A high level introduction to applying pandas to analyzing A/B testing result. In this example, we will look at the result of a simple A/B testing experiment, where users were divided into two groups - treatment and control group. The treatment group received a personalized email about a campaign, while the control group received a normal email with no personalization. Our target metric is the conversion rate, ie. we will investigate whethere the treatment group has a higher conversion rate.
This exercise will comprise of the following:
- User allocation: inspect the number of users being allocated to treatment group and vice versa
- Lift and statistical significance: evaluate the result of the A/B test using two key measurements, lift and statistical significance
- Segmentation: understand the result of the A/B test in the context of user segmentation
This is an exercise taken by DataCamp's Analyzing Marketing Campaigns with pandas Case Study, and all credits go to them. You can check out the course, and if interested purchase access to DataCamp here: https://www.datacamp.com/courses/analyzing-marketing-campaigns-with-pandas