Microsoft Uses Machine Learning and Optimization to Reduce E-Commerce Fraud

  • Jay Nanduri ,
  • Yuting Jia ,
  • Anand Oka ,
  • John Beaver ,

INFORMS Analytics Collections | , Vol 50(1): pp. 64-79

Many merchants conduct their businesses through e-commerce. One major challenge in tackling e-commerce fraud results from dynamic fraud patterns, which can degrade the detection power of risk models and can lead to them failing to detect fraud that has emerging unrecognized patterns. The problem is further exacerbated by the conventional decision frameworks that ignore the follow-up decisions made by other associated parties (e.g., payment-instrument-issuing banks and manual review agents). Microsoft developed a new fraud-management system (FMS) that effectively tackles these two challenges. It keeps features used by the machine learning (ML) risk models up to date by using real-time archiving, dynamic risk tables, and knowledge graphs. The FMS uses customized long-term and short-term sequential ML models to detect both historical and emerging fraud patterns. It also makes rapid real-time optimal decisions using a dynamic programming approach to optimize the long-term profit by taking into account the aforementioned multiple-party decisions. After implementing these innovations over a two-year period (2016–2018), Microsoft reduced its fraud loss by 0.52%, thus generating $75 million in additional savings; reduced the incorrect fraud rejection rate by 1.38%; and improved its bank authorization rate by 7.7 percentage points. The result was many millions of dollars in additional revenue. These innovations simultaneously prevent fraud and increase bank acceptance. In April 2019, Microsoft launched Microsoft Dynamics 365 Fraud Protection, a cloud-based service available for all e-commerce merchants.