A Data Science Approach in Product Analytics


A significant part of product development is A/B testing. Simply put, this is where companies and product managers test to see a new version of their product versus the older one to make sure it’s worth publishing their features to the entire user base. In order to do this, an A/B test needs to be set up where two user groups, test and control, respectively use the new and current version of the feature. The change in the feature can be as modest as changing the color of a button.

For an A/B test to be effective, the test and control