The Mystery of the Missing Loyalty Effect

It would be nice if the link between customer loyalty and profit were simple, clear and easy to prove. The theory is certainly attractive: Customers with high satisfaction and commitment levels form a bond with a company, and this bond leads to a variety of desirable behaviors. Loyal customers stay longer, spend more, complain less and make recommendations to their friends and family. If companies invest in loyalty, the theory goes, the rewards will come pouring in.

In practice, however, this scenario doesn’t always play out as predicted. The loyalty-profit connection can be complex and muddled, as the following case demonstrates:

At a large retail bank (which we will call Bank Z) analysts from the Finance Department teamed up with the Service Quality Group to examine the relationship between customer loyalty and profitability. The Finance analysts calculated profit at the individual household level. Service Quality managed a large survey program and calculated household loyalty scores, which were based on a combination of satisfaction, commitment, advocacy and brand identification ratings.

The teams linked the profit measures and loyalty scores, with the expectation that they would discover a nice, upward-sloping line showing that the least loyal customers made less money for the bank than the most loyal customers. What they found, instead, was a flat line. There was no difference in the average loyalty scores between low-profit and high-profit customers.

This finding caused considerable consternation. Bank Z spent a great deal of money on programs to measure and improve customer loyalty, and the loyalty metric was a key part of the executive incentive structure. If there was no return on these investments, what was the point of making them?

Further analysis revealed there were at least three factors confusing the results:

1. Competing Segments. When examined more closely, it was found that there were two large, distinct groups among the most profitable households. The first consisted of high-income, older customers who had long tenure with the bank. These customers generally used multiple products and services, and kept large balances in their deposit accounts. Their loyalty scores, as expected, were extremely high.

The second segment looked very different. This group consisted of younger, low-income customers who typically held a single free-checking account with a low average balance. However, they bounced a lot of checks, which brought in significant revenue from overdraft fees. Unlike the first segment, these customers had exceptionally low loyalty scores. In fact, they hated the bank. The only reason they stayed was that the situation would be the same wherever they went.

Taken together, the loyalty scores of these high-profit groups cancelled each other out.

But what about the low-profit customers? As it turns out, there were competing segments in that group, as well.

One segment consisted of customers who were highly dissatisfied with the bank. They had experienced errors or service problems, and consequently had begun shifting their business to competitors, resulting in lower profit levels. As one would expect, their loyalty scores were abysmal.

The second segment consisted of renters with modest but steady income. They did not have mortgages or equity loans, kept fairly low account balances, used free checking, and never had overdraft fees. In other words, they were stable, financially responsible people – and they made no money for the bank. They were also among the bank’s most “loyal” customers. And why not? They were getting a valuable service for free.

The existence of competing segments at the opposite ends of the profit spectrum may in itself have been enough to flatten out the loyalty curve. But there were other factors at play, as well.

2. Imprecise Profit Calculations. Profit, as we all learned in school, is calculated by subtracting cost from revenue. Companies are usually pretty accurate when it comes to assigning revenue to individual accounts. Banks, for example, know to the penny how much someone has paid in fees and interest.

But costs, particularly in service businesses, can be more difficult to assign to individual customers. This is important, because according to some experts, loyal customers are cheaper to serve. By staying with the company longer they learn more about its products and services, and they develop more realistic expectations about what the company can do. As a consequence they become less reliant on support services and less likely to complain. In addition, with time and knowledge they may migrate from high-cost to low-cost channel usage. In the case of banks, that means using ATMs and on-line banking instead of relying on tellers and call agents.

But if loyal customers at Bank Z were cheaper to serve, that fact was not reflected in the profit calculations for individual households. The Finance Department ignored such details; they simply took an average cost-to-serve for all accounts and applied it evenly, irrespective of the individual households’ actual behaviors.

Some of the cost details were, in fact, available at the household level. But the information was scattered about in various databases and was controlled by Marketing, Operations and other groups. It would simply have been too difficult for Finance to find and extract all the information needed to get a true picture of household service costs, so they took a shortcut – and consequently masked any insight about cost-related loyalty behaviors that might have been contained in the household profit data.

3. Financial Benefits Outside The Formula.
Not all of the predicted financial benefits of loyalty are included in the standard profit formula, at least at the level of the individual customer or household. One example is retention – or its flip-side, attrition. At Bank Z, those who severed their relationship with the company because of dissatisfaction or lack of loyalty were removed from the household profit figures altogether. The revenue lost from their defection would impact profitability at an aggregate level, but as individual households they had effectively ceased to exist, and would not have been included in the loyalty-profit analysis.

Another supposed loyalty benefit is new account generation from referrals. Analysis of customer survey data at Bank Z indicated a strong correlation between a customer’s satisfaction and the number of referrals given. Furthermore, about 40% of new customers said they had joined the bank primarily because of a referral from an existing customer. But household-level profit calculations have no way of accounting for the value of bringing in new business, and this benefit would, like retention, have been excluded from the loyalty-profit analysis.

Bank Z learned more than one lesson from this exercise. First, it stopped treating loyalty as a one-size-fits-all proposition, and began looking at how best to leverage loyalty-driven behaviors within specific customer segments. Second, it started searching out and consolidating cost-to-serve data from across the organization (an effort that is still far from complete). And third, it designed a loyalty benefits calculator that incorporated a wider range of variables than the profit formula used by the analysts in Finance. Even with these changes the loyalty-profit picture remains fuzzy, but it’s getting clearer all the time.

If you would like a PDF of this case study, please contact us.

Originally published in:

Loyalty Management Magazine, July 2009