Facebook recently introduced a new bidding type, "average bidding", as an alternative to the existing bidding type called "maximumbidding". One of our clients, bombabomba.com, decided to test this new feature and would like to do an A/B test to see if averagebidding increase interaction effect. A/B testing has been going on for 1 month and bombabomba.com is now asking you to do this. It is waiting for you to analyze the results of the A/B test. The success criterion for Bombambomba.com is Purchase. Therefore, the focus should be on the Purchase metric for statistical testing.
In this data set, which includes the website information of a company, there is information such as the number of advertisements that users see and click, as well as earnings information from here. There are two separate data sets, the Control and Test group. These datasets are on separate sheets of the ab_testing.xlsx excel. Maximum Bidding was applied to the control group and AverageBidding was applied to the test group.
impression: advertisement views count
Click: Number of clicks
Purchase: The number of products purchased after the advertisements clicked
Earning: Earnings after purchased products