SEO A/B Testing - Issue 49

SEO testing to measure the benefit of search engine optimization techniques

The truth is, there is no such thing as A/B testing for SEO. A/B testing is a cookie-based approach to showing different website treatments to different site visitors. In the world of organic search, however, there is really only one user, and that’s Google. So how can we determine the true incremental benefit of an SEO strategy?

An organic test requires a more permanent change to the content, but instead of varying the user experience, create a Test & Control approach where a similar set of products gets enhanced and the results are compared to a similar set of products that remain unchanged.

We start by removing both the best performing and worst performing products and focus on the average products that are representative of the site overall. Next, products get randomly assigned to either the test or the control group.

Page-level web analytics performance is then tracked for both product groups. By comparing the change before and after it happens, we can measure the improvement after discounting the change in the control group.

Tracking different tests against a control group helps teams optimize their product presentations for more conversions.