The New Economics of Merchant Fraud

Something has changed in the past two years. The merchants trying to get onto your platform, the sellers applying to your marketplace, the businesses requesting payment processing, they don't look like they used to.

The obvious fraudsters are still out there. But increasingly, the applications that should raise red flags look completely legitimate. Professional websites. Realistic product photos. Plausible business descriptions. Clean documentation. Everything checks out on the surface.

Underneath, some of these businesses don't exist at all.

The Asymmetry Has Shifted

Fraudsters have always tried to fake legitimacy. What's different now is how easy and cheap it's become. A synthetic business can be registered in a state with minimal verification requirements for less than $150. An AI-generated website with product pages, About sections, and customer testimonials can be spun up in an afternoon. Fake business documents, invoices, and even promotional videos featuring lifelike synthetic personalities are now trivial to produce.

The barrier to creating a convincing-looking business has collapsed. Between May 2024 and April 2025, scam pages created via generative AI quadrupled globally. That's over 38,000 new fraudulent pages per day. The people trying to get onto your platform have access to tools that didn't exist three years ago, and they're using them.

Meanwhile, the teams reviewing merchant applications are still working the same way they did in 2019. Open the website, click through the product pages, check the business registration, Google the company name, make a judgment call. This process worked when creating a fake business required real effort. It doesn't work when a sophisticated fraud ring can generate dozens of plausible-looking storefronts faster than your team can review them.

The Incentive Problem

Every platform faces the same tension. Merchant acquisition matters. Seller growth matters. The business wants to onboard as many legitimate merchants as possible, as quickly as possible. Friction in the application process means lost revenue.

At the same time, onboarding a bad merchant creates real costs. Chargebacks. Regulatory exposure. Reputational damage. For platforms in regulated industries, the consequences can be severe. For marketplaces, a single high-profile fraud case can erode customer trust.

The result is a constant pressure to approve faster while the risks of approving wrong keep growing. Manual review becomes a bottleneck. Backlogs build up. Reviewers face pressure to move quickly, which means less time to catch subtle signals. The sophisticated fraudsters, the ones who've invested in looking legitimate, slip through.

What Manual Review Misses

A human reviewer can spot obvious problems. A website selling clearly prohibited products will get flagged. A business with no web presence at all will raise questions. But the merchants who are actively trying to evade detection have learned what reviewers look for.

They know to create a professional-looking site with enough legitimate products to pass a quick scan. They know to bury the prohibited items deeper in the catalog, or hide them behind login walls, or list them using euphemistic names. They know to register their business in a state with minimal verification and create documentation that looks authentic. They know that a reviewer with a queue of 50 applications isn't going to spend two hours investigating each one.

Manual review also struggles with what's not visible on the surface. A website might look legitimate, but are the prices realistic for what's being sold? Does the business model make sense? Are the product images stock photos used on dozens of other sites? Is there evidence that this business actually operates, or is it a shell designed to move money?

These questions require investigation, not just observation. They require cross-referencing multiple sources, checking for patterns, reasoning about whether what you're seeing adds up. That's time-consuming work, and time is exactly what most review teams don't have.

Continuous Risk

Initial onboarding review is only part of the problem. A merchant who was legitimate when approved can change. Business models shift. Product catalogs expand into prohibited categories. A company under financial pressure might start cutting corners or engaging in fraud.

Most platforms don't have the resources to continuously re-review their merchant base. The focus is on getting new merchants through the door, not monitoring the ones already inside. This creates a window where a good merchant gone bad can operate for months before anyone notices.

For platforms that take on financial risk, like buy now pay later providers or payment processors with chargeback exposure, this ongoing risk matters as much as the initial decision. The merchant who seemed stable at onboarding might be a liability six months later.

What Deep Investigation Looks Like

Roe's Merchant Risk Agent approaches merchant review as an investigation, not a checklist. When an application comes in via API, the agent conducts a comprehensive review that typically takes 2 to 8 minutes depending on complexity.

The agent analyzes the merchant's website at depth, examining product listings, pricing, and business model claims. It checks whether prices are plausible for the category, whether product images appear elsewhere on the web, whether the site structure suggests a real operation or a hastily assembled front. It can detect content that exists in the site's HTML but isn't visible to casual browsers, a common technique for hiding prohibited products.

Beyond the merchant's own site, the agent searches public records, social media, news coverage, and other sources to build a picture of whether this business is what it claims to be. It looks up beneficial owners and checks them against sanctions lists and adverse media. It reasons about whether the pieces fit together: does this business make sense?

For platforms that care about more than just compliance, the same investigation informs underwriting decisions. A merchant with a legitimate web presence, consistent social media activity, and realistic pricing is probably a better partner than one with a thin digital footprint and prices that seem too good to be true.

The output maps to whatever decision framework you use. Approve, reject, or request additional documentation. Each recommendation comes with specific findings and evidence, so your team can verify the reasoning on edge cases.

For ongoing monitoring, the agent can periodically re-review merchant websites and catalogs, flagging material changes from what was present at onboarding.

The Path Forward

The economics of merchant fraud have changed. Creating a convincing fake business used to require significant effort. Now it requires a laptop and a few hours. The tools your reviewers use haven't kept pace with the tools fraudsters use.

Roe's Merchant Risk Agent conducts deep investigations in 2 to 8 minutes per merchant, analyzing websites, product listings, public records, and more. It works at onboarding and on an ongoing basis. It catches what surface-level review misses.

If your team is under pressure to approve faster while the risks keep growing, we should talk.

Schedule a demo to see how this works with your merchant risk workflow.