What is split testing?
Split testing, also called A/B testing, is a systematic way to optimise content by comparing multiple versions to see which performs best.
Just like any other scientific test, split testing uses a control and a variant. The control is your original, unaltered page. The variant is a changed version of the original page. Traffic coming to the site is distributed randomly between the two. The variants are then compared to the original to assess their performance through their conversion rate.
If the variant site has a significantly increased conversion rate to the original then the site can be changed to the variant with the confidence that it will perform better.
Why use split testing?
In digital marketing, conversion is key. Conversion is getting a visitor to complete an action that you want - it could be buying a product, clicking a link, or signing up to email list.
Split testing offers a simple and effective way to change your content to have a higher conversion rate. The data-driven approach takes the guesswork out, giving you confidence that the changes are effective.
Using split testing
The first step to split testing is deciding what to test. Almost anything on a site can be split tested.
Headlines are an ideal candidate. As one of the first things a visitor sees, it is important to get the headline right - a successful headline can be the difference between a visitor reading on or not. Split testing can help determine the right headline to use.
Call-to-actions are another obvious way that split testing can help. A call-to-action is what causes a user to perform an action on your content. Every aspect of these can be split tested. Should it be a button, or a hyperlink? Should it be at the start or end of the page? How prominent should it be? Split testing can give you the answers to these questions.
Overall site design is also important. Increasingly, the influence of colour and font on visitor behaviour is being recognised as worthy of close attention. More than just providing a background for the content, these design elements should be part of the overall content strategy.
Next, a hypothesis of the effect of the change should be produced. For example, if the colour of an aspect of content is being tested, the hypothesis could be that changing the colour from blue to green will significantly boost conversion rate.
At this point, the test can be run. Two versions of the site would be produced. The original would keep the colour of the element as blue. In the variant, the colour would be changed to green. These two versions of the site would be run simultaneously, with traffic distributed between the two, until enough data has been collected to give confidence in the results. The length of the test will depend on how the amount of traffic that the site receives.
When enough data has been collected, the conversion rate of the two sites can be compared. If the variant has a significantly higher conversion rate than the original, then the colour of the aspect can be changed to green with confidence.
Could split testing benefit you?
Split testing is used by everyone from companies with massive marketing departments to small companies taking their first steps into digital marketing. The simplicity and the effectiveness of it means that split testing is a method that can be beneficial at every stage of development.
Split testing can play an important part in shaping your content strategy. If you'd like more information about split testing, or about how you can use it to help your content strategy, then get in touch.