Overview

Testing is one of the fastest ways to improve performance. That’s why every campaign and automation on our platform supports A/B testing (or multi-variant “experiments”)—so you can test, learn, and optimize with confidence.

Screenshot 2025-04-13 at 11.09.13 AM.png
A simple A/B test to measure the effect of extended recovery messaging

What is an experiment? 🧪
An experiment allows you test two or more versions of a message (A/B/C/D, etc.) to see which one performs best. You can test things like:

How does it work?
We split traffic evenly across each variant and report the results on the A/B Tests analytics page. You’ll be able to compare key metrics like conversion rate and order value—so you can quickly see what’s working and by how much.

Screenshot 2025-04-13 at 11.11.41 AM.png
An A/B test with a holdout to measure if a recovery is more effective than not

Why it matters:

Whether you're testing a CTA line or rethinking your welcome flow, experiments are one of the simplest and smartest ways to drive better results.