ab-split-testingWebsites are tough work and the little; incremental things are the fine line between a good website and a great website. Of course, it’s often not easy to know what to do to get that little bit more from your site; however A/B split test maybe the answer.

If you’re in the game of conversions, whether in the ecommerce, lead generation or services field, A/B split testing is a must.

What is it?

A/B testing is quite simple in concept. You simply have two versions of a page and a metric that outlines accomplishment. You subject both versions to real world use and whichever version does better in a prescribed metric is the winner.

A/B testing is perfect for landing pages, online forms and other conversion based pages and allows you to determine which is the most successful in real life. In turn, this should help you increase your bottom line.

How to Do It?

Tested pages don’t have to be too disparate and you’d be surprised at how little changes can create significant successes. Like any sale or conversion, on screen or off it, the field needs to resonate emotionally with the potential client and they need to get excited about the product. This can take the smallest of alterations to showcase the worthiness of the conversion.

Define Changes

You need to define the changes before the test. Firstly decide on what alterations you wish to make, these may be a change of colour, wording or other visible aspects of the page. Change one of these things and you have an ‘A’ version, and then change something else that you think may affect the page positively and you have the ‘B’ version of the page. These can then be tested against each other directly.


Before Testing

Take a note of the performance of the original page. Factors such as the conversion rate and the bounce rate are important here. Taking note of these allows you to compare them against the two versions that are now being tested.

The Test

One of the most common mistakes people make when testing new versions of a page is to not allow enough test time or page views. If you’re thinking about changing a landing page or form, you want to have a large amount of data to back up the change.

It’s not uncommon to test for several months and thousands of views to collect the information necessary. After all the more of these the better and more accurate the information and the less likely there is to be random fluxes in numbers.

Testing should be monitored. It’s common for neither A or B page to work well and cause falls in sales. In such a scenario it’s best to make changes and perhaps try another A/B test.

In addition, A/B testing should be constant and done regularly. People change the way they consumer information and make decisions and regular A/B testing keeps you at the cusp of this evolution. The way we engaged with things a few years ago is very different to the way we do now and A/B testing helps such problems.

So, as you can see A/B testing is a very important part of conversions. It’s all good having great SERP rankings, traffic volume and social media presence. But, if you can’t convert this to leads, interest or cash, it’s a waste.

Written by Cormac Reynolds