Every time I fly Delta, I get a survey asking me how likely I am to recommend the airline to friends--a version of the famous Net Promoter Score. Hertz sends a link to a survey after each rental. Amex sends a note after every call to customer service. Hilton does something similar--indeed, sometimes I get an "Is everything all right?" email during a stay. It makes you wonder why customer satisfaction is so elusive when companies are so punctilious in seeking information. 

The reason's easy to see but not easy to act on. Satisfaction is an end product. Like a batting average, it tells you how you did, but not how to swing and hit better. Satisfaction is backward-looking. It doesn't correlate to future behavior. Far more important than a satisfaction score is a deep understanding of your customers' experience and how it manifests itself before, during, and after their time with you--that is, during the entire journey.  

Nine times out of ten, my response to one of those surveys is "It was okay." Would I recommend them to a friend? "Depends on the circumstances." Did the employees treat me well? "They were fine." 

The blanket "meh" that I routinely dispense shrouds a lot of opportunity to make an experience better. Here is how you uncover it that opportunity.  

1. Construct a program to examine past, present, and potential ways to delight customers. 

A Harvard Business Review article by Christopher Meyer and Andre Schwager, "Understanding Customer Experience," advises companies to keep tabs on past satisfaction by using surveys, focus groups, and the like, while probing present relationships with user-experience studies and periodic, structured heart-to-hearts with key customers for example. Also make note of future opportunities with ethnographic studies, focused market research, and pilots with emerging technology. Ten years old this year, the article hasn't been surpassed. 

2. Identify specific customer pain points and fix them--fast. 

Even little irritants can provoke fits in customers, and the situation can escalate if you don't fix the pain points. For example, departing passengers who come to Washington's National Airport via the Metro ought to see a sign listing flights' gate numbers right when they enter the terminal, but the sign is tucked off to the side, several dozen yards away, and not immediately visible. That goof bugs us more each time we see (or don't see) it. 

As Meyer and Schwager point out, it's especially important to remove stumbling blocks that stand in the way of a customer completing his business, whether it's proceeding to checkout online or getting the bill in a busy restaurant.

3. Experiment "in the wild" with customers. 

Very few transactions are simple one-and-done deals, like buying a Snickers in a deli. Most customer journeys have multiple steps and stages, and each may branch off in dozens of directions depending on the customer or occasion. You can't possibly observe them all in the controlled environment of a lab. You've got to get out and watch what real customers do. 

Long before the Internet, Walmart executives got together weekly to compare notes about the effect of promotions, configurations, and other variations across the company's network of stores. Whether such experiments are natural or planned, observing customer behavior will tell you more about them than any satisfaction survey.

4. Find the critical moments and be awesome in those. 

Every company has make-or-break moments with customers. Get them right--no, get them super-right--and you will create the kind of goodwill that cements relationships. Some of the most important moments are when something goes awry. A friend of ours recently ordered a desk from Wayfair.com--not an insignificant purchase. 

It arrived damaged. Her Twitter-finger was itchy but she decided to call first. She got an apology, a full and immediate refund, and, since there was no way to repair or resell the desk, the advice that she give it to a  community group that could use it. Wayfair got a loyal customer.