORE
A familiar metric that companies publish and discuss frequently with analysts is the company’s so-called net promoter score (NPS). Although debate continues among academics about the value of this metric and some people don’t consider it a satisfaction metric at all, some large companies, such as GE and Philips, have embraced it. Philips began publishing its NPS in 2008 as part of its environmental, social, and governance reporting.
An NPS is based on the question “How likely are you to recommend Company ABC to others?” According to Sundar Bharadwaj, a customer who gives a rating of 1-6 on a 10-point scale is usually considered “a detractor” whereas a customer who gives a rating of 9-10 is considered “a promoter.”
To calculate the NPS, survey analysts take the percent- age of people who scored their willingness to recommend as a 9-10 and subtract from that the percentage of people who gave a rating of 1-6. The result is the NPS.
Bharadwaj contends that NPS is a powerful way to mea- sure how a company is doing because it is forward looking. “Customer satisfaction is very historical: ‘What did you do in the past for me, and how did I perform?'” he says.
Paul-Valentin Ngobo and Claes Fornell, however, question the predictive power of NPS. Fornell argues that NPS is not a customer satisfaction metric because customer satis- faction is a predictor of repeat business (i.e., loyalty). “NPS is not relevant for repeat business. It is directed at first purchase. NPS measures word-of-mouth recommendations,” he maintains. “For a consumer who has already bought the product or service in question, such recommendations are almost always trumped by the consumer’s own satisfaction with the product and have very little, if any, effect. The consumer is likely to buy it again if he or she is satisfied and not likely to do so if dissatisfied. At that point, recommendations from others don’t really matter much.”
A MISSING VARIABLE
If customer satisfaction with a company is widely accepted as an indicator of the long-term economic performance of that company, why aren’t financial analysts considering the information consistently and on a wider basis?
Financial analysts lack a common reporting methodology for customer satisfaction information, argues Sascha Raithel, an assistant professor at Ludwig Maximilians University in Munich who studies the link between intangible market-based assets and financial performance. “If half of the companies report customer satisfaction and the other half don’t, it’s not very useful,” says Raithel. “Analysts don’t know if not reporting is a sign of poor customer satisfaction or perhaps the company just doesn’t do it. The only database that comes close is Fornell’s ACSI, but I’m not sure whether analysts really use it as a standard metric.”
Even if analysts used ACSI data, determining how satis- faction today translates to financial returns tomorrow would be difficult. “That’s really a hard thing to prove because the mechanisms are very complex and the effects are usually long term,” says Raithel. “Some research suggests a strong link between customer satisfaction and long-term stock performance, but right now, the data are not avail- able to financial analysts to reliably calculate a ‘return’ on customer satisfaction.”
If companies do calculate a return on customer satisfaction for their internal use, that information is not usually available to analysts and investors. Companies have no obligation to report satisfaction metrics. “If regulators forced firms to report on such metrics, this might change,” says Raithel.
Another possible reason that analysts underuse or don’t use customer satisfaction data is lack of training. Analysts have not learned how to incorporate intangible assets such as customer satisfaction into their models, according to For- nell. In addition, he says, customer satisfaction data are “so simple and so obvious” that somehow the data are almost ignored as a result. “I think this explains why financial analysts overlook the information,” adds Fornell.
Another point that could hold back financial analysts is a well-founded skepticism about the quality of customer satisfaction surveys. A large research industry has grown up around the surveys, with little transparency about differing methodologies and quality standards. Consequently, results are hard to compare across companies. Naturally, financial analysts will question the reliability of self-reported data about something that cannot be seen or touched.
Academics have followed two lines of thinking on mispricing. One group says financial markets misprice information on customer satisfaction, making it possible for investors to beat the market. The mispricing comes from a neglect of the information or a delay in its analysis. Another group says there is no mispricing; the information has already been priced into the market, with the exception of a few industries. Therefore, investors cannot beat the market.
According to Ngobo, the utility sector is an exception to pricing consumer satisfaction. The reason customer dissatisfaction does not automatically affect revenues is switching costs. Customers are not likely (or able) to switch; therefore, analysts following utilities are not responsive to a negative change in customer satisfaction. In the information technology sector, analysts of companies respond to a decrease in customer satisfaction with a lag, which probably reflects the complexity of the industry. This lag may explain why some academics have reported a mispricing effect from customer satisfaction information in the computer and internet sectors.
Ngobo and his co-authors argue that financial analysts do incorporate customer satisfaction but not necessarily ACSI data, which are reported with a delay because the data must be collected and analyzed.
“We say the market does not necessarily misprice satisfaction information per se. The analysts do respond to customer satisfaction information because there are ways to get this information other than the ACSI, such as press releases, conference calls, and discussions with managers,” Ngobo explains.
Marcus Pratsch, the head of sustainable investment research at DZ Bank in Frankfurt, uses customer satisfaction information in his model, albeit in a small way. DZ Bank rates all equity stocks it covers for sustainability but uses only information that is publicly available. As one of 180 “subindicators” for which companies can be awarded points in the model, customer satisfaction falls into the “social” part of the analysis because the bank examines the impact that products and services have on society.
Consider an example of how data are incorporated into the model: “A company can earn a maximum of four raw points for the customer satisfaction subindicator,” says Pratsch. “If the company conducts a customer satisfaction survey, it earns one raw point. If it conducts it on a regular basis, it can earn up to another raw point. If the survey is of high quality, the company can earn—depending on the quality level-up to two more raw points.” ”
At the end of the day,” adds Pratsch, “we see customer satisfaction data as one part of the puzzle.”
SEEKING AN EDGE “Markets have not historically emphasized metrics like customer satisfaction and loyalty, but they add incremental explanation to a stock’s story,” says Bharadwaj. “I think there’s a good arbitrage opportunity for firms that do use this type of information. Future generations of analysts will become more cognizant of this because we’re starting to train them on these things. And marketers are starting to talk about these kinds of metrics as well. I think that customer satisfaction information would even help retail investors.”
Customer satisfaction may gain importance as online shopping shifts power to consumers away from producers via transparent information about price and quality. “The buyer is gaining more power to dictate and become disloyal, and that means satisfaction will become even more important in the future,” Fornell says.
“If you believe that the most important asset a company has is its customers and the customer relationship,” says Ngobo, “it makes sense for you to try to use information about the quality of those relationships in financial analyses. You won’t find that information on the balance sheet; you’ve got to look for it elsewhere.”