You can conduct AB testing on three elements that influence email opens:
- Sender “from” names
- Subject lines
- Pre-header content
How does AB testing work?
10% of the total recipients are selected at random to be sent Version A of your message and another 10% are sent Version B.
After the last message in the test group has been sent, results are collected for two hours to determine the winning version.
The winning version is selected based on open rates. The version with the higher open rate is automatically sent to the remaining 80% of recipients. In the event of a tie, version A is chosen as the winner.
Undeliverable / bounced emails are excluded from calculating test results.
How to conduct AB tests
Click the “Create AB test” button in the sidebar of the email editor to get started.
Select the type of test you’d like to run and enter / select the content for versions A and B.
Send the email campaign as you would send standard messages without AB testing.
On the reports page you can view reports for version A and B independently or view combined results.
It’s possible to add emails with AB testing as part of nurture messages. In this case a winner is not determined: Version A or B are sent randomly and ongoing results are tracked.
Notes and restrictions
- A minimum list size of 500 contacts is recommended for effective AB tests.
- The total number of test recipients may be slightly higher or lower than 20%.
- AB testing is not available for legacy / custom HTML messages.
- It’s not possible to re-send completed AB tests to single contacts.
- Both version A and version B will be sent when you send test messages.