Showing two versions of a page, ad, email or offer to randomly split audiences at the same time, then letting the results decide which one wins.
Any time there is a debate about copy, design, offer or price and enough traffic to reach a reliable answer.
Subject line tests in email; headline and CTA tests on landing pages; ad creative experiments; Claude Hopkins keying coupons a century ago to see which ad pulled.
The core of evidence-based marketing in every channel. Flagged when tests are stopped early, run without enough traffic, or ignored when the result disappoints.
