Business Challenge
This leading supermarket in the Middle East wanted to devise an integrated marketing strategy based on customer purchase behavior in stores. It wanted to change the merchandizing based on the items being purchased by the customers.
Solution
We deployed a strategic segmentation and automated scoring algorithm with the following features.
- Stratified random sampling technique was used on the transaction data.
- 2% of transactions were selected as an unbiased sample to develop segmentation of transactions.
- K-Means algorithm was used to segment transactions and identify product affinity across transactions.
- Scoring algorithm was deployed for periodic scoring of customers.
Impact
- Actionable Insights:Client could rapidly translate insights into more effective merchandizing and CRM campaigns.
- +10% Increase in Trip Dollars: This was achieved by rewarding large baskets with extra coupons. This also significantly improved retention rates.
- +2% Increase in Transaction Profitability: Analysis indicated that ~60% of transactions were made in the evenings. Rationalizing impactful promotions post 5.00 PM helped the client achieve increased profitability.
