In high school, we had a class project to introduce a new product in our school store. We decided to launch a branded beanie as the weather was turning cold and football season was in full swing.

As part of our market research, we surveyed students to determine color preferences. We showed color samples and the overwhelming response was for blue beanies with a minority wanting white.

As you would expect, we ordered a lot of the blue and very little of the white. On the day we began selling the beanies, we were shocked to see that the white ones were selling like hotcakes and nobody wanted the blue ones. In fact, we sold out of the white within days. Parents were showing up at the school trying to buy as many as they could for their entire family.

We ended up placing several orders trying to keep up with demand, which we never did. And the blue ones? We discounted them almost to cost in order to empty the inventory.

Why did our market research fail? And what’s the solution?

We showed color samples to students. And yet, they acted very differently when it came time to pony up cash and buy.

You might try to put consumers in buying mode when they’re being surveyed or interviewed in a focus group, but consumers act differently than they think/say they do. In part, we all have the person we hope to be and answer based on that ideal vision of ourselves.

The solution is to test early and often gaining invaluable and otherwise unobtainable insight. (Click to tweet this quote)

In fact, companies such as Intuit are built on the premise of constant, little tests. Many meetings end with the boss asking, “So how can we quickly test your hypothesis?”

Related:  Does firm-created word of mouth or user-generated content drive purchase intent?

Google is constantly running tests with its search algorithm – SEOMoz reports 500-600 changes a year. You’ve most likely been a subject of at least one of Google’s tests but didn’t know it.

Don’t think that consumer testing has to be large, expensive or cumbersome

There are so many ways to quickly gauge real responses. For example, send an email to your subscribers advertising that new product or service you’ve been thinking about. See how many people click to sign-up, buy or get more info. Then, use the landing page to say something such as, “Thank you for registering your interest in Product XYZ. We’ll let you know as development progresses.”

Stop for a moment and ask yourself, “What theory / improvement / change / idea / hypothesis can I test today?” And now find a quick, cheap way to do some consumer testing.