5 Tips for successful AB testing

Posted by interexcel on June 18, 2018

ab testing

Successful AB Testing

At IEDM Digital Marketing, we run successful AB testing campaigns. We believe it is a powerful way of measuring digital success.

If you type “AB testing” testing into Google nowadays, you are likely to get a barrage of articles telling you that this method is now deader than a doorknob. This is not necessarily the whole truth. The fate of AB testing is a bit more nuanced, as we will discuss in an upcoming article. For now, let’s focus on the basics.

AB testing (also called split or bucket testing) lets you test two options to find which one works the best. Here is a simple example: let’s say you want to test whether a red or a green Call now! button leads to more clicks from your emailer. To do an AB test, you will create two (or more) versions of the emailer – one with a red Call now! button and one with a green Call now! button – and send one to half of your audience and the other to the other half. Based on how many people clicked on the red button vs. the green button, you will now be able to see which one is more effective.

Get the most out of your AB test

  1. Test one thing at a time.
    Putting more than one variable into a test will muddy your results. Even the smallest changes to layout, design, or wording can have a huge impact on your communication piece’s effectiveness.
  2. Test both options at the same time
    This may sound like a no-brainer, but it is important to test both variations during the same time frame. People are easily influenced by outside forces, and not testing simultaneously can lead to flawed results.
  3. Have a clear goal in mind.
    Think about what you want to find out before you send out the test, and formulate it as a statement. In the example mentioned above the test’s statement could be: “A red Call now! will lead to more clicks”. You will then do the test to prove or disprove the statement.
  4. Give it time
    AB testing likely won’t give you a clear answer overnight. To get a good result, you need to wait until you have collected enough data to prove the point, or in other words statistical significance. Figuring out what you consider statistical significant before starting your test.
  5. Take action
    Once you have the results from your AB test, you need to implement your findings. If there was no clear winner between the two variations, you can either assume that whatever you tested has no real impact and stick to the original version, or do another test to make sure of your results.

Let Interexel handle the AB testing for you

Our award-winning team has the knowledge and know-how to help you optimise your online presence. Contact us today for your free digital marketing consultation.

Sharing is caring!

Let's Talk