Editor’s note: Gene Leichter, president of Leichter Research for 13 years, has joined with New York research firm Eric Marder Associates to extol the benefits of choice research.

Over the last two decades, companies across corporate America have learned that they must listen to their customers to keep product offerings current, competitive, and on-target. Maintaining customer satisfaction is an essential element of building brands and building customer loyalty. Monitoring satisfaction with all aspects of the business enables marketers to anticipate and respond to perceived problems before they can cause damage.

Customer satisfaction research has become so widely used and specialized that many companies consider it an entirely separate discipline from marketing research. It has grown and matured over the years, as illustrated by the many ways companies use the results of satisfaction research. Satisfaction research data provide:

  • a single “point in time” report card to businesses;
  • a benchmark to track the success of product and service delivery;
  • a measure of management performance against goals;
  • an objective measure on which to base compensation;
  • an internal measure to evaluate performance of divisions, regions or other business units;
  • attributes to track satisfaction, not just a single score;
  • attribute ratings as a diagnostic tool for improvements.

The methods of data collection, analysis and reporting have also matured. Many years ago I worked on a customer satisfaction program that provided business reply cards to fast-food customers at the point of purchase. Data was often unavailable for several months. In another instance, I designed a program in which customers were interviewed by telephone to gather their opinions on a recent purchase. Reports were generated in only a few weeks.

Now we can collect the same amount of data using a self-administered questionnaire, completed at the customer’s convenience through an e-mailed link to a survey Web site. Tabular data can be compiled, updated and reported in real time.

In early customer satisfaction studies management was primarily interested in answering the perennial question “How am I doing?” As studies became more sophisticated and the interviews became longer, questioning covered more aspects of the customer’s experience. How this new data is used represents the next stage in the evolution of customer satisfaction research.

Considering the great expenditure of money and effort put into customer satisfaction research, you would expect there would be detailed analyses and actionable recommendations resulting in a direct impact on revenues. This is rarely the case.

Savvy marketers are beginning to realize you can get more from your customer satisfaction research budget. By using a choice research model, you can obtain all of the key customer satisfaction data you did before. In addition, you can generate accurate category share data and be able to conduct an unlimited variety of simulations and assess the impact of alternative strategies on repurchase rate – the “ultimate” measure of customer satisfaction.

Eric Marder Associates (EMA) has sought to understand and predict behavior. Through this process, EMA has developed a theory of choice behavior and proprietary techniques for gathering and analyzing data to predict how people will behave.

In his book The Laws of Choice, Predicting Customer Behavior Eric Marder summarizes his opinions about customer satisfaction research in his “Customer Satisfaction Principle.” He states: “Customer satisfaction is nothing but a brand’s deserved share among its own customers.”

EMA’s approach to measuring choice flows directly from the theory of how people make choices. Whether people are choosing a brand of facial tissue, or deciding which car to buy, the elements of the choice process remain the same.

  • People have specific needs.
  • People have unique perceptions about the choices available to meet those needs.

If we look at the intersection of each person’s needs and their subjective evaluation of the choices available to them, we can determine the value of each choice to that person. Through simulations, we can also determine the extent to which each choice would change if a new strategy were implemented.

These are the underlying principles of the SUMM (Single Unit Marketing Model) technique. This technique uses choice research to estimate deserved market share, which is what customer satisfaction is all about. This single research tool can provide a multitude of data:

  • traditional customer satisfaction metrics;
  • accurate estimates of share for your brand and for your category;
  • a tool for modeling and simulations.

SUMM predicts choice by integrating what people want (their desires) with what people believe. Data are not averaged; instead, all analysis is done at the single unit or respondent level.

The model differs from others in its reliance on respondents’ beliefs. The premise is that what people actually believe is more important than objective characteristics. For example, if a respondent believes your product has a great warranty, this belief influences their choice. The objective “fact” that your warranty is no better than the competition’s, is, for that respondent, largely irrelevant.

The SUMM technique has replaced traditional numerical and semantic scales with an open-ended behavioral scale to reduce order effect and increase sensitivity. It uses the unbounded write-in scale to collect ratings that represent desirability. This proprietary scale has several distinct advantages. Unlike bounded scales (1 to 10, very important to not important), the unbounded write-in scale permits full freedom of expression.

If a respondent likes one attribute a certain amount and then comes across something he likes even more, the unbounded scale does not limit him. Bounded scales cannot distinguish between a respondent’s most important choice element and others. The unbounded write-in scale can and does determine ranked importance for every respondent.

Bounded scales are subject to “bunching.” Typically, many respondents bunch at the poles or ends of the scale. For example using a +5 to -5 scale, 76 percent of people rated a food product a +5, implying that they liked this product so much that its desirability could not be increased further - an unrealistic conclusion. The unbounded write-in scale is more likely to generate a distribution of the desirability of a product or characteristic resembling a normal curve.

The model examines the desires and beliefs of each respondent and generates chooser shares, which closely parallel actual shares.

Simulations are the most valuable product of a SUMM study. Simulations, or games, allow the evaluation of countless strategies. In each game you are essentially asking, “If our advertising increases the perception that our product can do this, or our competitor initiates this new offer, or we make physical changes to our product, or all of the above occur, how will this impact people’s choices about our product?” In each case, the output provides unambiguous information about the impact of those changes on the choices people will make.

This simulation model is an enduring tool. Simulations can continue to be run for years, or for as long as the basic beliefs and value structure remain valid.

A client and dedicated user of choice research recently described how his company has integrated this modeling methodology into its customer satisfaction tracking process. “Instead of traditional satisfaction measures, we focus our attention on the ‘chooser share’ among customers. SUMM has been a powerful ally in maintaining and improving our marketing competitiveness. It lets us ‘simulate mistakes and introduce winners!’”

Traditional customer satisfaction research alone, or with the added benefits of a choice research modeling methodology. The choice is yours.