The Ultimate Guide to A/B Split Testing

We are sure that you have heard of A/B Split Testing before but have you really understood what is it about? Did you ever tried to create one of these A/B tests and see how they work? If your answer is no to at least one of our questions, this ultimate guide to A/B Split Testing is just the right thing for you!

split testingWhat is A/B Split Testing?

In web development and marketing, A/B testing or split testing is an experimental approach to web design (especially user experience design), which aims to identify changes to web pages that increase or maximize an outcome.

Split testing represents one of the easiest ways to increase conversion rate and it also gives you important insights about your audience. A/B testing can be really efficient in showing you what it is to be changed and how to change it. But it is important to initiate these simultaneous experiments having in mind just one change since split testing is limited to single elements of a page. Otherwise, you can use Multivariate Testing in order to determine a combination of changes that needs to be implemented.

How Does A/B Split Testing Work?

At its core, A/B Split testing is exactly what it sounds like: you have two versions of an element – A and B, and some kind of metric that defines success. A/B testing requires that you subject both variants A and B to simultaneous experiments in order to determine which one is able to score higher on your success metric. In the end, that version which was more successful will be used for the real-world use.

Here is a good example of A/B Split Testing used to improve Email Marketing Campaigns, from Mailchimp. In this example, they are testing subject lines for Email Marketing Campaigns.

split testing

What to Test with A/B Split Testing?

Usually, the choice of what to test will clearly depend on your goals. If your goal is to increase your conversion rate, you need to look at the elements involved with conversion and decide with which one to start experimenting.

Even though every split testing is unique, here are some examples of elements that can be tested during an A/B testing:

  • Headlines – size, color
  • Content – length, wording, font, placement on page
  • Layout and style of website
  • Social Media Buttons – placement, size
  • Product pricing and promotional offers
  • Call-to-action Buttons (“Buy Now”, “Sign-Up”) – size, color, placement on webpage
  • Images on landing and product pages

Tools for Creating Your First A/B Split Testing

Once you have decided on the element you want to submit for split testing, you need to choose the tools for testing it. There are several options: you can either integrate this experiment with your Google Analytics account, or you can create an experiment with another specialized service.

Google Analytics offers you the possibility to experiment in order to improve your conversion goal or any other element on your website. There is a simple process for creating and starting your split testing and Google Analytics offers you a complete guide for doing things right. You can read all about it on your Google Analytics account, on the Experiments tab.

split testing

Another way to start creating and implementing a split testing is to choose a specialized service to help you do it. You can choose, for example, Visual Website Optimizer. They offer you a complete guide to A/B Split testing, that teaches you basic concepts of Split testing and helps you implement your first trial test for free.

Whatever tool you choose to start creating your split testing, your should remember to never surprise your regular visitors while testing a core part of your website, but include only new visitors in the tests. Also, keep in mind that split testing is done for better conversion rate or other important goals and not for aesthetics, so try not to involve your gut feelings and never reject a result just because of your arbitrary judgment. Other than that, we wish you happy A/B split testing!

Image from Mailchimp