A controlled comparison of the effectiveness of variants of a website, email, or other commercial product. Usually, a novel or treatment condition is compared against a baseline or control condition to determine the effect of the treatment condition.