Practical A/B Testing

Working with Different Types of Data

The goal of A/B testing is to observe what happens as you introduce changes to the product. To do this, you need data, a lot of data. You also need different types of data. Having different categories of data will help you with the following:

• Enable you to cross-reference and validate outcomes to increase confidence in the analysis.

• Give you the capability to monitor your engineering system health that creates the product experience.

• Provide you with a deeper, richer context for data analysis.

Let’s look at the different categories of data that will create a deeper understanding of how your changes impact the product.

Defining User Engagement Data

To understand what users are up to on the product, you’ll ...