A methodology for comparing two versions of a webpage, email, or other marketing asset to determine which performs better. Users are randomly assigned to see either version A or B, and conversion rates are measured to identify the winning variation.