A/B Testing Foundations: The Math, the History, and Message Match (Three References Rescued)

A rescue-and-synthesis reference on A/B testing foundations. Three of the most-cited references on A/B testing — Pardot's 2012 primer (Jenna Hanington), Amazon's developer documentation walking through the math behind a confidence interval, and Monash Business School's marketing dictionary entry on "message match" — have all gone dead at their original URLs. This article reconstructs the substance from Wayback Machine snapshots and combines it with operator notes from 13 years of running CRO programmes. Topics covered: the history of A/B testing from Google's first SERP experiment to default-on at Amazon and eBay; the actual statistical maths behind a confidence interval (rather than the marketing-tool dashboard summaries); the Monash academic definition of message match between paid-ad headline and landing page headline; and a practical operator-led discipline for sample-size calculation, stopping rules, and interpretation of failure as information. Companion to the live article at https://www.gogochimp.com/blog/ab-testing-foundations-rescued and to the Wikipedia archive-rescue edit on the Landing page article (revision 1352184645, 2026-05-02).