PHPMatters Help You Better Hosting Your PHP-based Sites
How to Start WordPress A/B Split Testing

How to Start WordPress A/B Split Testing

No matter how large and what kind of websites you are running, you must want it to convert better than ever before. To achieve this goal, you have to make a lot of changes for optimizing your site, which include the aspects of overall design, web content, headlines, images and some call to actions, etc. However, the question is that can these changes really be effective to make your visitors click?

To figure out the result, it is better to carry out the A/B split testing than guessing or waiting for the changes that take effects. As many webmasters have no idea of how to carry out this practice, in the following, we’d like to introduce the steps as how to start WordPress A/B split testing.

What is A/B Split Testing?

To put it simply, this practice allows you to test different versions of your site and to figure out which version results in more visitors, more clicks and more traffic. For instance, the version A is your original site and the version B is your site that contains some changes based on the assumptions of they may work better for the website running. With the split testing, you can know exactly which version gives the better result, thus decide whether to update your site with these changes or not.

Some Do’s or Don’ts about A/B Split Testing

Before learning the detailed steps, you firstly need to know some do’s and don’ts for carrying out the split testing effectively.

Do’s

  • You have to decide the exact time range for this testing. After all, stop the test too early may cause some losses on the real meaningful results. If you test for a long time, the poor performing variations can do no good to your site.
  • In order to prevent blunders, you need to showcase the same variations for the same visitors.
  • You can do multiple A/B tests for the most precise results. Generally, you can get three outcomes from the tests – no results, positives ones and negatives ones. If you carry out tons of split tests and get the all positive results, you can make some modifications on your sites without any worry.

Don’ts

  • You can never test your two versions at different times. They need to be monitored simultaneously.
  • Do not conclude the result too early, but keep a deep insight into your website changes with repeat testing.
  • Do not affect your loyal and regular readers. You’d better only include the new visitors into your testing.

Start WordPress A/B Split Testing

At present, there are a lot of tools to carry out the split tests, all of which are easy to use. In the following, we have listed some great options for you to choose.

  • Optimizely – This tool costs you starting at $17/mo. It is really non-technical and is a good option for newbies. The only coding practice is to paste the testing code into your website files. Besides, Optimizely allows unlimited experiments and gives a point-to-click editor for the implementation of website variations.
  • AB Press Optimizer – Charged starting at $39 for each year, this tool is a premium one that integrates with WordPress dashboard directly for the starting of split testing, no coding stuff required. However, it is not suitable for some large projects but is great from some small fishes.
  • Simple Page Tester – This tool charges you $59 for the premium version. Its setup is straightforward and easy from the WordPress editor. With it, you can not only start the split tests for the whole webpage, but also can use some shortcodes to test the single elements.

As the above-mentioned options are all not free of charge, here, we’d like to introduce a free way for the setup of WordPress A/B split testing with the help of Google Analytics. Note that we only use its Experiments function. If you want to learn more functions of this tool, you can check this page.

Step one: create new experiments

First of all, you should navigate to the Behavior tab and click the Experiments button. Here, you can find a Create Experiment button. Simply click it.

Create Experiments

Step two: choose the objective

The content experiments of Google Analytics require a metric to get the testing evaluated. Therefore, the next step is to choose the objective for your testing. For instance, if you want to test your registration page, you can enter “Registration Page Testing” as the experiment name, and click the “Create a new objective” link to enter the metric. As for the “percentage of traffic” and the “email notification” option, you can leave the settings as default.

Choose Objective

Step three: configure the experiments

Now, you need to configure your experiments by entering the page link for both “version A” and “version B”. Note that the version A is your original page and the version B is the page you have made some changes. Here, you have to make sure that the version B is published already, being able to be accessed for all the readers.

Configure Experiments

Step four: insert the testing code

Now, Google Analytics will generate a testing code that needs to be inserted into the header.php file of your current template. Then, you can back to Google Analytics and click the Next button. Now, this tool simply verifies automatically whether you have the code inserted into your website properly. If yes, you can be presented with the following screenshot. Click the Start Experiment button.

Start Experiment

Now, you need to wait around 24 hours before Google Analytics showing the data. Note that your readers will be directed randomly to your “version A” and “version B”, and will continue to see the same version if they go back again. After waiting for a few days, you can know which version is working better by checking the tracking results of the split testing.