Editor's note: Alex Vayslep is division manager of Maritz Marketing Research Inc.'s Chicago office.

Bradley T. Gale's book, Managing Customer Value, has, at the very least, driven many of us to take another look at how we conduct customer satisfaction research. Based on our experience in implementing several customer value measurement (CVM) programs, CVM has distinct advantages compared to traditional customer satisfaction research. Generally, CVM:

  • Provides the ability to link customer value trends to market share and therefore develop a tracking mechanism that is more closely aligned with meaningful business objectives.
  • Measures price/value perceptions (as opposed to just satisfaction) in the market as a whole and therefore allows us to devise strategies to improve customer loyalty by comparing our client's performance versus the competition, as customers do.

As a strategic marketing tool, CVM is a superior design. However, discarding traditional - or what Gale refers to as "conformance-oriented" - customer satisfaction research can be hazardous to a company's customer retention efforts.

Before reviewing the strategic advantages and potential pitfalls of CVM, let's first review the basic principles of customer value measurement.

CVM's appeal was born from many marketers' frustration with their inability to tie customer satisfaction research results to traditional customer loyalty and customer acquisition metrics such as retention rates and market share. CVM addresses the two primary problems that limit the ability of customer satisfaction results to link with these traditional performance measures. Unlike most customer satisfaction research, customer value research addresses:

1. Performance perceptions of the competitions' customers as well as our client's customers.

2. All the broader issues relating to how customers select a product or service. This includes a more complete investigation of price perceptions as well as the less tangible buying influencers such as image and reputation.

The customer value concept is based on the principal that customers make buying decisions based on their relative value perceptions of two or more providers in the marketplace. More specifically, relative value is defined as perceived quality vis-a-vis price for all offerings in the market. The basic model is graphically displayed here.

The Consumer Value Map depicts both quality and price perceptions for two competitors. Positions on the map are determined based on competitor performance ratios (overall price and quality scores). The fair-value line represents points at which a given provider should neither gain nor lose market share based on the importance of quality and price to the buying decision. At any point along this line, a company is perceived as charging the appropriate price for the quality provided.

The farther a competitor's position is to the right (superior quality) and to the bottom (lower price) of the map, the better overall value position it is in to compete. For a complete description of the techniques used to develop the map, see Gale's book.

How do we identify and define value?

Qualitative research typically provides us with the quality and price dimensions customers use to make purchase decisions. Customer value perceptions are based upon broader price and quality dimensions. Issues such as customer service, product quality, etc., make up the quality factor. Within both the quality and price factors, specific attributes emerge as the issues that drive overall perceptions for the factor. These attributes become the basis for performance measurement in quantitative measurement.

CVM requires us to measure our competitors' performance on the same dimensions that most customer satisfaction programs limit to just current customers. This is the primary tactical difference between customer satisfaction and customer value approaches. Just as important, CVM doesn't shy away from evaluating customers' price perceptions of the offerings made by marketers.

The willingness to evaluate price perceptions is a key differentiation for CVM when compared to the manner in which many customer satisfaction programs are currently implemented. "Quality" managers have a propensity to shy away from price issues in both the measurement and reporting of customer satisfaction surveys for fear of losing focus on quality improvement efforts.

While their motivation is understandable, neglecting the role of price perceptions ensures the inability to link research results with how customers (and non customers) actually behave in the market. Moreover, since CVM approaches call for us to understand relative performance perceptions for all competitors in the market, we can more accurately predict both customer loyalty and market share.

Implementation of quantitative measurement

Most CVM surveys are conducted via telephone interviews in a "blind" fashion. In other words, the respondent is not made aware of the survey's sponsor.

A blind approach ensures that competitor evaluations are comparable. Sponsorship of a survey leads to significantly higher satisfaction or value ratings for the sponsoring firm -- particularly if the respondent is evaluating a product or service they currently use. Maritz has demonstrated the biasing effects of sponsorship on satisfaction and competitor preference through parallel tests of both blind and sponsored approaches. The research revealed the sponsoring organization's scores dropped by as much as 20 percentage points when the identical survey is conducted "blind."

Although some CVM methodologies gather customer ratings for only the respondent's current or preferred provider, we strongly suggest obtaining ratings for both the sponsoring company and the appropriate competitor(s). Reasoning for this approach is twofold:

1. Most customers make buying choices based on value comparisons for two or more companies in the market. These decisions are based on either actual experience or perceptions determined by what they have seen or heard.

In either scenario, the comparison is based on the buyer's opinions. If we simply gather respondent satisfaction for their current or preferred company, any value comparisons (analyses) are not reflective of how a given buyer actually makes a choice.

2. The key to the most actionable analyses is to gather respondent-level data for both the current and competitive provider. Much of the analyses described hereafter are made possible (or better) by respondent-level analysis.

Among calls placed to the sponsoring organization's customers, gather competitor ratings for only the broader factors (price and various quality dimensions) - no individual attribute ratings will be asked for the competitor. Ratings for the sponsoring company will be based on experience, however ratings for the competitor will be determined primarily by what the customer has either seen or heard. While respondents are capable of rating broader factors without having experience with a company, it is difficult for them to evaluate performance on specific attributes based solely on what they have seen or heard.

Among calls placed to competitors' customers, gather ratings for the sponsoring company for only the broader factors in the same manner described above.

This method will provide the most actionable data set possible while minimizing questionnaire length and respondent irritation over asking questions which they are unable to answer.

Preference modeling

Once we have perceptions for both the sponsoring firm and its competitors on key customer value issues, we can model the criteria customers use to select a particular provider of a product or service. Preference (or difference) modeling for CVM programs is a technique developed by Maritz to answer the following questions:

1. Which criteria drive customers' preference for one company over another?

2. What is the relative importance of each of the attributes that drive the customers' choice or preference?

3. What are the likely criteria used by customers to defect or switch to an alternative supplier?

Preference modeling is a form of derived importance analysis that uses score differences (or gaps) at the respondent level to determine which issues drive perceived value preference for one provider over another.

Preference modeling techniques are an excellent fit with the principles of CVM since the model focuses on competitively based perceptions as opposed to just satisfaction for one company. In addition, we have found that preference modeling techniques provide importance hierarchies with significantly greater discrimination among attributes. In addition, these analyses typically provide a predictive model that better predicts customer actions (R-square, or goodness of fit explanation).

Preference modeling techniques can help prioritize probable defection and acquisition areas based on the:

  • relative competitor performance scores on quality and price factors, and
  • the importance weights developed by the model.

As such, we can prioritize strategies for market share improvement based on relative perceptions of the competitors in the market.

Like any strategic analysis of customer perceptions, we must ensure that value perceptions and needs are analyzed by key customer segments. While it is always tempting to "roll-up" results to a corporate level, assuming homogeneity in the marketplace usually results in improvement efforts that are inappropriate for some customers. If our clients have not conducted a formal segmentation study prior to measuring customer value, we typically analyze the results by whatever demographic, financial, or psychographic schemes are available to ensure understanding of how different customer segments view the market.

Financial linkage

CVM results have been successfully tied to financial results or market share in a number of studies. After measuring customer value for a number of time periods, we can attempt to establish statistical relationships using one or more of the following data sets:

  • Customer value scores and financial performance over time for one or more market segments;
  • Customer value scores and corresponding financial performance for a large number of markets for one time period.

A correlation can be drawn between the two metrics in either scenario listed above. In addition, we can actually plot the two metrics as well.

Potential pitfalls

Appropriately, much has been made of some firms' success in tying CVM results with actual market share trends and other financial performance metrics. While these successes are enticing to organizations embarking on a CVM program, one should remember these results have been obtained in mature markets. While CVM approaches can still provide superior strategic information, growing industries can pose problems to the researcher wishing to link CVM results to financial metrics. Before proceeding with a CVM program, we suggest to our clients that they consider the following:

Source of market growth - Growth for companies that compete in new industries or that continue to attract new customers is as much or more a result of attracting first-time buyers as a result of stealing market share from current competitors. As such, surveying current customers for your firm and its competitors will not necessarily correlate to market growth or share.

For these markets, consider measuring perceived or expected customer value performance among likely adopters of the emerging product or service. While somewhat expensive, this can be accomplished by conducting the CVM study among the demographic and/or psychographic segments that are typical of the new customers entering the market for the first time.

The number of buying decision dimensions - For most industries, measuring price and quality dimensions is enough to help predict how customers actually behave in the real world. While plotting two dimensions on the customer value map makes analyses easy to present, in some circumstances, other criteria play an important role in how decisions are made and a method to deal with a third dimension must be developed.

Think about how you might decide from which service station to purchase gasoline -- you may prefer Billy Joe Bob's price and service offering, however, you never pass a Billy Joe Bob's station on your way between work and home. In this case, location has become an important third dimension which is clearly outside the realm of price and quality. Similarly, if your company credit card is not accepted by Billy Joe Bob's, good ol' Billy is not even considered for gas purchases for your company car. Finally, other buying dimensions may include product or service awareness, or other, less tangible brand equity issues.

In many instances, including the so-called third dimension within the quality factor measurements may suffice. In other circumstances, we may wish to limit market share analyses to customer segments or geographic markets that are applicable. Regardless, the important thing to understand is to consider all buying decision criteria in how you design both your measurement instruments and your analyses.

What about customer satisfaction measurement?

While we advise our clients to measure customer value as part of a comprehensive performance measurement program, we suggest they maintain traditional customer satisfaction research as part of their overall customer loyalty effort. Traditional customer satisfaction research has some distinct benefits not always attained through a customer value program. These include:

  • The ability to use an individual customer's comments and scores to take customer-specific actions to improve customer retention. Clearly, this is impossible with CVM programs as they must be unsponsored surveys and as such, respondent names can not be forwarded to our clients.
  • A focus on doing the basics. While an emphasis on customer perceptions in a competitive context is a key advantage to CVM, ignoring the basics such as customer service, repair or service issues is a sure-fire way of losing market share. Customer value research usually covers too many broader factors to gather scores for the operational details (attributes) addressed by traditional customer satisfaction research.

Most of our CVM clients also conduct customer satisfaction research programs to maintain the benefits listed above. Typically, this involves conducting research with customers who have recently completed a transaction with the company, such as a repair or new service order.

Customer satisfaction measurement helps an organization focus on the issues that keep customers happy. Certainly, that is the first step in any customer loyalty effort. However, there is no guarantee that customers will stay loyal just because they are happy with the product/service they receive. What if the competition offers superior quality, or an equal level of service quality at a lower perceived price?

Understanding value perceptions of market offerings as a whole is the only way we can truly understand customer behavior with regard to customer loyalty. Moreover, analyzing both quality and price perceptions of all the competitors in the market through a comprehensive CVM program allows us to develop a big-picture view of the market and develop superior strategies aimed at both keeping current customers and winning new ones.