A/B testing

A/B testing is an instrument of marketing research. During an A/B testing you can find out how changes to the site influence the visitors’ desired actions (submitted leads, purchases, sign-ups).

For example, you have two versions of a section that differ from each other a little: in design, presentation of the commercial offer, photos. You publish both versions to learn which one will receive more response from potential customers and save the successful version.

Here is an example of two identical sections with different background images.

There is no time limit for A/B testing: it is individual as you need to collect statistically valid data to get the whole picture.

A/B testing is set up at the section level. Several tests applied to different sections can run on one page at the same time.

To create A/B tests, click the “…” button in the section settings and then switch on the A/B testing toggle.

*******

Select an appropriate option in the pop-up window.

Copy current section
A selected section (version A) is duplicated and its duplicate (version B) appears right below it. It comes in handy if you want to make slight changes of the section.

Don’t forget to set up the version B and save the changes. Once the changes are saved, an A/B test starts automatically, i.e. the incoming traffic gets divided into groups A and B.

Create a new section
You choose any section from the catalog of  templates or create one from scratch. This is the best choice if your versions A and B should significantly differ from each other.

Hidden section option
Tests with a hidden section allow you to find out if any particular section affects the conversion rates of your site. 

When you enable the A/B testing, an icon of the versions A and B appears at the top left corner of the section. The active version is colored green.

There's a statistics icon next to the versions A and B.

Click the A/B testing statistics icon to open a pop-up that shows:

• the number of users in both groups
• the number of leads for each version
• the conversion rate of each version
• the percentage of conversion improvement / deterioration with the version B

You can pause or complete the testing.

If you pause the testing, the site's visitors will stop seeing the version B.

If you resume the A/B test, new visitors will continue to be divided into groups and the statistics will be collected further.

To complete an A/B test, click the Complete button. The last step is to select the better performing version, the other one will be deleted. Your choice should be a section with a higher conversion percentage.