The Simility Blog

Oh, those crafty fraudsters: Paying themselves for fake services with stolen credit cards!
Jayan Tharayil
June 14, 2017

Here’s a good case study for you…

Marketplaces have it tough, monitoring millions of services, products, buyers, and sellers for fraudulent activity. And when numerous newly registered users set up accounts with the intent of fraud…they have a huge problem on their hands.

When this marketplace came to Simility, they discovered that registered users were:

  1. Paying for services using stolen credit cards
  2. Setting up fake buyer and seller accounts to launder the stolen credit-card funds via the marketplace platform

The marketplace was using a fraud detection solution to block orders at the transaction stage by evaluating attributes like dollar amount of transactions, buyer name, seller name, and service.

Could they decrease losses further if they detected fraud at the earlier, account registration stage?

That’s where Simility came into the scene. With an adaptive fraud solution that combines flexible data ingression and machine intelligence, Simility deployed a customized device recon model. The details of what they were able to uncover surprised me. (You can read more about how they did it here.)

Simility discovered that organized fraud rings carried out many of the attacks. Most rings used the Pale Moon variant of the Firefox browser. Another finding was the fraud rings had Taiwanese email IDs using bots and scripts, spoofing multiple device IDs for the same device. It’s amazing to me that all of this can be so easily detected with such detail.

Thanks to this intelligence, the marketplace was now able to:

  1. Block 90% of fraud at the account registration stage, well before reaching the transaction stage
  2. Predict 3X more accounts as being suspicious, 65% of which ended up being confirmed bad by either of the following two criteria: connected to other accounts or transactions that were confirmed fraud or displaying other suspicious behavior at the transaction stage

It amazes me what big data, combined with machine learning and talented data scientists, can reveal that helps businesses more effectively combat fraud.

P.S. I love helping people fight fraud…and sharing stories of fraud attempts that have been stopped. Do you have a good fraud story to share? Let me know. I look forward to hearing from you.


About Lynn Strand

Lynn is VP of Marketing at Simility. With over 20 years experience in the security industry—she has a deep understanding of fraud prevention, security technology, and SaaS solutions.