An experiment in which two versions of a piece of content or product are shown to randomly divided audiences to determine which performs better according to a defined metric. Properly run A/B tests require sufficient sample size and statistical significance before conclusions are drawn. (See Chapter