Loyalty programs are proven methods for growing and sustaining market share. They increase growth, help retain customers & improve brand reputation. The specifics of each program may differ depending upon the business, which in this case is the production & refining of crude oil, natural gas & petroleum products. A loyalty program was established with the sole purpose of transferring the benefits to customers based on their incremental fuel consumption and thereby maximize profitability. However, malpractices followed at multiple levels coupled with complex interplay by entities involved, which led to revenue leakages and posed a serious threat to critical success factors. The key players involved here were basically Transporters, Owners, Drivers & Franchisees who play significant role in the transactions., which were uniquely identified using Map_Ids.
Our key hypothesis
Potential frauds are happening wherein the Loyalty slabs are played around for garnering more points.
A fraud committed by manipulating the system will not be a one-time activity. Hence there will more than one situations when the fraudulent transactions are carried out by a single entity or a group of entities. Hence, the process of detection of frauds revolves around triangulation while aggregating at various levels and cuts.
Benefits of the solution
The predictive analytics helped to pin-point the most accurate scenarios for likely occurrence of fraud and ensured that concerned stakeholders are notified about the same. Thus the data analytics solution had a significant impact on the baseline performance metrics as well as strategic resources assessment to prevent fraud for the loyalty program.
The insights generated in the initial phases laid foundation for prescriptive analytics. The algorithms developed could help identify churners and not so profitable customers. This helped the business operations to take nextbased action in order to decrease effort-return ratio considerably