A control in a growth test or experiment is your sample that remains consistent (i.e. doesn’t change) over time throughout your experiment. For example – let’s say you have a B2B SaaS platform and you’re currently seeing a 10% conversion rate of visit to free trial, and you’ve not released any major updates in 14 days. That can be your control. If you then run a growth test and change the free trial verbiage and form fields, that becomes your variant. You’ll measure success on how your variant or variants perform against your control. If variant “A” wins, then you can define that as your new control and run future variations against the new control.
Looking for ideas on how to build and measure a growth test? Check out our blog post here, or download Craig’s PDF below.