A/B testing
Short definition
An A/B test shows two versions of a page or advert to visitors at the same time to see which performs better.
Some visitors see version A, others see version B; a single variable such as a headline, image or button text is changed and compared against a metric like conversion rate. Testing only one variable at a time makes it clear which change caused the result. Change more than one thing at once and it becomes impossible to say afterwards which change actually mattered.
For a result to be trusted, it needs enough visitors and enough time; a low-traffic page tested over a few days can mistake a random fluctuation for a real difference. Decisions should not be made before statistical significance is checked.
An A/B test replaces guesswork with evidence; instead of asking «does this look better», it asks «does this convert more».
Why it matters
Without an A/B test, whether a page change works stays an assumption. It turns the real effect of a small change into a number, which stops decisions being made on the wrong basis. Skip this discipline and a page that merely looks different, without actually converting better, can stay live indefinitely.
Illustrative example
When a flooring manufacturer A/B tested changing its quote form button from «Get a Price» to «Request a Free Survey», the second version drew noticeably more form submissions.
