Implement Website Self-Service to Focus on What Your Customers Actually Do

We all know how important it is to understand our customers. We spend loads of money creating personas, anticipating wants and needs, conducting surveys and focus groups, and trying to “get inside the heads” of our target buyers. In light of these efforts, it is surprising (and sad) to reflect on just how little we know about what our customers actually do with our products. Sure, we think we know what they do, but how reliable are our intuitions? It can be hard to observe actual behavior, but it is of absolute importance if we’re to create successful products and services.

The importance of understanding customers’ behavior


People are notoriously bad at predicting what they want or would use. The most unreliable question a product marketer can ask is, “Would you use this?” Steve Jobs knew this well. But it is a mistake to think that Jobs was not customer-centered in his approach to innovation. He was a keen student of behavior, paying careful attention to customers’ problems and how his products were being used. In other words, he understood what his customers did

Despite Jobs’ reputation for “I know best,” there is humility in striving to understand your actual customers’ behavior. To do so means not to assume you know best, but to trust that your customers’ behavior can reveal important insights.

What customers can and can’t do for you

First, what customers can’t do for you: 

  • Customers can’t predict what they’ll like (or dislike). People do not know the future, even when that future is about their own tastes. Focus groups, interviews, and surveys shouldn’t attempt to discover preferences about things customers have never authentically experienced.

  • Customers can’t predict what they’ll use. Not only are people poor at predicting preferences, they are poor at predicting what they’ll find useful. They also can’t predict how they will use a thing, especially given that many things are used in unexpected ways.

  • Customers can’t substitute for designers. Customers will often tell you how to design your product or service. Don’t take their suggestions; instead, uncover their underlying needs. If they say, “You should make this blue,” find out why. Is it to be more noticeable? Or to match their website colors? Is blue is their boss’s favorite color? Each of these reasons has different implications for design. Uncover customers’ needs, and then meet those needs the best way possible.

Second, what customers can do for you:

  • Customers can tell you what they already like (or dislike). People can reflect on what they know, and they know whether they like or don’t like something they’ve experienced.

  • Customers can compare among alternatives. People can’t tell you much about an idea presented in isolation, but they can say whether they prefer A, B, or C when placed side-by-side.

  • Customers can take action. The most reliable signal from customers is what they actually do. It is more reliable than anything customers say. Unfortunately, it is also often the hardest signal to get.

So how do you discover what customers actually do?

Create conditions to observe actual use


There is a well-known design adage about how to determine where to lay a concrete path leading to the front door of a new house. Instead of thinking hard about where people ought to walk, create the conditions to observe where they do walk. How? Don’t pour concrete. Lay down a lawn. Then wait for a path in the grass to get worn and pave it.

The lesson here is to sink your energies not into upfront analysis, but into creating the conditions for observing what your customers do. Upfront analysis has its place, but humans are generally poor at predicting other humans’ behavior. A major cause of that inability is that we all have expert blind spots about things with which we’re familiar. We know too much, and that puts distance between us and our customers.


The need for data and patience

Creating the conditions to observe actual use is problem-dependent and can be done in a variety of ways, but in all cases, there are two things you must have: data and patience. It goes without saying that any data collected must result in something observable, ideally a trend or pattern. And as with the worn path in the lawn, it can take time—and patience—for patterns to emerge. But once you observe those patterns, you’ll have a reliable signal as to what your customers actually do. Such a signal is priceless. It is far more valuable than what you think they do, or what they say they do. Both are easier to come by, but often misleading.

AnswerDash’s example


AnswerDash takes a nontraditional approach to capturing question-and-answer content for website self-service. Rather than building a knowledgebase or FAQ with questions seeded by AnswerDash or the adopting company, AnswerDash provides a means for visitors to websites to submit the questions they actually have, right where they actually have them. Once answered by the adopting company, those questions and answers then live in a contextual knowledgebase that’s accessible via self-service by any future website visitor. In this way, questions and answers are “grown,” not “built,” reflecting actual use by actual users. Often, the most commonly accessed questions and answers surprise even the most data-driven website managers, enabling them to improve their sites by revealing their customers’ actual behavior.