At first glance, creating an army of satisfied customers seems an obvious way to build a business. But as a leading computer software company has learned to its surprise, satisfied customers aren’t necessarily good customers. Indeed, the company discovered in a recent survey that there was no correlation between customers’ satisfaction scores and their actual purchase behavior.
Why are customers who say they’re satisfied not necessarily repeat customers? Because satisfaction is a measure of what people say, whereas loyalty is a measure of what they actually do. Many managers still don’t recognize this fundamental difference, so they use customer satisfaction and customer loyalty interchangeably, as though they were synonyms.
What customers report in satisfaction surveys is their attitude, which usually reflects their recent experience in transactions with the business in question. The survey and the report it generates take the temperature of customer feelings about events that have already occurred. The importance that businesses ascribe to these surveys is profound. Indeed, measuring customer attitudes has become an industry in its own right, and such groups as J.D. Power and Associates, the University of Michigan Business School’s National Quality Research Center, the Customer Satisfaction Institute, and other similar organizations have become powerful shapers of business practices.
There’s no question that satisfaction measurements can be valuable. They allow customers to vent frustrations. They can highlight problems with product quality and customer service. But satisfaction surveys also have limitations. The larger the customer base, the more expensive and time-consuming it can be to survey. Because of the time and expense they require, surveys can be conducted only periodically, which means they may not reflect current attitudes. Additionally, surveys cannot include all customers — and results can be biased when customers either are excluded or don’t bother to respond. Most important, surveys measure opinion and are not reliable predictors of future behavior. Even surveys that ask customers about their intentions do not necessarily shed light on the future because customers don’t always do what they say they’ll do.
Loyalty (be it to a king, a brand, or a relationship) is most definitely not a matter of opinion. It is a measure of commitment and a strong indicator of future behavior. In a business setting, sales data (such as the transaction date, amount, and product description) can be used to profile customers’ past behavior, and can be a reliable basis for predicting their future actions. If, for example, past measurement shows that the Ajax Partnership has been buying supplies regularly every three months for the past two years, and then it begins purchasing smaller amounts at less frequent intervals, you can be fairly certain that Ajax’s loyalty is at risk.
In a small organization with few customers, this kind of behavior measurement is usually a matter of eyeballing the records. There are, however, new and sophisticated mathematical techniques that allow an enterprise with hundreds of thousands (or even millions) of customers to extract data automatically from accounting databases and convert it into an early warning system that segments customers on the basis of their loyalty profiles, and then identifies potential defectors. In this way, enterprise accounting records can be transformed into valuable marketing intelligence.
Loyalty profiles can predict defections and the amount of revenue that will be lost as a result of those defections. Loyalty measurement can also identify when customers will buy next, what they’re likely to buy, and how much revenue these sales will generate. It can identify the customers who are likely candidates to buy more than they now do, and predict how much enterprise revenues will grow if these candidates can be upgraded.