The only roadmap you need for AB testing

You can learn anything with a solid roadmap.

Why?

All it takes to achieve anything is consistent effort, step-by-step execution, and a clear path to follow.

When I only dived into Data science (10 years ago), the greatest roadmap I found is stanford lectures, the only challenge is that it was theory but practice.

When I started launching the experiments I couldn’t find anything even similar to this and I just had to pick it up through my mistakes, experience, papers and articles.

Here’s how I’d approach it today and what you should cover to understand how to launch the experiments in a right way.

I’ll start with the basics, which are enough to get you started, and then move to advanced concepts that will help you go faster and deeper.

The basics

To launch an experiment effectively, there are six main sections you need to understand. For each section, I’ve included the key questions you should be able to answer:

1. Reason - why do we need experiments at the first place,
   Why are experiments the gold standard? Why run them whenever possible?