A/B Testing
Comparing two versions of a message or page against each other on a single variable to learn which performs better, decided by data not opinion.
A/B testing pits two variants — version A and version B — against each other, changing one element at a time, such as a subject line or a call to action, and routing traffic so the winner is decided by results rather than instinct. It is how marketers replace guesses about what works with evidence from their own audience.
The discipline is isolating one variable and giving each variant enough volume to reach a meaningful result before declaring a winner. In outbound, A/B testing subject lines drives open rate and testing the opening line or offer drives reply rate; reading the right downstream metric, not just the easy one, is what keeps a test honest.
Related terms