A/B testing an email

Themes of this article


E-goi allows you to run A/B tests for your email campaigns, helping you determine which version performs best based on open and click rates.



What is A/B testing?

A/B tests are perfect to figure out how your email contents truly impacts engagement, as the choice of what works or not is made by your own contacts (without them realising it).

Which version of your email will have the best response? The one with a punchier subject line? Should you go for a "View item" button or a "Buy it now"? You can automatically leave that call to your own audience!

Just run an A/B split-test. E-goi will send both versions of your email to a slice of your contact list (usually 5%). Half of those people will get version A while the other half receives version B. E-goi then gauges how each version performs for a couple hours and sends out the winning email (the one with more opens or clicks) to the remaining 95% of your list.



Before you start



To perform an effective A/B test, it is essential that you have:

  • Two or more versions of the email you want to test, with variations in specific elements such as button text, images, or subject lines.
  • Ideally a segmented contact list in E-goi.
  • Clarity about the success criteria (higher open rate, clicks, etc.).



What you can test

E-goi lets you split-test any two or more email campaigns with any differences between them, including:

  • Subject line
  • Sender
  • Any content (layout, images, CTAs, etc.)
Even though you can test any number of variables in a test, try narrowing it down to one variable per test. Testing multiple variables will make it harder to figure out which brought you the best results.



How the the winner campaign is chosen

The winning email campaign can be decided by:

  • Unique opens
  • Total opens
  • Unique clickthroughs
  • Total clickthroughs



Step by step


Create your email campaign versions

1) Create an email campaign as usual. This will be the "A" version.

2) Create a separate version of this email campaign containing the differences you want to test. This will be the "B" version. Those differences can be as small (subject line only) or as big (a wholly different layout and pics) as you like.

A/B testing is usually done with two versions of an email campaign, but you can do it with up to five versions.



Prepare and run the A/B test

1) Hit the Engage menu and click A/B tests.

3) A new page will come up. Click Create A/B test.

4) Choose how the winner will be picked (eg. the one with the most clickthroughs wins!). Provide the test with a title and select the contact list (and its segments, if necessary) you want to run your test on.

5) Tick the email campaigns you want to test (these are the versions you've previously created).

6) Under Options, select how the test will be split across your contacts and for how long (1 hour is usually enough). E-goi defaults to run the test on 5% of your list or segment, but you can make it larger. These will be the contacts picking your winning campaign!

7) Now simply choose when you want to run the test (right there and then or schedule it) and what happens when it finishes: E-goi can auto-send the winning campaign to the remainder of your contacts or simply wait for you to check the test results and send the winner manually.

8) All set? Just click Enable A/B test at the bottom of the page.



Wait for the A/B test result

1) E-goi will then go (or wait until the time you scheduled) and A/B-test for the number of hours you've chosen.

2) When done, it'll be marked as Finished, so you can then hit Report to find out which email edged out for the winning spot.

3) If you've selected to send the winner email manually, you can now do it. Otherwise it has already been auto-sent by E-goi.



Frequently asked questions


Which variables should I split-test?

Some examples:

  • Subject line:

    • Test out different subject lines (short versus long, emoji-heavy versus emoji-free).
    • Use different approaches (urgent, engaging, enticing).
  • Email content:

    • Compare versions with different messages or offers.
    • Test stories versus hard data.
  • Design and layout:

    • Try different layouts (simple versus complex).
    • Test CTA buttons position and size.
  • Images:

    • Compare image-heavy emails versus no images.
    • Test different types of images (item photos, infographics, etc.).
  • Call-to-action:

    • Try different texts (including all caps) for CTA buttons.
    • Test the CTA placement (email layout header, body, footer).
  • Personalization:

    • Compare personalised emails versus generic emails.
    • Test different personalisation levels (contact name only, contact name and purchase history, etc.).
  • Offers and promotions:

    • Test different offer types (X per cent off versus fixed amounts).
    • Compare time-limited sales versus continuous offers.
  • Font size and style:

    • Experiment with different font sizes and styles.
    • Test using different font types (serif versus sans-serif).
  • Email signature:

    • Test different signatures (formal versus informal).
    • Compare signatures with sender photos versus no photos.
  • Content format:

    • Compare text-based emails versus image-based emails.
    • Test using videos with "play" overlays versus plain images for videos.
  • Title and snippet:

    • Test different combinations of subject lines and snippets.
    • Compare different snippet strategies (short versus long, straightforward versus intriguing).


Can I run A/B tests in my E-goi automations as well?

Yes. Automations can be split-tested at any stage. Check out how it works!