A/B testing
An A/B test compares two variants of a page, ad or email by showing one part of the visitors variant A and the other variant B — revealing which works better.
Instead of debating taste, real users decide: does the new headline lead to more enquiries? Does a shorter form bring more conversions? Does the second ad creative get more clicks?
A test is only meaningful with enough visitors and if only one thing changes at a time. With few visitors, tests take long or remain inconclusive — then user interviews or experience are often the better route.
A/B tests are especially easy for ads, because advertising platforms automatically run several variants against each other.
What to look out for
- Formulate a clear hypothesis: what should improve, and why?
- Change only one element per test.
- Define in advance how success is measured.
- Run the test long enough to rule out chance.
- Document results — including tests that brought nothing.
Common mistakes
- Changing several things at once and not knowing what worked.
- Stopping the test after two days because one variant is ahead.
- Tests without enough visitors that never reach a clear result.
Frequently asked questions
Until enough visitors have seen both variants and the result is stable — at least over full weeks so differences between weekdays do not distort it.
On the website often only to a limited extent, because too few visitors arrive. For ads, however, they work well because platforms test variants automatically.