Split testing lets you create different versions of an email and compare how each version performs.
You can split test Emails and email steps within Sequences. For example, you can test different subject lines, content, images, or calls to action to learn which version gets more opens or clicks.
For a clearer comparison, consider changing one element at a time. If several parts of the email are different, it can be difficult to determine which change affected the results.
Add versions to an Email
To set up a split test for an Email:
Go to Marketing > Emails & SMS.
Click + Email/SMS.
Select Email, enter a name, and click Create.
In the Design and configure section, click + New Version.
Kartra creates another version of the Email, such as Version B. You can add additional versions by selecting + New Version again.
Choose how recipients are divided between versions
You can control what percentage of recipients receives each version.
Click the three dots next to a version.
Enter the percentage of recipients you want to receive that version.
The remaining percentage is assigned to Version A.
Each eligible recipient receives only one version of the email.
Edit your email versions
Select Version A, Version B, or another version to switch between them and edit their content.
For each version:
Select the version you want to edit.
Click Edit message.
Change the element you want to test.
You can test elements such as:
Subject lines
Email copy
Images
Links or buttons
Calls to action
For example, you could keep the email identical except for the call-to-action button:
Version A: Sign up now
Version B: Get your download
Keeping the rest of the email consistent makes it easier to understand whether that particular change affected engagement.
Test your versions before sending
Before sending the Email to your audience, check each version to make sure its content, formatting, and links work as expected.
You can send a test email from the email editor for each version.
Test emails are sent from [email protected] rather than your configured email gateway. Placeholder variables, such as a contact's first name, display as TEST instead of using real contact information.
See Test emails before sending for more information about testing your emails before they reach your audience.
Add versions to a Sequence email
You can also split test individual email steps within a Sequence.
For the email step you want to configure, add another version from its email content settings. You can then edit each version separately and choose the percentage of eligible contacts who should receive it.
The percentage allocation works the same way as it does for an Email: the percentage assigned to the additional versions determines their share, and the remaining percentage is assigned to Version A.
You can then configure and edit each version separately. Each eligible contact reaching the email step receives only one version.
The versions apply to that particular email step. Other email steps in the Sequence can have their own versions and percentage distributions.
Review your split test analytics
After recipients receive your Email or Sequence email, you can compare how the different versions performed.
Email split test analytics
After an Email is sent, you can compare the performance of its different versions.
From the Emails & SMS dashboard, you can view information including:
Number of recipients.
Distribution of recipients between versions.
Open rate for each version.
Click rate for each version.
For more detailed results, go to Marketing > Emails & SMS > Analytics.
Sequence email split test analytics
To review the results of split-tested Sequence emails:
Select the relevant Sequence to view its individual steps.
If an email step has multiple versions, each version appears in the list along with its assigned percentage and performance data.
You can review metrics including:
Recipients
Spam complaints
Bounces
Unsubscriptions triggered from the email
Opens
Clicks
Goals
Revenue
Comparing the versions can help you understand how the changes you tested affected engagement and other results.
