I’m not going to give a full explanation of A/B testing, or mulitvariate testing, here. First, I’m no expert and second, Ayat Shukairy gives as good an explanation as any in her post on multivariate testing for the Invesp blog.

Briefly, A/B testing involves the selection a certain elements of your website or landing pages, such as the images you use, your call-to-action buttons. page layout, etc., and testing them against other options for those elements to see which option performs best.

A simple example would be an online form with a grey CTA button at the bottom of the form labelled “Submit”. If you suspected that a different button might increase form submissions, you would create another version of the form page with an alternate button and present that page to half your site visitors, while the other half get the original page. After a pre-determined testing timeframe, you would see which page generated the most submissions.

A/B testing is responsible for many of the best practices of conversion optimization, website usability and online customer experience.

Perhaps most importantly, A/B testing is cited in seemingly endless case studies and tests as significantly increasing ecommerce sales.

So just just how much of an increase can be achieved with A/B testing?

Results will always vary, but this should give you an idea: a variety of A/B test results found on Google showed an average increase in ecommerce sales of 27.25%.

Again, results vary and no one should expect certain returns from A/B testing. This average most likely represents some of the better test results, probably from sites that were not previously optimized.

But these sorts of results are repeated in test after test after test. And even if you slash this average by two-thirds, or get nasty and chop it by 90%, you still get a 3% increase in sales. I’ll take that any time.

The undeniable fact is this: I found no online evidence that properly executed A/B testing does not produce results, and overwhelming evidence that it produces significant increases in sales.

Which means there is one big question about A/B testing:

Why don`t more businesses use it?