Case Study:  Transforming Bluestem's Fraud Detection and Prevention with AI/Machine Learning

Company Profile

  • Consumer lender and online retailer of electronics and general merchandise
  • Parent company of seven eCommerce retail brands
  • Annual sales of US$1.5B

Goals

  • Reduce fraud incidences
  • Improve customer experience by creating frictionless process
  • Lower false positives
  • Improve fraud review timeliness
  • Increase fraud team’s efficiency

Results

  • 60% reduction in “Friendly Fraud” rate
  • $5M+ reduction in annual losses from Friendly Fraud
  • 20%+ faster time-to-approve good customers

The Problem

Lack of Visibility Hampered Fraud Team

Bluestem Brands is an eCommerce company with a diverse portfolio of brands.  In 2017, they sought a partner to help them reduce fraud, while improving the customer experience. As is true of all digital merchants, they were constantly under a barrage of a wide variety of fraud schemes, including:

  • Fraudulent Chargebacks,
  • Synthetic Identity Fraud,
  • Account Takeover.

Bluestem faced an additional challenge in finding a flexible solution that would work with their various brands, and payment and lending options. Because they were built through acquisitions, there was no single fraud system in place across the portfolio. Their fraud team was hampered by the manual effort required, lack of visibility, and slow turnaround times. This situation resulted in an inability to quickly and efficiently detect and prevent fraud. 

The Solution

Robust and Flexible Platform to Detect Fraud

To meet Bluestem’s needs, Fraud.net architected a solution that offered greater power and flexibility to detect and prevent fraud, along with creating a frictionless customer experience. Working together, processes were transformed through automation, data orchestration, and use of analytic tools. The solution included:

  • AI-based platform and models to stay ahead of the variety of fraud schemes encountered,
  • Robust toolset and unified dataset for fraud team.,
  • Data consortium to offer timely insights regarding the organized fraud rings that were especially damaging to Bluestem,
  • Linked entity analysis to bring data to life through visualizations for quickly identifying and addressing the fraud attacks.

“Fraud.net has helped us immensely in simplifying our fraud team’s work. It has freed them from tedious manual reviews and pulling data together from disparate systems. The Fraud.net system provides us with all the tools and relevant data for improved fraud detection and prevention.” 

VP Fraud Prevention, Bluestem Brands


AWS Infrastructure + Fraud.net’s Intelligence Layer =
A Complete, Cloud-born Fraud Management Platform

The more you move mission-critical processes into the cloud, the greater your agility as an organization, and the faster you will pull ahead of your competition.  But in doing so, you need new frameworks that offload non-core functions to partner specialists offering highly scalable, highly resilient systems-of-record on which you can rely.  Only then can you focus on thrilling your customers by doing what your organization is known for.

Let Fraud.net on AWS help you to:

  • Isolate fraud to deliver immediate bottom-line improvements,
  • Gain real-time risk intelligence across your applications and services,
  • Remove friction from your new customers’ applications and transactions,
  • Grow revenues by confidently expanding into new products and geographies.
Get Started Today