A new way to prevent payment scams in real time

Scams are personal. I’ve seen it in my own family.

When my mom retired, she started receiving scam calls almost immediately. In fact, on the very day she retired, the number of calls seemed to increase dramatically.

They became so frequent that she eventually decided to disconnect the landline altogether.

Most of the scams were obvious, and she recognized them for what they were. But there was one that almost worked.

She was about to send money to a scammer and could have lost her savings. Fortunately, she was talking to my father, and together they realized what was happening before she made the transfer.

I still think about how different things could have been if she'd sent that money. It would have had a significant impact on her life and on our entire family.

What bothers me most is how deliberately scammers exploit people. They take advantage of vulnerabilities and use whatever tactics they can to manipulate someone into sending money.

And this is happening everywhere. People are losing savings they've spent years building, sometimes with devastating consequences for their families.

I imagine most of us know someone who has been scammed or has come dangerously close.

That's a big part of why I'm so excited about what we've been working on at Incognia. I believe we have an opportunity to make a meaningful difference in preventing these crimes.

I think the industry has given up too easily on scam prevention

One of the biggest challenges with scams is that everything can look perfectly legitimate on the surface.

The customer is using their own account, on a trusted device, from a familiar location. They're the ones authorizing the transaction, even though someone has manipulated them into sending money.

And identifying the recipient isn't necessarily easier.

Many scams involve mule accounts that have passed identity verification and may have been operating normally for months before being used for fraud.

Financial institutions have developed different ways to identify these scams, but I'm skeptical about relying too heavily on some of the signals being used today.

Take hesitation during a transaction, for example. Some solutions try to identify whether a customer seems uncertain or is taking longer than usual to complete a payment. But that could just be someone who's unfamiliar with the technology or maybe distracted while doing it.

Or consider whether someone is on a phone call while sending money. That can certainly be an indicator of a scam, but people make legitimate payments during phone calls all the time.

These signals can be useful, but they're noisy. And I think we need stronger evidence to identify when someone is being manipulated.

Sometimes I feel like parts of the industry have become too willing to accept that scams are simply too difficult to prevent.

I don't think we should accept that.

At Incognia, we could have followed the same trends as everyone else and invested in similar capabilities. There's always pressure to pursue whatever technology is getting the most attention from investors or competitors.

But I've always believed there's more value in going deeper into what makes your technology unique.

For us, that's our ability to understand physical-world behavior through apartment-level location intelligence.

So we kept investing in that capability and looking for new ways to apply it to difficult fraud problems.

And when we started exploring how it could help with scam detection, we found something that genuinely surprised us.

A new way to identify scam risk in the physical world

We started investigating whether historical location behavior could tell us something about the relationship between the person sending money and the person receiving it.

Specifically, we wanted to understand whether evidence of a real-world connection between two people could help us distinguish legitimate payments from scams.

And what we found was more compelling than we expected.

Legitimate payments frequently happen between users with physical-world connections

Think about going to a restaurant with friends. One person pays the bill, and everyone else sends them their share afterward.

Or think about sending money to a family member or someone you know.

In both cases, there's a good chance you've spent time together in the physical world. You have a real-world relationship. And that relationship can provide useful context about the payment.

When we analyzed our data, we found that approximately 79% of legitimate transactions occurred between users with evidence of a physical-world relationship.

Meanwhile, approximately 86% of reported scams involved transactions between users with no observed physical-world relationship.

We were able to identify signals indicative of these relationships by looking at historical location behavior.

What I find fascinating is how intuitive this is once you think about it. Of course people send money to people they know and spend time with.

But until now, it's been difficult to use that information in real-time fraud prevention. Traditional location signals like IP addresses and GPS coordinates lack the precision and historical context needed to reliably identify these patterns.

Scams were concentrated among payments without apparent physical-world connections

When we looked more closely at transactions between people without an apparent physical-world relationship, we found a much higher concentration of scams.

In fact, just 3% of transactions accounted for 65% of the scams in our analysis.

That was a big discovery for us. It suggested that understanding the relationship between the sender and recipient could provide a valuable new signal for identifying suspicious payments.

Of course, plenty of legitimate transactions happen between people who have never met. Someone might be paying an online seller or sending money to a person who lives across the country.

A lack of an observed physical-world relationship doesn't automatically mean a payment is fraudulent.

But when we combine this intelligence with other risk signals, we have a new way to identify potentially suspicious transactions in real time, while there's still an opportunity to intervene.

How we’re bringing this into Real-Time Scam Prevention

These discoveries became the foundation of Incognia's Real-Time Scam Prevention.

The idea is to help banks identify potential scams by looking beyond the person sending the money to understand who is receiving it, their relationship to the sender, and how they're connected to other accounts.

At the center of the product is our Physical Social Graph, which uses historical location patterns and persistent device intelligence to find signs of real-world relationships between people.

This is possible because of the location technology we've spent years developing.

Our patented Trusted Locations technology uses multiple sensor signals to identify places that devices visit frequently. With apartment-level precision, we can identify patterns that traditional location signals like IP addresses and GPS miss.

We also use persistent device intelligence to recognize devices even after they change or reset. Together, these technologies help us build a more complete picture of people's physical-world behavior and how they may be connected.

We can also find connections across accounts that might otherwise appear unrelated.

For example, if several accounts receiving suspicious payments are being accessed from devices that frequently visit the same apartment, that could be a sign of a coordinated scam or mule operation.

We combine this information with other fraud signals, such as signs of mule activity, active phone calls, remote-access tools, and device tampering.

Banks receive a real-time risk assessment and individual signals they can use alongside their own information to decide whether a payment needs further attention before the money moves.

In an analysis of transactions from a bank with more than 80 million active users, Incognia identified approximately 70% of scam transactions.

Changing the economics of scams

When I was asked what excites me most about this launch, my answer was simple: Bankrupting the scammers.

Today, scammers can target people anywhere in the world without ever meeting them. They can operate remotely and convince people to send money to accounts that appear legitimate.

I believe our technology could make that much harder.

If banks can identify signs of a scam by looking at the real-world relationship between the sender and recipient, scammers may need to change how they operate to avoid detection.

And those changes could make their operations more expensive and harder to scale.

Of course, I don't expect scams to disappear overnight. Scammers will find ways to adapt, and it will take time for more financial institutions to adopt this technology.

But I believe we have a real opportunity to make scams much less profitable.

When I think about what happened to my mom, and all the other people whose lives have been affected by scams, that's what makes this work so meaningful to me.

If we can help banks catch more scams before the money moves, we can help protect people from losing the savings they've worked so hard to build.

—

Read more about Incognia's Real-Time Scam Prevention and how physical-world intelligence can help banks identify scams before money moves.

See how your bank can catch scams that traditional fraud controls might miss. Schedule a demo.

Subscribe to The Signal: FinServ for more insights on scams and fraud prevention every other week.