The Strategist June 19, 2009, 11:40AM EST

The Customer Satisfaction Survey Snag

Satisfaction isn't enough. If you mean to beat the competition, your surveys should measure customer loyalty

Story Tools

Customer satisfaction has been a major business buzz phrase for more than a decade. In his 2001 book, The Loyalty Effect, Fred Reichheld claimed that customer loyalty is a dominant determinant of success in business. He purported, for example, that a 5% improvement in customer retention for an advertising agency can create a 95% increase in average profit per customer over time. He claims similar (though less extreme) effects in most other industries.

To address the loyalty question, consultants often persuade companies to spend time and money trying to ensure that they know precisely how pleased their customers are. Theoretically, this is vitally important information. Empirically, however, it can be of little worth. In our experience, many large companies kid themselves about how accurate a picture they really get. So how do you get truly worthwhile feedback?

A survey conducted by a client of mine, a large phone company, revealed that 94% of customers were "completely satisfied" with their experience. However, in a separate, concurrent survey conducted by the same company, 30% of customers claimed that given the option, they would switch to a new provider. The former survey indicated only 6% of their entire customer base is less than completely satisfied. Assuming those 6% were among those who would switch, at least 24% of their customers who claimed complete satisfaction felt no loyalty to the company.

The problem is, companies are interpreting satisfaction to mean loyalty. Sure, a customer may be satisfied, but if the customer believes she would be equally (or better) satisfied with any other provider, she'll switch. A new scale is required to measure loyalty. In this new scale, the middle answer is "Perfect. I have no complaints. On the other hand, I didn't feel it was anything special." Frankly, this is likely what most of your customers actually think about you. This level indicates that you can expect an average rate of attrition from this group. All else being equal, you will lose your customers at the same rate as your competitors.

Better than the Rest?

The next-higher rating should be "Better than I could expect from another provider." Here, loyalty begins to assert itself. However, be careful interpreting results in these tough economic times. Recent University of Michigan surveys on the fast-food industry show that the recession is causing consumers to be willing to reduce their own "total satisfaction" in favor of "purely the lowest price."

If you want to measure the ultimate test of loyalty—the willingness to tell friends—the top rating should be "This was so great I will mention it in conversation later today." (Remember that your customer has lots of things to talk to friends about—a referral only comes when your product makes it to the top of that list.)

As anyone who has had a car repaired lately knows, as you leave the dealership, the salesman lets you know that you will be getting a call about feedback. He also informs you that the company's goal is a perfect score, that anything less is considered unacceptable, and asks if there's anything he can do to make that happen. While the guidance is disguised as an offer of service, you know very clearly what they really care about is getting a good score. When you indulge the company that way, its employees never learns that you really didn't think they were anything special.

Long ago, research about marketing showed that customers will phrase their reactions in whatever way you coach them to. Coaching is insidious because it teaches your customers to phrase "I have no complaints" in a way that you interpret as "I loved it and could not possibly be happier." Employees at one hospital recently achieved the pinnacle of coaching. They were given preprinted cards providing guidance on how to answer any survey questions ("we want to be a perfect five")—just to avoid any misunderstanding on our part as to what box to check.