An Indian fast-food restaurant wanted to know if its in-app, online coupons were driving incremental sales.
Analytic Edge was engaged to design a geo-test, to measure the incremental impact of their coupon program.
Our approach included:
- Identifying 3 test cities where the client could publish coupons for the test
- Creating synthetic controls from remaining cities in the client’s operational footprint
- Determining the final impact of the coupon program
AE built custom synthetic controls for the 3-city test cell, using a 2-month pre-period and store-level sales data.
We were able to build controls for the block of 3 test cities, and for each city individually, to look for local effects.
The coupons were offered for a 30 day period, giving us plenty of time to measure the impact.
Test City vs. Synthetic Control Pre-period Match
The test showed little to no impact from the coupons. There was no measurable lift in store sales or orders.
This result held city by city for the 3 test cities, and across all 3 test cities as a group.
AE and the client decided that the coupon, available directly from the ordering app, was probably subsidizing existing customers and purchases with unnecessary discounts. Funding for the program was redeployed to higher ROAS / ROI activities.