What is A/B testing?
A/B testing is an experiment that shows two versions of a page or element to randomly split groups of visitors at the same time and compares how each performs against one goal, such as sign-ups. It shows whether a change caused an improvement, rather than relying on before and after guesses.
How a test runs
Pick one goal metric, write a hypothesis, build the variant and split the traffic, often half and half. Let the test run for full weeks until it reaches the planned sample size, then read the result once rather than stopping as soon as it looks good.
How much traffic it needs
Small differences need a lot of data. Pages with only a few conversions a week may be better served by user research and bolder changes than by tests that never reach significance.
