A/B Testing Triggered Emails

An A/B test allows you to test two different versions of an email in order to decide which one performs better. In the test one email rule is assigned to group A and the other to group B and applicable visitors will only receive one of them, chosen by random.

The test is not limited to the email content, you can also test different sending times, subject lines or segments.

Quickly creating a copy of an existing rule

If the changes that you want to test are minor (for example, a different subject line) then you can quickly create a copy of an existing rule and apply your changes. To create a copy click on the Save and Duplicate link at the bottom of the rule setup:

saveandduplicate.png

after clicking the link you will be editing the new rule. Make your changes and then save the rule. It's recommended to:

  1. Change the rule name (for example 1st cart abandonment email - B)

  2. Change the tracking parameters so that you can distinguish traffic from both emails in your analytics software (example utm_campaign=CartReminder1B)

Assigning the email rules to the A/B groups

Edit the rule and click on Exposure Settings - > Triggered Emails A/B testing:

a-b-test.png

Assign one of the rules to group A and the other to group B and then activate both rules.

Comparing the results

Results can be compared either in your analytics software (if you made the recommend change above so each email uses slightly different tracking parameters) or from viewing the statistics next to each rule (accessible by clicking on the icon marked below:

statlink.png