
Statistical Methods in Online A/B Testing
Take your UX & CRO tests to the next level!
The book
Welcome to the most indepth book on statistical methods in online controlled experiments, a.k.a. A/B testing! It is the culmination of many years of work by the author Georgi Georgiev and aims to bring much needed clarity and context to the application of statistical tools for business decisionmaking and risk management through experimentation.
"This is the book I wish I had when I started learning A/B testing stats!"
 Georgi Georgiev, the Author
What Readers Are Saying
Some of the reviews given by readers of "Statistical Methods in Online A/B Testing"
Peep Laja
Founder and Principal, CXL Institute
"If you're running online experiments, you absolutely need this book. It will help you avoid costly testing mistakes stemming from misuse and misunderstanding of statistics."
* you absolutely need this book *
Alex Birkett
Senior Growth Marketing Manager, HubSpot
"This is simultaneously the most comprehensive and the most accessible book on A/B testing statistics I've encountered. If you run tests, then, this book should be required reading. [...] Read it, and then gift it to the rest of your team to read, and watch your program's maturity rise in front of your eyes."
* comprehensive and accessible *
Barry Garner
Sales Development Manager, SiteGainer
"This book transforms the oftencomplex area of statistical modeling into an understandable and useful framework for those will little or no math background. Starting from the ground up, it introduces the reader to important statistical concepts with the goal of making the science behind A/B testing interpretable and useful."
* understandable & useful *
Philip Cross
CEO & Conversion Rate Optimizer, Premeni Services Company
"Finally, a comprehensive reference on statistics for my own industry! No longer do I need 8 different books on statistics that were each bought for only one or two chapters each, because Georgi has delivered a text whereby each and every chapter is applicable to success in my daily work. Thank you for making such a complex subject both interesting and easier to understand."
* a comprehensive reference *
Hristian Kambourov
Chief Conversion Craftsman, Revise
"One of the few books that would actually make you a better marketer and bring you more sales. Period. No fluff, straight to the point, practical manual you need on your desk. Pick up a copy of this book and finally understand how to properly evaluate, analyze and use your data when running A/B tests. All masterfully delivered by a true practitioner and an excellent teacher of the somewhat complicated world of statistics."
* no fluff, practical manual *
Topics covered
A list of the main topics covered in the book. Click here to see the full Table of Contents, as well as a preview of the introductory pages of the book.
Risk management & Decisionmaking
The role of statistics in business decisionmaking. The art and science of balancing risk & reward through optimal test design.
Types of Statistical Hypotheses
Testing for superiority, noninferiority, and strong superiority. Onesided versus twosided hypotheses.
Statistical Models in A/B Testing
Statistical assumptions, assessing model adequacy, nominal vs. actual error rates. External validity of experiment outcomes.
Multivariate tests a.k.a. A/B/N tests
Understanding MVTs. Statistical adjustments for pvalues and confidence intervals. Complications in power and sample size calculation.
Pvalues & Confidence Intervals
Calculation and interpretation of pvalues and confidence intervals. What is and what isn't warranted by a statisticallysignificant outcome.
Sequential Testing
Statistical tests and decisionmaking in the presence of continuous monitoring. Group sequential tests with alpha and beta spending.
Statistical Power & Sample Size Calculations
Calculation and understanding of the concept of statistical power and the false negative error rate. Sample size formulas.
& Much more
Handling segmentation, multiple KPIs, percentage change as KPI, running multiple tests at the same time, communicating statistical results...
Learn how to gain business insights and control risk through statistical methods
This book will help you navigate the statistical jungle of jargon & formulas. You will understand the advantages and limitations of statistical approaches for assessing the uncertainty of data from online controlled experiments. You will learn how to design efficient statistical tests and to adapt statistical models in a number of practical scenarios, primarily from ecommerce A/B testing.
Who is this book for?
This book would be most useful to:
 Conversion rate optimizers
 Landing page optimizers
 Data analysts & web analytics specialists
 Product managers & growth experts
 UX & user testing specialists
About the author
Georgi Z. Georgiev
Founder of AnalyticsToolkit.com
Georgi is a veteran digital marketer, data analyst, and statistician specializing in statistical methods for online controlled experiments. He is the author of the popular blog at AnalyticsToolkit.com, several technical whitepapers on statistical methods in A/B testing, a comprehensive glossary of A/B testing statistical terms, and primary contributor to the OneSided project. He has developed a number of statistical tools used by the A/B testing industry. Georgi is a lecturer at several industry events and the author of the CXL Institute course "Statistics for A/B Testing".