AB Testing - Shoefly

An A/B testing analysis project for Shoefly, a fictional shoe company. The goal is to determine which of two ad variants performs better using click-through rate data.

Dataset

ad_clicks.csv contains ad impression records with the following columns:

Column	Description
user_id	Unique identifier for each visitor
utm_source	Traffic source (google, facebook, twitter, email)
day	Day of the week the ad was shown
ad_click_timestamp	Time of click (null if not clicked)
experimental_group	Ad variant shown to the user (A or B)

Analysis

The analysis is split across two files:

• script.py — standalone Python script
• ab_testing_shoefly.ipynb — Jupyter notebook with the same analysis

Steps

1. Traffic source breakdown — count views per UTM source
2. Click-through rate by source — percentage of users who clicked per source
3. A/B group distribution — verify even split between groups A and B
4. A/B click comparison — overall clicks for Ad A vs Ad B
5. Day-by-day breakdown — click-through rate per day of the week for each ad

Conclusion

Ad A outperforms Ad B on most days of the week (all except Tuesday), and has a higher average click-through rate overall. Ad A is the recommended choice.

Requirements

• Python 3
• pandas

Usage

pip install pandas
python script.py

License

This project is licensed under the MIT License.