Reset Password
If you've forgotten your password, you can enter your email address below. An email will then be sent with a link to set up a new password.
Cancel
Reset Link Sent
Password reset link sent to
Check your email and enter the confirmation code:
Don't see the email?
  • Resend Confirmation Link
  • Start Over
Close
If you have any questions, please contact Customer Service

What the Fake is Going On.  

thewildcouple969 46M/45F
2 posts
4/6/2022 5:19 am
What the Fake is Going On.


The Fake Profiles that We Deal With

6 APR 2022

If there’s one thing we know, it’s that fraud comes in many unsavory flavors. From purchases made with stolen credit cards to phishing schemes, fraudsters are always looking for new and novel approaches to scamming victims.

Back in October, we did a deep dive into transactional fraud, sifting through Sift data to discover which U.S. states had the highest fraud , as well as creating a profile of the fraudulent person in America. Now, with Valentine’s Day approaching, we thought we’d focus on a different (but also painful) type of fraud: fake<b> profiles </font></b>on dating sites.
The high cost of romance scams

There are a variety of reasons someone might create a fake profile on a dating site, from the curious (“I wonder if anyone would respond to someone like this?”) to the insecure (“What if I looked like this instead?”) to the downright criminal. Sometimes, fake<b> profiles </font></b>are set up by organized crime rings that use bots to send phony messages and coax victims into parting with their .

Romance scams are a huge, costly, and disturbing problem. According to the FBI, romance scams cost victims more than $82 million in the last six months of 2014 alone, with the average victim losing more than $100,000. Yes, that’s five zeros. Ouch.

For the dating sites that host these fake profiles, the problem can also come with damaging consequences. Their brand reputations are at stake. User experience suffers. And internal teams often find themselves devoting more time than they’d like to identify and dealing with these pests, which – despite the company’s best efforts – keep popping up again and again.
Scope of the problem

We’ve already learned that romance scams – however they’re perpetrated – can be costly. But how rampant are fake dating profiles? We analyzed a sampling of more than 8 million<b> profiles </font></b>created in the past year on dating sites that use Sift to find out how many phony<b> profiles </font></b>had been blocked during that period.

The results? We discovered that % of all new dating<b> profiles </font></b>created were fake. We also found that:

Male<b> profiles </font></b>are 21% more likely to be fake than female<b> profiles
</font></b>The most common age listed on fake<b> profiles </font></b>is 36
However, users listing their age as 64 had the highest fraud . One factor contributing to this is the relatively of dating site users in this age group.

Location, location, location

Location is a common signal used, in conjunction with other clues, to determine whether a is a fraudster. So, what about dating site users? Typically, location is determined via shipping, billing, or IP address – but in this case, we took the location directly from what someone had filled in on their profile.

When looking at where the “users” in these<b> profiles </font></b>hailed from, we discovered that Nigeria, Ghana, the Netherlands, Romania, and South Africa had the highest fraud . Surprised? Most people are familiar with Nigeria’s reputation for scams. However, we don’t recommend blocking users based on a single factor like country – even if it’s appearing at the top of our list. Creating rules like this is far too black-and-white to effectively deal with something as nuanced as fraud, and you run the risk of inadvertently blocking good users.

welcome--2x
Fighting fakes at scale

That’s why dating sites – and other sites in which users create profiles, like social networks, marketplaces, and job sites – often turn to a machine learning-based solution to help automate the discovery of phony<b> profiles. </font></b>While many of Sift’s customers use us to reduce chargebacks, a significant segment is more focused on weeding out fake users and<b> profiles </font></b>before they actually harm their legitimate customers.

Our algorithms process a variety of potential fraud signals, both industry standards (like IP address, account age, location, etc.) and custom data chosen by the individual site (like, say, whether someone has uploaded a profile photo) to identify the<b> profiles </font></b>most likely to be phony before an unsuspecting person has a chance to get conned.

The good news is that the<b> profiles </font></b>we looked into never saw the light of day since they were preemptively blocked or deleted after being flagged as phony. Still, users of dating sites should – as always – stay vigilant and practice healthy skepticism.

Become a member to create a blog