How to Verify Customer Intent with Split-Testing

Senior copywriter Jennifer Hoppe wrote last week about how we can now track data about customer behavior like never before and use it to focus on conversions, conversions, conversions.

At the risk of wearing out the Mad Men references, Sunday’s episode reinforced the point. Whereas Don Draper aims for creative as a solution to driving customer behavior, in this case a consumer research agency is pitching Don’s advertising agency on the value of its psychological research for developing campaigns. As they point out, there is a difference between what people say they want—and what they do.

With split-testing and conversion tracking technology, we don’t have to guess what they do or figure it out after the fact; we can see it in action.

Let’s take a look at how we’ve put this technology to work.

We’ve been reviewing a lot of research that indicates that using a landing page headline that closely matches language in an ad will convert the best. We wanted to confirm this best practice with a landing page campaign we’ve been running for one of our clients.

Using the Google Website Optimizer split-testing tool, we sent visitors to the original landing page (the control), as well as multiple versions of the landing page, each of which had varying headline copy but identical designs.

The test ran for several days until the tool could collect statistically significant results. At that point, we were able to compare the conversion rates of each and, in fact, there was no comparison. The matched headline version won, hands down.

With these numbers in hand, we briefed the client on the test results and rolled out the winning page to all visitors for maximum conversions overall. Since then, we’ve also used it as a basis for future landing pages.

It just goes to show that it takes a little of everything to get to a winning result: customer research and good old creative know-how up front to generate a control, industry research to determine what to test, and an optimization tool to find the best-converting combination.

Your Turn

  • What have you learned about customer behavior through testing?
  • When has testing challenged an assumption you had about customers?
  • What would you like to understand better about your customers?
Scroll To Top