Understand How Users Think They Do Things

A few weeks ago, I heard an interesting story that did a nice job of revealing the differences between how people think they act and how they actually act. The story was about a new sandwich from a fast-food restaurant.

See, the marketers did lots of research before releasing the idea upon the public. They asked a bunch of people if they would find appealing the idea of ordering a low-carb version of their hottest-selling cheeseburger. Resound-ingly, people said they would, indeed, love to take their usual trip to the establishment and order something they know is good for them and their families. The marketers knew they were on to something. So they whipped up a plan, sent the recipe-makers into action, and released the sandwich, sure that their hours of hard work would pay off and earn the company big dividends.

Reality kicked in a short while later. The sandwich failed miserably and quickly disappeared from the menu.

Why, you ask?

People often don't do what they think they would do. They don't act the way they think they would act. We can talk for hours about how we would respond in any given situation, but we don't really know what will happen when the hypothetical becomes real. The sandwich failed to live up to its promise because the promise was based on meaningless conversations with people who thought they would do the smart, responsible thing and make the healthy choice.

The marketers, I'm sure, didn't mean to have meaningless conversations. It's more likely that they asked leading questions, such as "Would you choose to eat the healthier version when given the choice?" It's a question designed to make the person who says, "No, I wouldn't" feel like an idiot for doing so. It's a question designed to get a "Yes, I would."

And even if the questions were presented in an unbiased way, you can't just walk up to people on the street and ask them what they would do. No one really knows what he or she would do. History shows us that people don't always make the right choices. They make comfortable choices. They make safe choices. More to the point, they make the choices they know how to make.

It's difficult to predict how we'd make decisions in hypothetical situations. Hardly anyone can do this well. When we intentionally begin making better choices, we usually do so in increments, making small improvements for a short time. At some point we find ourselves in a stressful situation and immediately revert. We fall back on the types of choices we've been making our whole lives. The ones we know how to make.

One kid starts screaming or crying, another starts begging ruthlessly for a kid's meal with the shiny new toy from his new favorite cartoon movie, and the money in hand goes toward whatever will get them to settle down. Forget the healthy sandwich, just give me one kid's meal, please.

When people use computers, it's usually because they need to accomplish something. Your application stands between them and their mission. And while many people who use the Web regularly feel pretty confident with it, we all know someonesomeone closewho can never quite figure out whether or not the one-click purchase they just made on Amazon.com is going to ship to the new house or the old one.

Every time I'm about to make a one-click purchase on Amazon, I make about 20 other superfluous clicks first so I can double-check the address and credit card information Amazon has on file. Of course, if someone from Amazon asked me if I would use the one-click checkout, I'd say yes. I wouldn't admit to the 20 other clicks, because I prefer to believe I can trust the one-click checkout process enough to be sure my order will be shipped to me correctly.

I don't feel comfortable making a one-click purchase on Amazon.com, because I don't trust the one-click settings (which means the design is a little too obvious). But I'd never admit that to an Amazon usability specialist if he asked me outright.

People prefer to think they know how they would act in any given situation. But very few of them chose the low-carb cheeseburger.

You can't ask users outright what they want. You get theoretical answers. You don't get the answers that result from real choices in real situations. You don't get the truth about how people think and work.

Now you know.