SELF-GUIDED BA PRACTICE Packages available · Access after confirmed paymentSee packages ↗

Business analysis glossary

A/B Test

What it is

An A/B test compares two versions of something, such as a page, message or flow, by showing version A to one random group of users and version B to another, then measuring which performs better on a chosen metric. Statistical checks show whether the difference is likely to be real rather than chance.

Why it matters

A/B tests replace opinions with evidence about what users actually do. Analysts help define the hypothesis, the success metric, the sample size and the test duration, and they read the results carefully, because ending a test early or checking too many metrics often produces false winners.

Example

The analyst tests whether showing the delivery date on the product page increases orders. Half the visitors see the date and half do not for two weeks. Orders rise from 3.1 to 3.4 percent with a confidence level the team agreed in advance, so the change is rolled out to everyone.

CHECK BEFORE CONTINUING

Keep your work safe