The hidden cost of payment failures: a $500B problem

The payments industry talks a lot about fraud. We talk far less about the mirror-image problem: false declines. A false decline is when a legitimate transaction is rejected by the issuing bank or payment processor. It happens roughly 15 times more often than actual fraud — and it's costing the global economy an estimated $500B annually in lost revenue.

The scale of the problem

According to data aggregated across Zupay's merchant base, approximately 8.2% of payment attempts are declined. Of those declines, our analysis suggests that roughly 40–45% are false positives — legitimate transactions incorrectly flagged. That means for every 100 payment attempts, around 3.5 genuine purchases are being turned away.

Why do banks decline legitimate transactions?

Issuer decline logic is a black box — banks don't publish their rules, and they change them constantly. But from pattern analysis, the most common causes of false declines are:

• Velocity triggers — multiple purchases in a short window flags as suspicious, even when it's obviously legitimate (e.g., buying rounds at a festival).
• Cross-border pattern mismatches — a UK cardholder buying from a Singapore merchant triggers geographic anomaly detection, even when travel is common.
• Unusual merchant category codes — first-time purchases at new merchant types are disproportionately declined.
• Amount anomalies — a purchase 3× larger than a cardholder's typical spend is flagged, even when it's a planned large purchase.

What the industry can do

The structural solution is better data sharing between merchants, processors, and issuers. 3DS2 and its exemption management framework is a step in the right direction — it allows trusted merchants to pass additional context to issuers that can reduce false declines. Network tokenisation (replacing card PANs with merchant-specific tokens) also significantly improves approval rates by providing issuers with more stable identity signals.

What Zupay does about it

Our approach operates on three fronts. First, intelligent retry logic: when a transaction is declined with a soft decline code, we automatically retry through an alternative acquirer within 800ms. This recovers approximately 18% of soft declines. Second, we use network tokenisation by default for all merchants, which has increased our baseline approval rate by 2.1 percentage points. Third, our smart exemption engine automatically requests Transaction Risk Analysis (TRA) exemptions from 3DS for low-risk transactions, reducing unnecessary authentication friction that causes cart abandonment.

Looking forward

The industry is slowly moving toward real-time issuer communication standards that would allow merchants to flag disputed declines instantly. Until that infrastructure matures, the best tools we have are smart routing, network tokenisation, and better use of the exemption frameworks that 3DS2 already provides. We expect approval rates to improve industry-wide by 2–3 percentage points over the next 24 months as network tokenisation adoption reaches critical mass.