How to A/B test widgets
A/B testing, also known as split testing, is a powerful tool that allows you to compare widgets against each other and find out what words, images, and promos attract more leads. In other words, A/B testing takes the guesswork out of your engagement strategy.
Interested? Here are the instructions to effectively run A/B tests on your widgets.
1. Create and set up the widgets
First, sign in to your Getsitecontrol account.
Next, start creating your widgets. You need to create at least two versions of the same widget to run an A/B test in your Getsitecontrol dashboard.
It’s important to change only one variable when creating widget versions for the test. Otherwise, it will be difficult for you to establish what really made one version more successful than another.
For example, your widgets may feature a different image, advertise a different promo, or have different wording — but not everything at the same time.
Take a look at the example widgets below to get an idea:
To save time and avoid having to create all of your widgets from scratch, make a copy of the original widget using the Clone button. Then apply the necessary changes to the copy.
❗Please, note that all widgets participating in the A/B test should have identical targeting settings. Otherwise, some widgets may get more views from your visitors than others, and the test results would be unreliable.
Once your widgets are ready, proceed with the steps below to set up an A/B test.
2. Add the widgets to the test
To include a widget in the A/B test, click the A/B test button at the bottom of its card in the All widgets section. Choose a test, e.g., Test 1.
Make sure you are adding all widgets to the same test.
That’s it: now each widget will be shown to an equal percentage of your site audience.
3. Monitor the results
The data on your A/B tests is stored in the corresponding section of the dashboard, which you can access from the menu on the left-hand side of the screen:
Click on the section to see how the tests are going:
You can stop the test at any time: just click the Stop button on the A/B tests page. Then you can choose either to deactivate the less successful versions of the widget or delete them:
❗There is no one rule for how long you should run the test for the results to be reliable, as too many factors play a role in that: the average number of monthly visitors, the current conversion rate, the widget’s targeting settings, etc.
Therefore, we recommend that you regularly monitor the reliability of the results you got using an A/B testing significance calculator such as this one. The percentage of reliability that you should aim for is >95%.
Now that you know all about the science of A/B testing, it’s time for you to put it into practice on your own widgets.
If you have any questions or issues with A/B testing, please contact our Tech Support team. We’ll be happy to assist you!