Optimisation (Payment Optimisation)

Payment optimisation is the ongoing work of making payments succeed more often, cost less and feel smoother. It is not one product or one switch. It is a set of small, testable changes across the whole path: the checkout screen, the data sent with each request, the route chosen, and what happens after a failure. The aim is to stop losing sales that the customer meant to make. That covers a card that could have been approved, a retry that did not happen, and a form that asked one question too many.

The idea took hold because payment success is not fixed. The same card, at the same shop, for the same amount can be approved or declined based on the data sent, the route used and the time of day. Once teams could see that spread, they started treating payments as something to tune rather than plumbing to install. Rules play a part too. Authentication rules in the UK and the EU include carve-outs for lower risk payments, and using them well is a large part of the job. Results vary widely by market, card mix and product, so this is an area to test rather than assume.

Where The Losses Hide

Four places, mainly. The checkout, where a long form or a missing local method costs a sale before any bank is asked. The request, where thin data gives an issuer less to work with. The route, where one acquirer may do better than another for the same card. And the retry step, where a failed payment is dropped rather than tried again in a sensible way. Each one is easy to measure on its own. That is what makes the work doable.

Data Quality In The Request

An issuer decides with what it is given. A request with a clean address, a full merchant name, a sensible category code and complete cardholder data reads as a normal purchase. A thin one reads as a risk. Small fixes matter here. A billing descriptor the customer knows cuts disputes. And a correct merchant identification number keeps history tied to the right shop. None of this is glamorous, and it is often the cheapest gain available.

Routing And Acquirer Choice

Where a payment is sent shapes whether it is approved. Local acquiring in a market usually reads better to a local issuer than a cross-border request. Some acquirers handle certain card products better than others. Transaction routing rules can act on card country, product type, amount or past performance. The value comes from measuring, not guessing: run a share of traffic each way and compare. That is the thinking behind smart routing as a live tool rather than a fixed setting.

Identity Checks And Carve-Outs

A strong identity check cuts fraud and adds a step. That step costs some sales. The rules allow for that balance. A low value payment exemption can apply below a set amount, with a running total ceiling and a cap on how many in a row. A second route allows risk analysis on each payment. It applies where a firm's fraud rate sits under a set level. The figures and conditions are set in the rules, revised from time to time, and differ outside those markets. Using an SCA exemption well means knowing which one applies and who carries the risk when it does.

Reading A Decline

A decline is a message, not a verdict. A soft decline asks for another try with more detail, often an identity check. A hard decline says do not ask again. Treating them the same way wastes good payments and annoys issuers. Reading the decline code and acting on it is the core skill: retry the retryable, step up where a check is wanted, and stop where the answer is final. Scheme rules limit how often a request may be repeated, so a retry plan needs care.

Keeping Stored Cards Alive

Repeat billing leaks money quietly. Cards expire, get replaced, or change number. An account updater service refreshes stored details before a payment fails. Network tokenisation goes further. It swaps the card number for a token the network keeps current. Both cut the number of customers who churn for no reason but a stale card. For a business with repeat billing, this work is often worth doing first.

How To Measure It

One number is not enough. An approval rate that rises while the conversion rate falls means the checkout got harder. Fraud and dispute rates belong in the same view, since a gain bought with losses is not a gain. Cut the data by market, card product, route and channel. A blended figure hides the cases worth fixing. And set a baseline before changing anything, so improvement can be told apart from noise.

Where To Start

Fix data quality first, since it is cheap and helps every later step. Test routes rather than picking one on a hunch. Build a retry policy with rules per decline code. Turn on stored card refreshes for any repeat business. Review checkout friction with real session data, not opinion. And keep a change log, so a shift in numbers can be traced to a cause. This guide to checkout optimisation and payment setup covers the front end, and this piece on improving approval rates across markets covers the back end.

‍