Matt DeLauro, President, GTM at SEON, explores the core imbalance in modern retail: the gap between how quickly retailers are scaling online and how effectively they are managing fraud risk.
Retail has never been better positioned to grow. Today’s e-commerce platforms make it possible to launch storefronts quickly, expand into new markets and process transactions at global scale. Payments are embedded, customer journeys are streamlined and operational barriers are lower than ever. Speed has become the defining advantage.
But that speed comes with a blind spot. Modern retail is designed like an open, high-traffic space – easy to enter, easy to move through, optimised to keep people flowing. That’s great for customers who expect seamless experiences. It’s also great for fraudsters, who benefit from the same openness.
Here’s the core problem: fraud controls remain heavily concentrated at checkout, while fraud itself spans the entire customer journey. Retailers are detecting threats at the point of purchase rather than at the point of entry. And that gap is getting wider as they scale.
Built for speed, not for scrutiny
Most of retail’s fraud strategy still lives inside the e-commerce platform. Built-in controls are simple and fast – pre-configured rules, automated checks, seamless payment integrations. For teams moving quickly, that convenience is hard to pass up. According to the 2026 Fraud and AML Leaders Report, 62% of retail organisations rely mostly or entirely on platform-native tools for fraud prevention.
The problem is that these tools were designed to enable transactions, not to defend against sophisticated fraud. They assess risk at the moment of payment and not much else. It’s like installing cameras at the till but not at the entrance. Among retailers that depend on these controls, nearly half report that fraud losses are growing faster than revenue. The tools that support scale are not keeping pace with the risks that scale introduces.
More growth, more gaps
Nearly 88% of retailers report revenue growth of 10% or more over the past year. That’s more customers, more accounts and more payment options. Each one expands the number of ways fraud can get in.
At the same time, responsibility for fraud is rarely owned by one team. It’s spread across e-commerce, payments, operations and finance, with each function working from different systems and different data. That makes it very difficult to connect the dots. A minor anomaly in one system might be part of a much larger pattern, but no one sees it because the data isn’t shared.
Fraudsters understand this. Rather than targeting the most visible controls at checkout, they focus on earlier stages of the journey where oversight is weaker, such as account creation, login and promotion sign-up. In many cases, fraud is already well underway before a transaction ever happens.
Detecting fraud in the wrong place
The data confirms this. The most damaging types of fraud in retail all begin before checkout. Account takeover (26%), promotion abuse (18%), return fraud (18%) and loyalty fraud (13%) all originate well before a payment is processed. By the time a fraudulent transaction occurs, the real damage has already been done. Accounts have been compromised, identities manipulated and incentives exploited.
Retailers are catching fraud at the till when it started at the door.
AI has helped surface more of these signals, but it hasn’t solved the underlying problem. Ninety-seven percent of organisations now use AI in fraud workflows, yet 97% also plan to increase fraud headcount. More signals are being surfaced, but those signals still need people to interpret them, investigate them and act on them. Retailers can see more fraud than ever. They just can’t stop it fast enough.
Moving detection to where fraud actually starts
If fraud begins at account creation and login, that’s where detection needs to be – well before checkout controls.
That means paying attention to how customers behave across the full journey; not just at the moment of purchase. Patterns in login activity, account changes and promotion usage can flag risk early, but only if that information is captured and connected across systems. Fraud signals rarely exist in isolation. They show up across multiple points and only make sense when those points are linked.
AI is important here, though it’s only as good as the data behind it. More signals alone won’t help if those signals are scattered across disconnected tools. What matters is whether the right information reaches the right system at the right time.
This isn’t about adding more controls. It’s about extending visibility to the places where fraud actually begins, so that by the time a transaction reaches checkout, most of the risk has already been addressed.
The real risk
Retail’s growth trajectory isn’t slowing down. Customer expectations for fast, seamless experiences will only increase. And fraud will evolve alongside both, becoming faster, more automated and harder to catch with traditional tools.
The retailers that manage this well won’t be the ones with the most controls at checkout. They’ll be the ones that figured out the fight starts much earlier.
The doors are open. The question is whether you know who’s walking through them.
