What is an A/B test?
An A/B test shows two versions of something, like a button, to two similar groups of visitors, and compares which one gets more sign-ups or sales.
Last updated
In real life
Like a baker selling two cookie recipes side by side for a week and counting which tray empties first.
Try it yourself
The demo sends 1,000 visitors to each of two versions of a landing page: A has a "Submit" button, B has the one you pick. "Start my free plant plan" wins clearly. "Click here" is too close to call, so the demo says to keep testing.
The landing page (a made-up app)
Leafy
Never forget to water a plant again
Leafy reminds you when each plant needs water. Free for up to 5 plants.
Start my free plant planA/B test
This lessonVersion A keeps “Submit”. Version B uses “Start my free plant plan”.
An A/B test shows two versions to similar visitors at the same time and counts the results. Try "Click here": a small difference isn't proof.
A/B test vs asking people
Asking people what they'd click tells you what they think they'd do. An A/B test shows what they actually do, which is often different.
Common questions
How long should an A/B test run?
Until each version has enough visitors for the difference to be real, not luck. Small differences need far more visitors. Testing tools tell you when there's enough.
Can I test more than one change at once?
Change one thing per test, like the button text. If you change the headline and the button together, you won't know which one made the difference.
Where you will use this
Quick check: A/B test
Question 1 of 3