False declines cost U.S. merchants $157 billion a year — and rejecting more orders doesn't even stop more fraud
Every business running a fraud filter believes it is protecting revenue. The 2026 data says the opposite is often true: merchants are losing more money turning away their own legitimate customers by mistake than fraudsters ever manage to steal from them — and the merchants blocking the most orders are not the ones catching the most fraud. A false decline never shows up as a complaint. The customer just buys somewhere else, and the system that rejected them logs it as a win.
A regular customer's card gets declined at checkout. Nothing is wrong with the card, the funds are there, and the order looks exactly like the last ten they placed. Somewhere in the payment stack a risk score crossed a threshold, or an issuer's fraud model flagged something that has nothing to do with this particular purchase. The customer doesn't see any of that. They see "payment declined," try once more out of politeness, and then leave for a competitor. No one calls support to complain. No one files a chargeback. The order simply never happens, and the business never learns why.
That failure has a name — a false decline — and 2026 is the year the industry finally put a number on how much it costs. PYMNTS Intelligence, running its Optimizing Payments Tracker® series with PayPal Open, put U.S. merchant losses to false declines at an estimated $157 billion a year, on top of payment failures affecting roughly one in five eCommerce orders worldwide — an estimated $47 billion more in annual revenue leaking out globally. Most businesses budget for fraud losses as a cost of doing business. Almost none budget for the cost of rejecting the customers who were never a threat at all.
This piece is about the size of that problem in dollars and in customer behavior, what merchants themselves admit about how often it happens, the industry data showing that blocking more orders doesn't even catch more fraud — it's the maturity of the fraud program that does — and the practical shift that closes the gap without loosening anything that actually matters.
Start with the scale, because it dwarfs what most businesses think they're losing to fraud. PYMNTS Intelligence's Optimizing Payments Tracker® series, run with PayPal Open, put U.S. merchants' annual losses to false declines at roughly $157 billion — money left on the table not because a customer changed their mind, but because a system that was supposed to protect the sale blocked it instead [1]. Zoom out globally and payment failures of every kind — false declines included — affect an estimated one in five eCommerce orders, a leak PYMNTS puts at $47 billion a year worldwide [1]. Fraud losses get a line item on most businesses' books. This number rarely does, because it never looks like a loss — it looks like an order that simply never came in.
The consumer side of that number explains why it stays invisible. A separate PYMNTS Intelligence study found that 56% of U.S. consumers experienced a false decline in the prior three months, and that about 42% of shoppers who hit a failed payment abandon the cart entirely rather than retry [2]. Almost none of them call the merchant to say what happened — they assume the card is the problem, or the site is broken, and they go finish the purchase somewhere that lets them pay. A false decline doesn't generate a support ticket. It generates silence, and silence is exactly what makes it easy for a business to believe its fraud system is working well.
Merchants themselves are starting to admit how often this happens. In PYMNTS' 2026 survey of merchants directly, 47% estimated that up to 5% of their legitimate orders get incorrectly declined as fraud — a self-reported admission, not an outside accusation — contributing to an estimated $50 billion in lost revenue industry-wide [3]. That is merchants describing their own checkout as a place where roughly one in twenty good customers gets turned away by mistake, and still treating the resulting fraud rate as if it were the only number that mattered.
The sharpest data point is the one that breaks the assumption that stricter filtering buys more protection. The Merchant Risk Council's 2026 Global eCommerce Payments and Fraud Report, built from 1,278 merchant professionals across 37 countries with Visa Acceptance Solutions and Verifi, found that merchants with a mature fraud program reject just 2.8% of orders as fraud while holding an actual fraud rate of 0.6% of revenue. Merchants without a mature program reject nearly double the share of orders — 5.2% — and still carry a higher fraud rate of 3.9% [4]. The businesses blocking the most orders are not the ones with the cleanest books; they're the ones with the blunter tool. And the same report found only about 64% of merchants track their false-decline rate at all [4] — meaning more than a third are running a filter with no idea how many good customers it's quietly costing them.
Missing payment methods are a related, separate leak on the same checkout screen. A PYMNTS study with PayPal, surveying 2,179 U.S. adults in June 2026, found nearly 56 million U.S. consumers abandoned a cart in the prior 30 days because their preferred payment method wasn't accepted — about 26 million of them specifically over a missing digital wallet — and one in four digital-wallet shoppers say they'd leave a merchant entirely rather than switch methods [5]. A fraud filter that wrongly blocks a good card and a checkout that never offered the wallet a customer wanted are different failures with the same result: a real buyer, ready to pay, turned away by the system rather than by their own decision not to buy.
A low reported fraud rate feels like proof the system is doing its job, but the MRC data cuts against that reading. It's just as consistent with a filter set so aggressively that it also blocks a meaningful share of legitimate orders alongside the fraudulent ones — you never see the good customers it turned away, because they don't show up as anything. A low fraud number and a high false-decline rate can be the exact same filter, described from two different sides.
That would be true if precision and permissiveness were the same lever, but the MRC comparison says they aren't. Merchants with mature programs reject fewer orders and see less fraud, simultaneously — 2.8% rejected against 0.6% fraud, versus 5.2% rejected against 3.9% fraud for everyone else. The gap isn't how loose the rules are; it's whether the program uses richer signals to tell a real customer from a fraudulent one, instead of one blunt threshold applied to everyone.
Providers execute the rules, but merchants set the thresholds — block every first-time international card, flag every order over a round number, decline on a single risk-score cutoff. Those are business decisions made once, usually out of caution, and then left untouched for years. A provider can hand you better signals; it can't decide for you that a $200 order from a returning customer shouldn't be treated the same as a $200 order from an anonymous new one.
Track your false-decline rate as its own number, next to your fraud-loss number, not folded into it — the MRC data shows two-thirds of merchants don't even do this much. Wire the payment result — the actual issuer decline code, not just "failed" — back into the system that owns the order, so a soft, retriable failure can route through a different path instead of dying silently at checkout. And revisit the blanket rules a provider configured years ago against the customer data you actually have now, because a program that rejects fewer orders while catching more fraud isn't a trade-off — the mature ones in the MRC data are already doing both.
- 01Does your business track how many of your declined orders were real fraud — or do you only ever measure the fraud that got through, never the good orders your own filter turned away?
- 02If a returning, legitimate customer was wrongly declined at your checkout this month, would you know about it — or would it just look like an order that never happened?
- 03Are your fraud rules a policy someone reviews on a schedule, or a set of defaults a payment provider configured once, years ago, that nobody has revisited since?
- 04If 42% of customers who hit a failed payment just leave instead of retrying, what is that actually costing you against the fraud those same rules are preventing?
- 05Is your checkout declining orders your system quietly can't tell from fraud — or is it also missing the payment method your customer actually reached for?
- [1]PYMNTS Intelligence, Optimizing Payments Tracker® Series with PayPal Open — "Merchants Find Revenue Hiding Behind the Decline Button" (August 2026): U.S. merchants lose an estimated $157 billion a year to false declines; payment failures affect roughly 1 in 5 eCommerce orders worldwide, an estimated $47 billion in annual revenue leakage globally.
- [2]PYMNTS Intelligence — "56% of Consumers Have Faced a False Payment Decline" (2026): 56% of U.S. consumers experienced a false decline in the prior three months; about 42% of consumers who hit a failed payment abandoned the cart entirely rather than retrying.
- [3]PYMNTS Intelligence — "47% of Merchants Report False Declines Cost Them Sales" (2026): 47% of merchants estimate that up to 5% of their legitimate orders are incorrectly declined as fraud, contributing to an estimated $50 billion in lost revenue industry-wide.
- [4]Merchant Risk Council — 2026 Global eCommerce Payments and Fraud Report, with Visa Acceptance Solutions and Verifi (1,278 merchant professionals surveyed across 37 countries): merchants with a mature fraud program reject 2.8% of orders while holding a 0.6% fraud rate by revenue; merchants without a mature program reject 5.2% of orders yet carry a higher 3.9% fraud rate; only about 64% of merchants track their false-decline rate at all.
- [5]PYMNTS Intelligence, with PayPal — "The Hidden Cost of Checkout Gaps: What 56 Million Abandoned Carts Mean for U.S. Merchants" (June 2026 survey of 2,179 U.S. adults): nearly 56 million U.S. consumers abandoned an online cart in the prior 30 days because their preferred payment method wasn't accepted, about 26 million specifically over a missing digital wallet; one in four digital-wallet shoppers say they would leave a merchant entirely rather than switch payment methods.
We build payment systems that tell a real customer from actual fraud — instead of guessing at both with the same blunt rule.
Felukaa wires payment results — issuer decline codes, retry paths, method coverage — back into the system that owns the order, so a legitimate customer isn't rejected by a rule nobody has revisited since setup. If you don't know your own false-decline rate, we'll help you find it before it costs you the next sale.
Book a free 15-min consultation