Every fraud loss has an origin story, and most start at the same place: the moment an account is opened. New account fraud is what happens when that front door fails, when a fraudster uses stolen, fabricated, or synthetic credentials to create an account that looks legitimate from day one. Once they are inside, everything downstream, from the first transaction to the eventual bust-out, is cleanup on a problem that should have been stopped at the threshold.
I am the President and Chief Commercial Officer of Socure, where we exist to make it less painful for good people to prove they are who they say they are, when opening an account, when coming back to authenticate, at any moment trust has to be re-established. When Filip Verley put five rapid-fire questions to me on Liminal’s Friday Five, my answers kept returning to an uncomfortable pattern: the industry knows where fraud starts, and it still is not fighting hardest at the front door. Here is the longer version, with the room the format does not allow.
The basis point I thought everyone would fight for
Filip asked what assumption about my market turned out to be wrong. Candidly, I thought customers would care more about stopping fraud. I thought they would go and fight for that extra basis point of fraud capture, because at the scale of a bank or a large fintech, a basis point of new account fraud is real money, real victims, and real synthetic accounts maturing quietly inside the portfolio. It turns out that with so many competing priorities and so many understaffed fraud teams, the fight is not getting the prioritization I expected. That is disappointing, and I say that as someone who wants customers to push us harder, not less.
The economics make the deprioritization stranger. A fraudulent account caught at creation costs almost nothing. The same account, caught 6 months later, has a transaction history, a credit line, possibly a role in a mule network, and a remediation bill many multiples larger. Buyers in this market are already acting on that math in one respect: 66% say they are seeking to replace their traditional onboarding solutions because detection is falling short. The appetite to change tools exists. The willingness to fight for the last basis point of capture, the part that separates adequate from excellent, still does not.
The next 12 to 18 months: further behind, faster
For what will matter more over the next 12 to 18 months than most people are preparing for, my answer is uncomfortable. In so many cases, we are already far behind where fraud is today. What worries me is how much further behind we will be, because I do not see the level of acceleration in investment needed to keep up with the pace at which fraudsters are innovating.
New account fraud is where that gap compounds fastest. Generative tools have collapsed the cost of producing convincing identities, documents, and application histories, so attackers can now open accounts at a volume and quality that manual review was never built for. Our own research found fraud rings generating thousands of synthetic identities and launching tens of thousands of attacks inside a single month, in one case pushing nearly 25,000 synthetic identities into more than 35,000 applications, often within 48 hours of the identity being created. The broader numbers show the same direction: reported consumer fraud losses in the United States hit a record of about $16 billion in 2025, an increase of roughly 25% in a single year. The attackers are compounding. Defense budgets, mostly, are not.
The fix starts with how you buy: run a proof of value
Filip also asked what trade-off buyers in our category still underestimate, and I answered: value. Buyers underestimate how they build the business case that determines which new account fraud-prevention solutions they should use. They do not test the products as much as they should, and they mostly buy off marketing. That needs to change.
The discipline that changes it is a proof of value. Not a proof of concept that shows the technology works in general, but a measured answer to the only question that should close a deal: on your traffic, your population, and your fraud patterns, how much does this solution capture, and at what cost to good customers? Take a retrospective sample of your own account opening decisions, include the fraud you caught late and the good applicants you declined, and let each vendor score it. The numbers that come back are the business case. They are also how a fraud team wins the prioritization fight internally, because a basis point of capture with a dollar figure attached stops being a technical metric and starts being a budget argument.
The hill I will die on: show up on time
Filip closed by asking for a small hill I would die on that most people think is trivial. Mine is punctuality. I spent 11 years in Army Special Operations, and being on time, especially in combat, meant the difference between life and death. I have carried that standard ever since. It sounds far from fraud strategy, but it is the same conviction underneath: standards are not trivial, and the ones you hold when nobody forces you to are the ones that decide outcomes. Stopping fraud at the front door and testing before you buy are those kinds of standards. Hold them.
Key takeaways
- New account fraud is the origin point, and it is underprioritized. Fraud caught at account creation costs a fraction of fraud caught months later, yet understaffed teams and competing priorities keep the front door underdefended.
- The basis point of fraud capture is worth fighting for. At bank and fintech scale, marginal capture is real money and real victims. Most buyers have stopped fighting for it, and 66% are looking to replace onboarding solutions that fall short.
- Fraud is compounding faster than defense investment. Socure research found fraud rings pushing nearly 25,000 synthetic identities into more than 35,000 applications, often within 48 hours of creation, while United States consumer fraud losses hit a record of about $16 billion in 2025.
- A proof of value beats buying off marketing. Testing vendors on your own traffic and population produces the business case, wins the internal budget argument, and closes the gap between marketed and measured performance.
