TersbyWhen you need it
When you aren't sure what works better: a different headline, price, first photo or order of blocks. A test shows both versions to different visitors and counts the leads for each.
A/B tests are available from the Flow plan.
Create a version
- Open the site in the editor and go to the “History and tests” tab.
- Add a version: it starts as a copy of the main page.
- Change what you want to test in it. Changing one thing is best: then it's clear what made the difference.
- Set the share of visitors who see the version.
A visitor always sees the same version, however many times they open the site. There can be several versions.
How to read the result
For every version you see visitors, leads and conversion. Below them Tersby writes a conclusion:
- Too early to tell: there's little data, and the numbers can't be trusted yet.
- No difference: the versions perform the same.
- Looks better or looks worse: there's a lead, but it may be chance.
- Better or worse: the difference is real, with the stated confidence.
While the conclusion is “too early to tell”, don't stop the test: the first leads almost always fall unevenly.
Finish the test
The button that makes a version the main one moves the winner's content to the main page and stops the test. If the main version won, simply delete the other.
What else you see
In the stats, leads and visits can be broken down by version, and a lead says which version the buyer saw.