โ04-03-2023 01:59 PM
What's the best way to A/B test within a canvas? Since there isn't the ability to A/B test different variants like with a campaign, we set up an 'experiment' but that only shows data but conversion rate and not opens/clicks/unsub/etc. With the experiment, the data is messy in the report builder and when you try to delete a step after the test it says you'll lose all the users in that step.
โ04-16-2023 09:31 PM - edited โ04-16-2023 09:32 PM
Hi @megantu
I often run two variants at the top of the canvas and a control. This leaves me with the option to compare variant data through each step as well as the ability to compare everything against a control group.
It's not perfect for detecting small changes like an campaign A/B test and the reporting is 'okay' but it is still helpful to see how each variant performs against one another.
๐
โ04-26-2023 08:11 AM
Pretty much what David above says, also consider the goal of the test. And importantly how long will you have to run it for in order to get significance?