Skip to content
  • There are no suggestions because the search field is empty.

A/B Testing

Not sure which version of your text will perform the best? 

You can now test the performance of two versions of a broadcast message BEFORE sending to your full audience. 

Here's How 👀

  1. Navigate to the New Broadcast tab. Select the audience you'd like to sample from first by selecting the Edit Audience button and choosing your audience. Once that has been selected, you'll click the A/B Test icon in the bottom right toolbar.to begin your test set-up. 


    Screenshot 2026-05-11 at 4.40.15 PM

  2. Once you click the A/B test icon a configuration modal will appear titled Configure A/B Test. 

    Screenshot 2026-05-11 at 5.31.31 PM
  3. The modal displays:

    • your total broadcast audience size
    • a visualization of the audience split 
    • two input rows for setting the percentage of your audience that will receive each version
    • a third "remainder" segment that represents the subscribers who the winning version can be sent to after you evaluate results
  4. From here you will set your desired split.

    For example
    • 25% Version A 
    • 25% Version B 
    • 50% Remainder

    ** If you want to send both versions to the entire audience with no remainder, set a 50/50 split. The Remainder will show as 0%, and no winner delivery step will be needed.

     

  5. Click Continue to confirm.
  6. The composer will then split into two stacked composers: Version A and Version B.

    Screenshot 2026-05-11 at 5.43.45 PM

  7. From here, you can compose different message content in each version section. This means you can even include different text, links, or images in each version! 

    ab-1
  8. Once you've finished composing your messages, you can preview them by clicking on the "eyeball" or preview button in the bottom right toolbar. This preview shows an SMS mockup of your messages and you can toggle between the options to verify the format. 

    Screenshot 2026-05-18 at 3.48.24 PM


  9. ***Optional Step: You can use the scheduling flow to set a date and time for each version. 
  10. Or you can save/send the broadcasts immediately according to your needs. After sending, you will see a confirmation banner on the Broadcasts page.

    Screenshot 2026-05-18 at 3.52.55 PM
  11. Review Results. Once both versions have been published, the Broadcasts listing page will show your A/B test broadcast with a special A/B Test badge and a Version A / Version B segmented toggle. Click each tab to view the performance metrics for that version, including response rate and reply count.

    Screenshot 2026-05-18 at 3.54.28 PM
  12. When you are ready, click Select as Winner on the version you want to send to the remaining audience. A modal will appear with two options:

    Screenshot 2026-05-18 at 3.56.10 PM
  • Send Now: Deliver the winning version to the remainder group immediately.

  • Schedule: Choose a date and time for the winning version to be sent to the remainder group.

After confirming, the listing page will update with a banner alert confirming your selection.

** If you scheduled the winner, you can modify or cancel the schedule before it sends.

And voila, your testing is complete! 💯

Screenshot 2026-05-18 at 3.59.57 PM

Once a winner is selected, a third tab, Winning Version, appears in the version toggle on the broadcast row, where you can see the metrics for this broadcast as well. 

Any questions?

Just let your CS Manager know and they'd be happy to guide you through the testing process. 😎