Major project
Retail, E-commerce & Consumer Goods
Uplift model: who should get the offer?
Predict who buys because of an offer, not who would have bought anyway, and compare the profit with mass mailing.
The brief
Sending an offer to everyone wastes discounts on people who would have bought anyway. Using a randomised advertising experiment, build an uplift model that finds the customers an offer actually changes, and compare profit from targeting them against mailing everyone or the likely buyers. Good looks like: an uplift curve, a profit calculation with the cost of the offer, and a plain explanation of why this beats a normal response model.
What you will build
- Load the experiment with treated and control groups
- Check that the groups are comparable
- Build a response model as a baseline
- Build uplift models (two-model and class-transformation)
- Evaluate with Qini and uplift curves
- Calculate profit under each targeting policy
Free, open data to build it with
Sign up free to get the data and links
It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.