Revenue reporting is the central nervous system of business. Every decision, every bet, every course correction runs through numbers we trust to be complete. But here's the question no one is asking: what if they're not? What if we've been optimizing around a blind spot, and calling it performance?
There is a system inside your business making revenue decisions every day. You didn't build it for that. Your fraud program was built to stop bad actors, and it does. But every transaction it blocks is also a revenue decision. And not one of those decisions appears in your revenue reporting. Not one. Yet everything else follows from it.
The number not in your board deck
Every transaction that reaches checkout gets a decision: approve or decline. That decision is made billions of times a day by systems built to stop fraud, not protect revenue.
The gap between them has a price tag. Research consistently estimates that roughly $50 billion in legitimate U.S. orders are declined annually. Not fraud stopped, but real customers with valid payment methods who couldn't complete their purchase. Industry estimates put the false decline rate at 2% to 10% of all declines. For a mid-sized retailer processing five million transactions a year, even the low end means approximately 150,000 orders that never closed. At the current U.S. e-commerce average order value of $183, that's around $30 million in vanished revenue with no alert, no report, and no flag.
Then it compounds. About 40% of customers who experience a false decline don't return. You paid to acquire them. You got them to checkout. Your fraud system handed them to a competitor. That outcome doesn't appear in any fraud report. It surfaces, if it surfaces at all, as a quiet gap in repeat purchase rates attributed to pricing, product, seasonality—anything except the actual cause.
This is a measurement problem, not a technology problem
The reason customer insults go unmeasured isn't missing data but missing accountability. The data exists. The problem is that businesses rarely build a metric around it, so no one owns it. Fraud teams are held to fraud losses, chargeback rates, false approval rates. Those metrics are clean, visible, and well-managed. Measuring customer insults is challenging and circumstantial, so it's rare for businesses to optimize for it. That's not a failure of rigor. It's a rational response to the incentive structure. False declines sit outside that perimeter because they were never put on anyone's scorecard, not because they’re invisible. You get what you measure. Nobody measured this.
I lived this at Sun Basket. As CFO, I watched teams optimizing hard for metrics that looked great in isolation and were quietly eating the business at the boundaries. The dashboards looked fine. The business wasn't. The fix was rarely more sophisticated tooling. It was changing what we held people accountable for. When the measurement changed, the behavior changed.
In retail fraud, the measurement problem is compounded by a structural one. Fraud prevention typically reports into security, risk, or technology, not into the revenue function. The person accountable for customer acquisition has no visibility into how many acquired customers the fraud system is rejecting. The person running fraud has no stake in the acquisition cost of the customers they're blocking. There is no shared metric that forces those two functions to look at each other's numbers. That gap is the problem. Technology is secondary.
Why the status quo is getting more expensive
The problem is only getting worse. Two reasons.
Fraud has gotten more patient. The high-volume attacks that rules-based systems were built to catch are giving way to slow-burn approaches like accounts built slowly to look legitimate. Harder to catch. More expensive when they land.
Meanwhile, legitimate customers have gotten harder to recognize. AI shopping tools, new device patterns, accelerated checkouts—a loyal customer using a purchase assistant can look, to a legacy system, like a threat. A fraudster with three months of account history may transact with less friction than a good customer on an unfamiliar device.
Without the best technology, you lose on both ends. Approving fraud. Blocking revenue. Without the right measurements, you'll never catch up.
The two questions that change everything
What did we stop? And what did we cost ourselves stopping it?
Those two questions, asked in the same meeting, by the same senior leader, are the difference between organizations that manage this problem and those that don't. The fraud operations that get this right have a CFO, COO, or CRO holding both numbers simultaneously.
Most retail organizations can answer the first question in detail. Very few can answer the second at all. Close that gap. That asymmetry is a choice; usually an unconscious one. Not because technology doesn't matter (it does), but because no technology solves a problem that leadership hasn't decided to measure.
