Satisfaction is relative to competitive alternatives. Improving financial performance lies in understanding how customers rank your business.

New research shows that the relationship between customer satisfaction and customer spending behavior is very weak. The return on investments in trying to increase customer satisfaction is often trivial or even negative.

It turns out that what does matter is how customers rank a brand in satisfaction relative to its competitors.

Why? According to Timothy Keiningham, Sunil Gupta, Lerzan Aksoy and Alexander Buoye, authors of “The High Price of Customer Satisfaction,” in the Spring 2014 issue of MIT Sloan Management Review , single-brand loyalty, which was common a generation ago, has been replaced in many sectors with loyalty to multiple brands.

“Because of this divided loyalty, more customers partially defect (in other words, they give more of their business to a competitor) than completely defect from a business or brand,” the authors write.

That’s a critical piece of information. It means that there’s important upside to be gained from current customers.

“Improving customers’ share of spending with your brand tends to represent a far greater opportunity than efforts to improve customer retention,” the authors write.

As an example, the authors cite a study of the hotel industry by Deloitte that found that, on average, approximately 50% of hotel guests’ spending is with hotels other than their most preferred brand. Similarly, a study of the banking industry by McKinsey found that, on average, only 5% of bank customers actually close their accounts each year and that, instead, 35% of customers reduce their share of deposits. It’s that reduction in engagement, rather than the out and out departure, that represents a far bigger loss to banks – and, similarly, to retailers facing the same kind of migration of business.

So what to do? That’s not an easy question to answer. “Unfortunately, because of the weak relationship between satisfaction and share of wallet, managers are typically unable to identify what their companies must do to capture a greater share of customers’ spending,” write the authors.

The problem with looking solely at a company’s satisfaction is that “it is a poor indicator of the relative preference that customers have toward the brands they use,” they write. The authors put it this way: image Janet and John use three brands: yours, plus Brand A and Brand B. Both rate their satisfaction with your brand a “9” on a 10-point scale (where 10 is the highest). Janet rates her level of satisfaction with Brand A “9” and her satisfaction with Brand B “10.” John rates his level of satisfaction with Brand A “7” and Brand B “8.”

“Even though Janet and John both rate your brand a ‘9,’ your brand is John’s clear first choice,” explain the authors. “For Janet, your brand is tied for last.” Result: John will allocate a substantially higher share of his category spending with your brand than Janet does.

“While the idea of ranked satisfaction might seem radical, it actually builds upon what researchers have known for a long time: Satisfaction is relative to competitive alternatives,” write the authors. “In fact, our research finds that the relationship between satisfaction and share of wallet is a function of satisfaction’s relationship to the relative rank.”

This article draws from “The High Price of Customer Satisfaction,” by Timothy Keiningham (Ipsos Loyalty), Sunil Gupta (Harvard Business School), Lerzan Aksoy (Fordham University) and Alexander Buoye (Ipsos Loyalty), which appeared in the Spring 2014 issue of MIT Sloan Management Review . (Fair ribbons image courtesy of Flickr user Susan Ujka.)