How to increase payment approval rates

First define what you are measuring

The approval rate looks like a simple number: successful transactions divided by all attempts. In practice operators calculate it in several ways and end up comparing figures that cannot be compared. One provider reports approval at the authorisation stage, another counts only transactions that passed 3-D Secure, and a third excludes attempts abandoned by the player.

Before you change anything, agree one definition across all providers and methods. Good practice is to track three levels at once: the share of attempts that succeed, the share of players who ultimately manage to deposit within a single session, and the share of new players who make a first deposit. The last of these shows most clearly how much of your acquisition spend you lose in the cashier.

Break declines down by cause

Decline codes fall into several groups, and each calls for a different response:

• Issuing bank declines – insufficient funds, limits, suspected fraud, a block on the transaction category. Some are final, others are "soft" and may succeed on a retry or with a different method.
• Authentication declines – failed or abandoned 3-D Secure. The user journey is often to blame here: a redirect to the bank's page on a phone, an expired code, no banking app installed.
• Risk declines – the provider's or operator's anti-fraud rules, set too strictly for genuine players.
• Technical errors – response timeouts, misconfiguration, provider unavailability.

In early projects it often turns out that a large share of declines are not bank decisions at all, but the result of configuration, risk rules or abandoned sessions. That is good news, because these are causes the operator can change itself.

Routing and cascades

An operator with more than one card provider can send each transaction where it is most likely to succeed: by card issuing country, card type, amount or player history. A cascade means that after a soft decline the transaction is automatically retried with another provider, without any action from the player.

Cascades work, but they require discipline. Retrying hard declines (for example a card reported lost or stolen) will not improve results and may damage the operator's standing with card schemes and providers. Rules should therefore be built on lists of codes that genuinely have a chance of succeeding on retry, and reviewed regularly.

Authentication without unnecessary friction

In the European Economic Area, strong customer authentication is mandatory for most card payments. It cannot be skipped, but friction can be reduced: pass full transaction data to the 3-D Secure 2 protocol so that the bank can approve more transactions without an additional challenge, make sure redirects work properly in mobile apps, and use tokenisation for repeat deposits by the same player.

Local methods instead of fighting the card

If banks in a given market routinely decline card transactions in the gaming category, giving prominence to a local method is often a better answer than adding more cascades. In Poland the natural choice for licensed operators is BLIK and fast transfers; in other markets, e-wallets or open banking payments. The order of methods in the cashier should follow data on the player's country and device, not the history of integrations.

Withdrawals are part of approval too

A player who waited several days for a withdrawal, or received it by a different method from the one used to deposit, is less likely to return. Fast, predictable withdrawals via the same method as the deposit reduce support tickets and chargebacks. Time from withdrawal request to funds arriving deserves the same careful measurement as deposit approval.

How to measure the effect of changes

Each change to a routing rule or the order of methods is best tested on a portion of traffic, with a control group and a test duration fixed in advance. Compare not only approval but also cost per successful transaction, chargeback rate and deposit value within the cohort. Higher approval at the price of more fraud is not an improvement.

Where to start in practice

1. Standardise the definition of approval and collect raw data with decline codes from all providers.
2. Group declines and estimate how many transactions can be recovered in each group.
3. Fix configuration errors and risk rules that block genuine players.
4. Design cascades for soft declines only and test them on a portion of traffic.
5. Review the order of methods in the cashier by country and device.

If you would like to work through these steps with an external team, describe your situation in the contact form. You can read more about our approach on the Payment solutions page.