Why A/B Split Testing Can Help Improve Your SEO Efforts

A/B split testing is a common SEO practice to check which version of a given page is performing better. It can be used to check, for example, if a variation of the copy writing may improve the conversion rates, or if moving a button around can affect user experience. It’s a simple technique that evaluates which version outperform the other (either A or B) by changing  one small detail at a time, although some marketing experts also use more radical test approaches.

A core for many marketing strategists, A/B split tests provide some unexpected and amazing results when employed in your strategy. It paves the way to some popular theories on what is the optimal font type or size is, where a button should be placed or what amount of text your landing page should have.

However, many marketers  still tremble in fear whenever you talk about making an A/B split test. Because of some outdated myths, many people in this industry still think that having two versions of the same page may cause a website to suffer from penalties such as cloaking, duplicate content or reduced on-page traffic. Let’s see why all these fears are actually largely unmotivated.

  1. A/B split testing is not content cloaking

Cloaking was a popular practice back in the 1990s and early 2000s. Back then keyword stuffing provided amazing results to rank a website as high as possible, and old-school smart SEO experts used to build two pages. The first one was just full of keywords for spider and engine bots to crawl, while the second, one was displayed to human visitors. Search engines later evolved to detect and penalize sites who used this practice but… almost 20 years have passed, and everything changed in the meanwhile.

Keyword stuffing is not working anymore, and since content cloaking is not a viable tactic, newer search engine algorithm changes did not include it among the penalized practices. Modern websites are not static HTML ones as they used to be, so dynamic content swapping done with JavaScript is not seen as cloaking anymore. All the largest search engines, including Google, do a lot of A/B split tests on their own sites. So unless they’re penalizing themselves (and they don’t), there’s nothing to be scared of.

  1. A/B split test does not generate duplicate content

The greatest fear of the A/B split testing opponents, is that variation pages may cause a duplicate penalty since their content is extremely similar (if not identical). Obviously, if you’re A/B testing pages which include substantially different content to test out why a given pages is failing, this threat does not even exist.

On one hand, you should know that, in theory, duplicate content penalty only applies to content you copied from other websites. So if you simply use different variations of your own content, you are free to use them as you want.

However, there’s still a risk of an internal duplicate content lingering above your site and some people swear this largely hypothetical penalty actually exists (note: it may be nothing but just myth). Even if we want to believe this tall tale, an internal duplicate content can be easily avoided by applying a rel=“canonical” script to your pages, so search engines will know they just need to crawl one of the pages you’re testing instead of them all. You can also manually de-index the pages which may give you a duplicate penalty by either removing them from the .xml sitemap or using Webmaster Tools’ specific function.

  1. A/B split testing will reduce your on-page traffic

Since you have different variations of the same page, your traffic is going to be split on several ones instead of being focused on just one of them, and some people fear this effect may negatively affect that page’s ranking. Although it’s true that high traffic can give a bonus to a page’s ranking, this bonus is so quite minuscule, so you can safely skip it without even noticing any change.

However, if you still want to avoid it altogether, you can simply split test group of pages instead of single ones. Rather than comparing the performance of two versions of the single page, you test two variants of control groups inside your site. For example, if you have an ecommerce, you may use Variant A for all pages selling glasses, shoes and trousers, while Variant B to test hats, jackets and t-shirts. By averaging the results of the two groups you will still know which one is performing better.

As you can see, there’s no reason to think that a good old A/B testing is going to negatively impact your SEO or that you can’t handle the task at hand.. If you still have some doubts, though, try monitoring your website performance with an  SEO auditing tool for a little while. If something bad still happens (but it won’t), you can simply revert your changes without suffering any long-term penalty.