On the Web we need to measure whether customers are actually successful, not whether they are satisfied. 

Professor Jim Saker of Loughborough University writes about a presentation he was at where “the announcement was made that the franchise customer satisfaction index had gone up by a whole five percentage points in the past year. There was wild applause and self-congratulation until it was pointed out on the next slide that sales and profitability had gone down by an even larger percentage.”

I remember once testing a website with a woman who had a really hard time completing basic tasks. At the end of the test, I asked her how satisfied she was with the website. “Really satisfied,” she replied.

This sort of response is not unusual. The question is: what will most affect a customers’ likelihood of returning to a website? The fact that they were able to do what they needed to do quickly? Or the fact that even though they failed miserably at their tasks, they told the nice researcher that it was a nice site?

In an article entitled "Stop Delighting the Customer," Matthew Dixon, Lara Ponomareff and Anastasia Milgramm explain how a division of American Express found that “customer retention rates remained flat even as satisfaction rates rose.”

Often, satisfaction is not a good measure of whether a customer will do more business with you. However, it was found that American Express customers who experienced “low-effort or effortless experiences were more likely to complete future transactions than customers who had high-effort experiences.”

Reliant, an American energy company, became committed to reducing customer effort. As part of this new strategy, “Reliant removed the talk-time metric,” according to Dixon et al. This metric measures how long a customer support rep stays on a call. The result of such a metric is that “reps paid more attention to the clock than to customers' concerns, rushing customers off the phone when they had hit their limit and leaving the customer without full resolution.” By instead focusing on reducing customer effort, Reliant has significantly improved key metrics such as first contact resolution rates.

The equivalent of ‘watching the clock’ in the web content world is a culture of producing large volumes of low quality content. A great many content authors get paid based on the quantity of content they publish rather than on how many customer problems it helps solve.

Recent studies have shown that satisfaction with Facebook compares very badly with other social sites such as Google+. Does this mean Facebook is much worse than Google+? Or does it mean that when you have almost a billion members it’s hard to keep them all feeling satisfied? McDonalds has the same customer satisfaction problem. And we all know, of course, that nobody is satisfied with Ryanair. But if there is such dissatisfaction with Facebook, McDonalds and Ryanair, why are so many people still using them?

In many situations, satisfaction is not a good measure of future customer behavior.