Every deposit attempt ends in one of three places: the issuer says yes, the issuer says no, or the transaction never reaches the issuer because something upstream stopped it. The approval rate is the ratio of the first outcome to the attempts that were actually sent. It is the cheapest revenue lever a trading firm has, because the traffic has already been paid for.
The uncomfortable part is that most operators cannot state their own number. They know what their processor reported last month, aggregated across everything, which hides the only breakdowns that would tell them what to fix.
The number is meaningless until you cut it
An overall approval rate is an average of very different populations. Split it at least four ways before drawing any conclusion. By issuing country, because a card issued in Germany and a card issued in Brazil face different rules and different risk appetites. By card brand and product type, because consumer debit, consumer credit and commercial cards behave differently. By amount band, since a first deposit of fifty units and a top up of five thousand trigger different issuer models. And by whether the payment was a first attempt or a retry.
Once you cut it, the picture usually changes. A firm with a middling headline rate often has one country or one BIN range dragging everything down while the rest performs perfectly well. That is a routing problem, not a product problem.
Track the pre authorisation drop separately. If your own fraud rules reject ten percent of traffic before it reaches an issuer, a beautiful issuer approval rate is hiding the real loss. Both stages belong on the same dashboard.
Why trading deposits start at a disadvantage
Merchant category codes carry reputations. Deposits to brokers and prop firms land in categories that issuers have learned to associate with disputes, which is a separate topic covered in the chargeback guide. Some issuers apply blanket blocks by category and country, and no amount of gateway tuning will move those. Others simply apply a tighter score.
Deposit behaviour compounds it. Traders fund accounts in round numbers, often repeatedly within a short window, sometimes from a device or IP that does not match the billing address country. Every one of those signals is a mild negative in an issuer's model, and stacked together they push transactions across a threshold. Our earlier piece on card approval rates for trading firms goes through the acquirer side of the same problem.
Data quality is the boring fix that works
Issuers approve on confidence, and confidence comes from the fields in the authorisation message. Sending a full billing address, a verified email and phone, a correct merchant descriptor and consistent device data raises the score before any fancy routing is involved. Sending a blank address and a generic descriptor invites a decline.
Descriptor mismatch deserves its own attention. If the name on the statement bears no resemblance to the brand the client signed up with, you get two costs: declines now, and disputes later when the cardholder does not recognise the entry. Fix the descriptor first, then argue about routing.
3DS2 is a tool, not a toll booth
Strong customer authentication under the European rules shifted liability to the issuer when a payment is authenticated, and issuers respond by approving authenticated traffic more readily. The mistake is treating 3DS as a binary switch. The protocol supports exemptions and frictionless flows, so the right setting depends on the corridor, the amount and the issuer's own behaviour. We break the mechanics down in the 3DS2 walkthrough.
Sending everything through a challenge screen will lift issuer approvals while cutting the number of customers who finish the flow at all. Sending nothing through authentication will keep the checkout smooth and hand you declines and liability. The useful configuration sits in between and needs measuring per corridor.
Routing, retries and the multi provider question
Two acquirers rarely perform identically on the same traffic. One has better local connections in Latin America, another has a stronger relationship with a large European issuer group. Once you can see the split by corridor, sending each corridor to its stronger route is straightforward arithmetic. That logic is what a payment orchestration layer exists to hold, and we describe the architecture in the orchestration guide.
Retries need discipline. Card scheme rules limit how often a declined transaction may be reattempted and how those attempts must be spaced, and they distinguish between decline reasons. A soft decline for a temporary issuer condition is a candidate for a second attempt on a different route. A hard decline for a closed or reported card is not, and hammering it produces fees and scheme attention rather than deposits. Our breakdown of decline reason codes maps the common ones.
Make one team own the number
Approval rate falls between departments. Marketing counts registrations, finance counts settled volume, support counts tickets, and nobody counts the attempts that died in between. The fix is organisational before it is technical: one weekly report, cut by country and BIN, showing attempts, pre authorisation rejections, issuer approvals and retry recoveries, with a named owner.
The reporting has to sit where the client record lives, otherwise you end up reconciling processor exports against account balances by hand every month. In our Broker CRM the deposit attempt, the decline code and the client account are one object, which is mostly a bookkeeping convenience until the day you need to explain to an acquirer why a specific corridor spiked.
One caution on chasing the metric too hard. Approval rate can always be raised by accepting riskier traffic, and the bill arrives sixty days later as disputes. The pair to watch is approval rate alongside dispute rate. A rising first and a rising second is not an improvement.
"When a firm tells me its approval rate is fine, I ask for it split by issuing country. Nine times out of ten there is one corridor bleeding, and it has been bleeding for months because nobody looked below the average."
— Roman Onta, Executive Director, SINGUARD
Key Takeaways
- An aggregate approval rate hides the problem: cut it by issuing country, card product, amount band and attempt number.
- Complete billing data and a recognisable merchant descriptor lift issuer confidence before any routing change is made.
- Retry only soft declines, within scheme limits, and preferably on a different route; hard declines must be left alone.
- Watch approval rate next to dispute rate, because looser acceptance buys short term approvals and long term chargebacks.
Frequently Asked Questions
What is a payment approval rate?
It is the share of attempted card payments that the cardholder's issuing bank authorises. It is measured on authorisation attempts sent to the issuer, so a payment blocked earlier by the gateway or by a fraud filter is counted separately. Both numbers matter, because a merchant can have a healthy issuer approval rate while quietly rejecting good traffic upstream.
Why do trading firms see lower approval rates than ordinary online shops?
Deposits to brokers and prop firms are coded to merchant categories that issuers associate with higher chargeback risk, some issuers block those categories outright by country, and typical deposits are larger and more repetitive than retail purchases. The combination triggers issuer risk models more often.
Does retrying a declined card payment help?
It depends on the decline code. Soft declines caused by temporary issuer conditions or insufficient funds can succeed on a later attempt, and card scheme rules set limits on how many retries are allowed and how they must be spaced. Hard declines such as a closed account or a stolen card flag must never be retried, and doing so can attract scheme penalties.