“Fraudsters Just Look Different” And Other Trends From Our Quarterly Report
We’ve launched our quarterly trends report, revealing original insights from Feedzai Research that show fraudsters are attacking in methods that are faster and faster – and continuously new. The time horizon for responding to new fraud is shrinking. That’s why we’re calling the report: The Future is Now: Q1 2018 Fraud Trends Report.
It goes without saying that merchant fraud used to be simpler. It looked like someone waiting in line at the return counter with an item they stole. Today, fraudsters are using advanced technology, coordinating with each other to iterate and innovate, and they’re taking advantage of the complexity of the financial ecosystem itself.
In The Future Is Now: Q1 2018 Fraud Trends Report, we surface the data behind some of the signature characteristics of today’s fraudsters, such as speed, adaptability, persistence, and expansiveness.
For example, the graph below shows that the fraud rate for a given card increases when it has more rule-blocked denials by the issuer. This means that a bunch of seemingly unrelated rules could combine to create a strong signal.
Because our technology stack allows for the rapid iteration of training models, and the rapid deployment of those models into run-time, we can rapidly arrive to new fraud insights like the ones we show in our quarterly fraud trends reports. Then we can operationalize these insights into highly effective models that let our clients detect new fraud patterns as soon as they appear.
The ATO surge
This report also shines a spotlight on “The ATO Surge.” The 2018 Javelin Research study on identity fraud found that 16.7 million people had their identities stolen last year. It’s the highest number since Javelin began tracking the data in 2003. As a result, according to the study, account takeover (ATO) fraud is growing, with ATO losses reaching $5.1 billion last year.
What can merchants do to keep pace with the feverish growth of ATO fraud? In our report, we recommend 6 key actions to take:
- Deactivate sleeper accounts
- Request/require password resets from customers in a given period of time
- Use 2-step verification for sleeper accounts
- Use safe known device security and send a warning when device changes
- Train your customer service teams to recognize the tactics fraudsters use before committing ATO fraud (e.g. the trending pattern of fraud by phone)
- Partner with an agile machine learning platform for risk
Meet Camouflage Fraud
The Future Is Now also features an article from our senior fraud analyst, Joel Carvalhais, who’s named a new kind of fraud: camouflage fraud. He discovered this scheme at a customer before a ring of fraudsters had the chance to wipe out seven figures’ worth of losses.
In camouflage fraud, a merchant will have transactions come in that seem to be made by different customers, when in fact, they’re coming from one fraudster using several stolen cards, devices, and identities.
This is the first time where we’ve seen one fraudster attempting to look like multiple people, and it’s undetectable using traditional methods. At Feedzai, we only discovered it after several analyses where we crossed information along fraudulent transactions to discover the source of what was hurting our merchants. We would have missed it if not for our machine learning technology, which identified similar patterns across different transactions, and our internal data science tools, which allowed us to perform rapid data analysis.
Customer 1 buys; then Customer 2 buys; then Customer 3. The human analyst sees three different people. The rule sees three different transactions. But the machine sees one person – a fraudster – and blocks the transaction before it happens.
Download the report to discover the six key merchant fraud trends of Q1 2018, and to read more about the threats of ATO fraud and camouflage fraud.