This document provides an overview of A/B testing. It explains that A/B testing involves developing two versions of a page and randomly showing them to users to track their behavior and evaluate which performs better. It discusses why companies do A/B testing, what can be tested, limitations of the G-test typically used to analyze results, important metrics to measure, and common mistakes to avoid such as comparing results across different time periods or traffic mixes. The goal is to help the reader understand how to properly design and analyze an A/B test to identify impactful changes.