Will Link Churn Remove the B from A/B Testing?

Posted by dave on January 17, 2006 at 03:09 PM

Which is better 1 or 2? 2 or 3? 2 or 4? As you sit in the optometrist's chair blurry eyed and dilated, a marketing marvel is happening to you: A/B testing. Your doctor of vision is building on a proven practice, which they know will help you reach the goal of clearer sight.

A/B testing for web sites works in the same way. It tests and builds on what marketers already know, to come up with a conclusion that helps focus on a specific goal.

Simple and effective, A/B testing utilizes a constant from which aspects of a page, such as the image and link text, can be tested. To be truly successful, A/B testing may involve changing link text or an image multiple times until that eureka has been reached and users begin to convert.

But, with the ambiguous advent of Link Churn, could Google be threatening this form of testing?

Point 59 of a recent patent application states that adjusting linking in the short-term may result in penalization. If this occurs, will A/B testing become a bad-business practice? Is Google essentially saying that we have to choose one link and be bound to that link for months? Which is better, 1 or 1?

So, is the sky falling? Is Google going to start to penalize the sites that utilize A/B testing? Is this once innovative and proven Internet marketing tactic going to go the way of the Dodo because of Link Churn? Not yet. Google understands the importance of conversion testing. In fact, in their own analytics tool they have a section dedicated to conversions. A/B testing still offers you with a way to measure the actual behavior of your real-world customers with confidence.

In short, keep testing and keep learning. Your bottom line will thank you for it.