Problem
Small - mid, even large scale retailers struggle to find insights amongst their data. Without loyalty programs, they do not receive demographic or customer information. Even with loyalty programs the usage rate across transactions is low (<50%), therefore there are large data gaps.
A common problem is that ‘Brand A’ who co-funded the cashback offer for the retailer often ends up funding the purchase of a competitor product. Due to dispersed use of the Cashback reward, the funding is overall less, and the offer to the customer is diluted.
Goal
When a customer redeems a cashback offer, to be able to verify the product that is being purchased in the basket and restrict the cashback offer to only certain products and/or categories.
By verifying cashback on particular products, retailers are able to improving the value propositions for brands and increase the co-funding they receive from brands to power cashback offers. Through limiting products they can also increase the cashback offer to the customer, therefore driving greater retail foot traffic.
Output
Created a Python script to match Point of Sale/Receipt data to offer redemptions, therefore enabling card-linked cashback offers to be limited to specific products and/or categories. The offers were surfaced in Commowealth Bank app which is powered by Mastercard, with Slyp performing the SKU validation.
The in-market trial of this was Australian first for SKU and category level automated cashback offers involving Slyp, MasterCard, Commonwealth Bank (Australia’s largest bank), and Chemist Warehouse (Australia’s largest Chemist; 500+ retail outlets).
Implementation:
This involved importing the offer files from MasterCard into Slyp’s database and creating a Python script to find unique fields to accurately match the offer to the existing receipt data in the Slyp platform.
There is no one unique ID available to directly link the records, therefore statistical analysis was performed on the various payment identifiers that are available (i.e., Merchant ID, Card last 4, Card last 6, authorisation codes, time of purchase, purchase location) to identify what would return an accurate match without being overly strict that it rejects true matches and customers do not get their cashback reward (False Negatives)
This logic was embedded in the pipeline and the verified offers were sent back to MasterCard to complete the reward redemption to the customer.
To the left is the live offer in the Commonwealth Bank app that was used for the in-market trial