Soft declines are costing you revenue — here’s how topropay recovers 38% of them
Here is an uncomfortable number for any high-volume merchant: somewhere between 10% and 15% of attempted card payments end in a decline that the issuer would have approved on a different routing path. Not fraud. Not insufficient funds. A “soft” decline — the kind triggered by issuer caution, network noise, or a stale BIN heuristic — and one that quietly drops off the order page before anyone in finance sees it as lost revenue.
topropay was built to catch exactly those payments. Its cascade engine retries soft declines server-side, in the same authorization request, against the next-best acquirer in a ranked waterfall. The published recovery rate sits at 38%. For merchants who want to see the routing logic behind that number, the smart routing and cascading page breaks down how each retry is scored and routed.
The rest of this piece is a closer look at what soft declines actually are, why they are so expensive at scale, and what changes once cascading is doing the work.
What a soft decline really is
A hard decline is final. The card is blocked, the account is closed, the funds are not there. A soft decline is everything else — a temporary rejection from the issuer that, on a second attempt through a different route, will often clear.
Common soft-decline codes include:
- 05 — Do not honor (the catch-all; often issuer-side risk caution)
- 51 — Insufficient funds (sometimes timing, sometimes a stale balance check)
- 91 — Issuer unavailable (transient network issue)
- 62 — Restricted card (BIN or country mismatch)
- N7 — CVV2 failure (often a stripe-read issue, not actual fraud)
The catch is that issuers behave differently depending on where the authorization arrives from. A request through acquirer A in Germany can return code 05, while the same card, on the same day, through acquirer B in the Netherlands, returns approved. That asymmetry is exactly the gap topropay is engineered to exploit.
The math on what you are losing
Pick a merchant processing €100M in annual card volume with a 12% decline rate. That is €12M in attempted payments not clearing.
- Suppose 70% of those declines are soft (industry midpoint): €8.4M
- Even a conservative 30% recovery rate would clear another €2.5M
- topropay’s published 38% recovery rate would clear roughly €3.2M
- For subscription businesses with churn risk attached to failed rebills, the downstream value is materially higher
This is the line item most merchants underestimate, because it never appears in any report. The customer either retries on their own, abandons, or — worst case — completes the purchase with a competitor. Nothing in your funnel analytics distinguishes “lost to soft decline” from “did not want it badly enough.” topropay turns that invisible drain into a recovered, measurable revenue line.
How the topropay cascade engine works
Cascading is not the same as “trying again.” A naive retry against the same acquirer will almost always return the same decline, because the path is identical. What recovers soft declines is changing the path.
The topropay engine does this in a single round trip:
- Initial authorization is sent to the highest-scored acquirer based on BIN, currency, MCC, 3DS outcome, and live approval rate
- If the response is a soft decline, the engine keys on the exact ISO-8583 code, BIN, currency, and issuer behavior to pick a different acquirer
- A retry fires immediately, against that next acquirer, in the same server-side request
- The cascade continues up to a configurable hop depth until the payment clears or the waterfall is exhausted
- Your checkout receives one clean result — approved or declined — with no visible delay to the shopper
Total decision time, end to end, lands around 47 ms. The shopper sees a successful payment. Your engineering team sees a normal approved response on the webhook.
Why most merchants cannot build this themselves
Plenty of payments teams know cascading would help. Very few actually ship it. The reasons are practical:
- It requires multiple acquirer contracts and integrations already running in parallel
- You need a live, per-acquirer scoring model — not a static priority list
- ISO-8583 decline mapping varies subtly between acquirers and needs continuous calibration
- Retry depth, retry timing, and idempotency keys all have to be exactly right to avoid double charges
- The whole thing has to fit inside one HTTP round trip from the checkout’s perspective
topropay handles all of that as infrastructure. The cascade is not a feature you opt into per transaction — it is the default behavior of the network, and it tunes itself based on live per-acquirer approval data flowing through every merchant on the platform.
What gets retried, and what does not
A common worry from compliance and finance teams: “is the cascade going to retry things it should not?” The short answer is no.
topropay only cascades on decline codes that are demonstrably recoverable on a different path. The engine does not retry on:
- Fraud-flagged declines from issuer-side risk systems
- Hard authentication failures (3DS rejection)
- Closed-account or stolen-card responses
- Customer-side cancellations or timeouts
What it does retry on are the issuer-cautious, network-transient, and routing-sensitive codes — the ones where the live data plainly shows a different acquirer will clear the same card.
Where the 38% number actually lives
The 38% recovery rate published by topropay is not a synthetic figure. It is the median lift observed across merchants on the platform with cascading enabled and at least three acquirers in the waterfall. The distribution is wide:
- Single-acquirer merchants who add one secondary route typically see 18% to 25% recovery
- Three-acquirer setups land in the 30% to 40% band, which is where the median sits
- Merchants with deep local-method coverage (iDEAL, Bancontact, SEPA alongside cards) push the number higher by routing to an alternative method when cards exhaust
The lever, in every case, is the same: when an issuer hesitates on one path, give it another path before the shopper notices.
Where the impact lands hardest
The merchants who see the biggest swing from topropay’s cascade engine tend to share a few traits:
- High order frequency or recurring billing (subscription, SaaS, streaming, gaming)
- Cross-border European checkout, where each border adds issuer caution
- High average order value, so each recovered transaction is worth more
- An existing multi-acquirer setup that is not yet orchestrated
Subscription businesses in particular see the compounded effect. A recovered first bill is not just one saved transaction — it is the full customer lifetime value that would have otherwise been lost at the first failed charge.
The takeaway
Soft declines are the most expensive payment problem nobody on your team is reporting on, because by definition they never show up as a completed event in any system. topropay surfaces them, retries them on a different acquirer inside the same request, and recovers an average of 38% of the revenue that would otherwise have silently disappeared.
If you want a concrete read on what that translates to for your specific volume, the topropay team will model the lift against your last 90 days of authorization data in a short discovery session. The exercise is brief, and the number is usually larger than expected.