Reprinted with permission from Marketing Research, Spring 2005, published by the American Marketing Association.

By Rob Kroenert, Leah Spalding, Brian Cooper, and Liz Le

Making the Link

A customer loyalty index can uncover the link between financial performance and customer loyalty.
MOST BUSINESS EXECUTIVES WILL AGREE THAT CUSTOMER loyalty has a major effect on the financial performance of their company. Acquiring new customers is typically much more expensive than maintaining existing customers, and customers tend to become more profitable the longer they stay with a company. In addition, loyal, satisfied customers are more likely to generate new business by providing positive references to friends and colleagues. In his 2001 book, The Loyalty Effect, Frederick Reichheld contends that customer loyalty can create tremendous competitive advantage, boost employee morale, produce unexpected bonuses in productivity and growth, even reduce the cost of capital. But demonstrating the link between something as ambiguous as loyalty and something as specific as revenue or profit can be

marketing research 21

Executive Summar y
It is difficult to demonstrate the link between something as ambiguous as customer loyalty and something as specific as company revenue. Because of this, many companies lack the information necessary for critical business decisions. An effective and widely applicable method of measuring customer loyalty combines three elementsreferenceability, repurchase intentions, and future purchase levelsto create a customer loyalty index. This can then be used to relate customer loyalty to a companys financial performance.

Future purchase levels. Does the customer plan to spend more, less, or maintain current spending levels in the future? Perception of competitive advantage. Does the customer feel that the companys products provide a competitive advantage? Satisfaction. How satisfied is the customer with the product or company? The best approaches to measuring loyalty are directly related to desirable customer behaviors or attitudes as well as the surrounding context (the presence or absence of competitors, barriers to entering and exiting the relationship, etc.). Ideally, the approach chosen for a specific company should be universal enough to apply to all relevant customers across a wide variety of situations. If a company with more than one product chooses an approach that is product specific, for example, it can prevent an apples-to-apples comparison of customer loyalty across the entire product line. And comparing customer loyalty across different companies requires an approach that is general enough to encompass an even wider variety of situations. Based on more than 20 years of experience conducting primary research in the information technology (IT) industry, TNS Prognostics has found that one of the most effective and widely applicable methods of measuring customer loyalty for our clients involves three of the elements listed above: referenceability, repurchase intentions, and future purchase levels. Many of the customer surveys we conduct for our clients include a question about each of those three elements, and the answers to those questions are then combined to create a customer loyalty index (CLI). The standard referenceability question we use is: Do you consider yourself a favorable reference for <company name>? with answer choices of yes, no, and unsure. The repurchase intention question is: Based upon your experience, are you very likely to continue purchasing <company names> products? with answer choices of yes and no. The future purchase levels question is: Do you expect future purchase levels of <company names> products will be with answer choices of higher, lower, and same. CLI is calculated as the percentage of respondents who answer positively to each of the three questionsyes to referenceability, yes to repurchase intentions, and higher or same to future purchase levels. To be included in the CLI calculation, a respondent must have answered all three questions. We have found the CLI measure to be both internally reliable and stable over time, with a Cronbachs alpha (a tool for measuring the reliability of scales) of .77 and an average test-retest correlation of .64 over the last three years (all with p < .06). The questions used to create the CLI are included in an annual benchmarking survey conducted by TNS Prognostics. The benchmarking survey targets a wide range of business customers in the IT industry and uses a standardized questionnaire. Each respondent is asked to provide feedback on a single company, and respondents are screened out of the survey if they are not current customers of that company. The results of

extremely difficult. One of the biggest challenges is the wide range of factors not directly related to loyalty that can influence the financial performance of a company. The overall condition of the economy, for example, can have a dramatic effect on a companys revenue. Although companies with a highly loyal customer base may be better positioned to prosper in tough times, customer loyalty does not make a company completely immune to the ups and downs of the economy. As a result, many companies do not have all the information they need to support critical business decisions. Without understanding the relationship between customer loyalty and financial performance, how can a company decide whether or not to invest valuable resources in initiatives designed to improve loyalty? And once a company decides to invest in improving loyalty, what types of initiatives are likely to have the greatest impact?

Measuring Customer Loyalty

No single definition of customer loyalty applies to all companies and industries. To an investment banking firm, for example, loyal customers might be businesses that use their investment banking firm exclusively and plan to continue doing so, while a soft drink company might consider loyal customers to be consumers who are more likely to choose their soft drink over any other, even if they also buy other soft drinks. The definition of a loyal customer can encompass a wide variety of attitudes and behaviors, and the best definition to use depends on the specific situation. Elements of customer loyalty may include: Psychological commitment. Does the customer feel an emotional connection to the product or company? Referenceability. Does the customer consider himself a positive reference for the product or company? Purchase history. Has the customer purchased the product (or from the company) consistently in the past? Repurchase intentions. Does the customer plan to buy the product (or buy from the company) again in the future?
22

Spring 2005

this annual benchmarking survey provide a rich source of historical customer loyalty data for a wide range of companies in the IT industry.

Investigating Links

The first step in our effort to uncover potential links between customer loyalty and financial data was to gather and organize historical customer loyalty data from the TNS Prognostics benchmarking survey. The specific IT companies included in our benchmarking survey change slightly from year to year, due in many cases to mergers/acquisitions, but each year the survey focuses on the hardware products of approximately 10 companies and the software products of approximately 20 companies. Given the greater number of companies covered in the softtomer loyalty measures and revenue growth over the following ware benchmarking, those results became the focus of our year (2002-2003). investigation, and we started with the 2002 results. In 2002, the software benchmarking results included 19 different companies, with an average of 301 survey responCorrelating Growth dents per company. The companies ranged in size from $90 The most compelling finding was the strong correlation million to more than $50 billion in gross annual revenue, and between CLI and revenue growth over the following year. (See all but two were publicly traded. Most were primarily softExhibit 1.) The analysis showed a significant positive relationware companies, but several sold hardware in addition to softship between CLI and next years revenue growth (r = .79, ware. For each of these companies, we gathered the survey p < .01), indicating that the companies with more loyal cusresponses for all three of the questions used to create the CLI. tomers in 2002 tended to experience higher revenue growth Our next step was to gather historical financial perforthe following year. More specifically, the companies in our mance data for the target companies. Annual revenue growth, dataset show an average 12.5% gain in annual revenue which is widely reported in public financial reports and less growth for every 10% increase in customers who are loyal. dependent on costs that may be unrelated to customer loyalty, The relationship also holds true across companies of various was selected as our primary financial performance metric. sizes, as the magnitude of the correlation remains fundamenFour of the 19 target companies either did not publicly report tally unchanged even when controlling for differences in their gross annual revenue or were acquired during the time annual company revenue (r = .80, p < .01). period we were investigating, and those four companies were therefore excluded from our analysis. For the remaining 15 Exhibit 1 Correlating loyalty with future growth target companies (5,242 total respondents), we calculated the annual revenue growth percentage for two time periods: from the year before the survey to the year of the survey, and from the year of the survey to the year following the survey. We then had 2002 customer loyalty data alongside revenue growth data for 20012002 and 2002-2003, which provided the core information for our analysis. We began by correlating the customer loyalty data to the financial data to measure the strength of their relationship. CLI and its three individual components (referenceability, repurchase intentions, and future purchase levels) were correlated to the revenue growth percentage for both time periods (2001-2002 and 2002-2003), and our analysis uncovered a significant 2002 customer loyalty index relationship between several of the cus-

he companies in our dataset show an average 12.5% gain in annual revenue growth for every 10% increase in customers who are loyal.

marketing research 23

association; for example, the relationship between loyalty and financial performance Exhibit 2 Correlating referenceability with future revenue growth might ultimately be driven by financials, as would be the case if a financially healthy company had the resources to invest in building better customer relationships. If that was truly the causal relationship, however, we would expect to see stronger correlations between customer loyalty and revenue growth over the year preceding data collection, assuming there is not a lag between a year of solid growth and a companys decision to invest more resources in customer loyalty. That correlation, although positive, was considerably weaker than the correlation with revenue growth over the year following the data collection (r = .41, p < .13). The pattern of these correlations supports the argument that loyalty drives revenue growth, rather than the reverse. 2002 referenceability Our findings certainly do not change the fact that customer loyalty is just one of the important factors that influence financial performance. Other external factors, such as the general state Of the three individual components of CLI, referenceability of the economy or product innovation, can also have a major was found to have the strongest correlation with revenue influence on the relationship between loyalty and revenue growth over the following year. (See Exhibit 2, which shows growth. From 2001-2002, for example, average revenue revenue growth correlated to the percentage of people who growth among the companies we surveyed in the IT industry answered yes to the referenceability question). But the correlawas much lower than it was from 2002-2003, despite the fact tion between revenue growth and referenceability, though that the average CLI in 2001 was only four percentage points positive and significant (r = .68, p < .01), is weaker than the lower than it was in 2002. Even the companies that maincorrelation between revenue growth and CLI. A positive but tained strong loyalty scores were not immune to the economic weaker relationship is found for the correlation between revdownturn, although they were better positioned to prosper enue growth and repurchase intentions (r = .48, p < .08) and when the economy eventually picked up. net future purchase levels (r = .41, p < .13, measuring the percentage of respondents answering higher minus the percentage answering lower). The benchmarking surveys also Linking Up asked respondents to rate their overall satisfaction with the Completing the link between financial performance and total performance of the company on a 0-10 scale, and the customer loyalty requires an understanding of the specific correlation between revenue growth and the average total performance rating was also positive (r = .49, p < .06), but again not as strong as the correlation with CLI. Based on this data, there is clearly a strong relationship between the loyalty of the companies customer bases and the revenue growth of those companies over the next year. Although referenceability appears to be a very important component of loyalty, the CLI measure, which also includes repurchase intentions and future purchase levels, appears to be even more closely connected to revenue growth. Given that our analysis was based on correlations, we cannot draw definitive conclusions about the causal nature of the relationship between customer loyalty and financial performance. There are alternative ways to describe this

here is clearly a strong relationship between the loyalty of the companies customer bases and the revenue growth of those companies over the next year.

24

Spring 2005

Exhibit 3 Customer loyalty drivers

Revenue growth

Economic factors

Customer loyalty index

Other factors

Quality of products

Quality of service/support

Quality of prof. services

Ease of doing business

Value received relative to price

Quality of sales process

Service driver #1 Service driver #2 Service driver #3

Service driver #1 Service driver #2 Service driver #3

Service driver #1 Service driver #2 Service driver #3

Service driver #1 Service driver #2 Service driver #3

Service driver #1 Service driver #2 Service driver #3

Service driver #1 Service driver #2 Service driver #3

issues that drive customers to be loyal or disloyal. Without this understanding, a company cannot determine the specific actions that are likely to lead to the greatest loyalty improvements. Measuring CLI over time can help a company understand how loyal its customers are, but CLI by itself does not reveal the factors that cause a companys customers to be loyal. At one company, for example, improving product quality may lead to a dramatic increase in customer loyalty. At another company, improving the quality of service/support might drive the greatest improvement in customer loyalty. To make things even more complex, the issues that have the biggest impact on the loyalty of a companys customer base will almost certainly change over time. When TNS Prognostics measures CLI for a company, we typically also ask a series of questions that measure customer satisfaction across a wide range of issues. For most companies we first measure satisfaction across high level issues (e.g., overall quality of service/support), and then we measure satisfaction across specific issues that roll up into each high level issue (e.g., timeliness of responses to service requests). These measurements can then be analyzed with CLI data to identify the issuesboth high level and specificthat are the primary drivers of customer loyalty. (See Exhibit 3.)

areas. To ensure that action is taken, companies should then incorporate the implementation of their improvement initiatives into companywide and individual objectives for the coming year. Consideration should also be given to linking customer loyalty to employee compensation. Effective companies will also track their progress by measuring customer loyalty again after their action plans have been implemented. But all too often companies fail to take action. Improving customer loyalty can require a significant investment of both time and money, and many companies are hesitant to make that investment without knowing how much they expect to get back in return. For the companies included in our analysis, we found that a 10% increase in customer loyalty was associated with an additional 12.5% increase in annual revenue growth. Armed with that type of direct link to financial performance, companies are in a much better position to build a powerful business case for investing in customer loyalty. Additional Reading Reichheld, F. and T. Teal. (2001), The Loyalty Effect: The Hidden Force Behind Growth, Profits, and Lasting Value, Cambridge, Mass.: Harvard Business School Press.

Investing in Loyalty
When companies know exactly which issues drive the loyalty of their customers, they are in an excellent position to take action and make real improvements. Equipped with an understanding of key loyalty drivers, companies can begin to take action by assembling cross-functional teams that develop specific initiatives designed to make improvements in key

Rob Kroenert is the director of consulting services at TNS Prognostics in Palo Alto, Calif. He may be reached at [email protected]. Leah Spalding, Brian Cooper, and Liz Le are all research consultants at TNS Prognostics. They may be reached at [email protected], [email protected], and [email protected].

marketing research 25