Project developed for the module eCRM, directed by Prof. Dr. Philippe Baecke. The project consists in a logistic regression analysis of the Dunnhumby database to improve the predicted coupon redemption rate. Tool used: Excel, KNIME for data mining.
1. Increasing Profit
by Improving Predicted
Coupon Redemption
Group 6
Aparna Sosale | Melanie Kaiser | Lara Zaccaria | Tove Perlhede | Christer Lundmark
2. Strategic targeting to narrow
down the most valuable
customers.
Increase
ROI
WHY PREDICTIVE MODELLING?
• To support business decisions
with strong, actionable insights
gathered from customer data.
• To highlight future possibilities for
better investment opportunities.
1
Optimal number of
people to send the
coupons to
3. BENEFIT OF PREDICTIVE MODELLING
2
14.5%
25%
39%
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%
40.0%
45.0%
RANDOM MODELLING HUMAN JUDGEMENT PREDICTIVE MODELLING
7. RESULTS OF DIFFERENT CUT-OFFS
5
Maximum
ROI
12% 12% (55.6%) 20% (€495,000)
Beyond 15%
Net Profit
increases, ROI
decreases
Best Cut-
Off
Maximum
Profits