Payments stopped being plumbing, and most boards haven‘t noticed

No serious business keeps all of its money in one bank. Treasury teams diversify across institutions as a matter of basic prudence: to manage counterparty risk, negotiate better terms and avoid being held hostage by a single relationship.

 

Yet a striking number of fast-growing global companies do precisely the opposite when it comes to the function that actually moves that money. They process every transaction through a single payment provider, integrated years ago when the company was small. Payments are plumbing, and as long as the pipe doesn’t burst, no one looks at it.

 

That assumption is now expensive, and gets more so with every decline in growth. For digital businesses scaling across the borders between subscription apps, cross-border retailers and marketplaces, payments now affect margin, retention and the speed at which a company can enter a new market.

 

The hidden tax on growth

 

The numbers are easy to miss because they hide inside transactions that look successful enough on a dashboard. A business might see an authorisation rate of 85 per cent and conclude that things are fine. The gap between 85 per cent and 92 per cent is real money – money that was attempted, intended and then declined for reasons that often have nothing to do with the customer’s ability to pay.

 

Here’s a concrete example. JP Morgan is the largest card issuer in the United States. If a British software company expanding into the US routes all of its American transactions through a European acquirer, every JP Morgan-issued card on its checkout is being processed by a bank that has no relationship with the issuer. The transaction looks fine on paper, but the acceptance rate quietly underperforms.

 

Move that same flow to JP Morgan as the acquirer and the rate jumps, because the issuer recognises its own data and trusts it. This is basic relationship economics applied to payments, and most businesses don’t run their stack this way because they can’t.

 

The acceptance rate difference between two reputable, tier-one acquirers on the same transaction can reach as high as 5 per cent. On a sufficiently large book, that gap is the difference between a profitable quarter and a flat one.

 

There is also the matter of what happens after a decline. A failed payment is not necessarily a lost customer. In many cases, it is a temporary fault: an issuer flagging a routine charge; a token mismatch; a network glitch. Sending the same transaction through a second provider, a process the industry calls cascading, recovers a meaningful share of these, typically around 8 per cent of initially declined transactions. According to our estimates, nine in ten businesses don’t run cascading at all.

 

Why payments became a board-level issue

 

According to Benchmarkit’s 2025 SaaS performance survey, the typical software business now spends around two dollars in sales and marketing to win each dollar of new annual recurring revenue – a 14 per cent rise on the previous year. The bottom quartile of companies spend close to three. Every percentage point lost at the checkout is therefore not a clean operational cost. It is a multiple of hard-fought acquisition spend that evaporates at the final step.

 

The shift to subscription and recurring revenue has turned payment continuity into a retention problem. Involuntary churn – customers lost when their card expires, is reissued or hits a routine decline – is now one of the largest single sources of subscriber attrition. Research by Recurly across 1,200 subscription businesses found that the average company stands to lose around 7.2 per cent of its subscribers every month to payment failures alone, unless decline-management tools are actively running.

 

And the default has become global. A fitness app launching out of London now sells into Brazil, Germany and Indonesia within months of its first product release. In Brazil, the dominant consumer payment method is PIX, the instant-transfer rail used by more than 180 million Brazilians – around 90 per cent of Brazil’s population. A European payment provider that doesn’t support PIX leaves the business invisible to most of its addressable market in the country. A single processor cannot deliver best-in-class performance across all of them, and pretending otherwise is now visibly costly.

 

In our experience, the threshold at which this stops being theoretical and starts costing real money is around $20 million in annual processing volume. Below that, the optimisation gains are usually too small to justify the engineering work. Above it, the leakage compounds.

 

The status quo trap

 

The obvious question is why so many businesses haven’t already diversified their payments. The answer is usually complacency. Most companies pick a payment provider in their first year and integrate it tightly. Two years later, the same provider is woven through the checkout, the subscription engine, the reconciliation tooling and the fraud workflow. Adding a second provider feels like duplicating effort, and can, indeed, cost engineering time and headcount to manage.

 

Meanwhile, the cost of staying put is diffuse – a basis point of margin here, a few churned customers there – while the cost of changing is concrete and immediate. The diffuse cost wins, year after year.

 

The deeper problem is structural. No single payment provider can give independent advice on routing across the full market, because no single provider operates across the full market. And while many processors have added orchestration features to their own platforms, the real payment orchestration layer sits above the providers, not within any one of them. A processor that also offers routing can optimise across the acquirers it operates. What it cannot do, purely structurally, is route across the full market, including to its own competitors.

 

The competitive consequence is already visible

 

In the older model, the business outsourced the entire function to a single provider and accepted whatever performance came with it. In the model emerging now, the business retains control of the routing logic, customer data and commercial relationships through an orchestration layer that sits above the underlying providers. Multiple acquirers and payment methods plug in as interchangeable connectors underneath.

 

Routing then becomes a strategic choice rather than a technical default. A business might keep a local acquirer in the first position because the commercial pricing is favourable, even at a slightly lower acceptance rate, and cascade to a tier-one acquirer for retries to capture the recovered volume. Or the inverse: lead with the acquirer that maximises authorisation in a given corridor and use cheaper rails as the fallback. These are commercial decisions, not engineering ones, and they belong to the business.

 

Two businesses with identical products, marketing and funnels can produce materially different financial outcomes simply because one of them runs a more sophisticated payment infrastructure. The one with higher authorisation rates retains more revenue per dollar of acquisition spend. The one that can switch on local payment methods enters new markets faster. The one that recovers involuntarily churned subscribers compounds its retention curve.

 

None of this is exotic technology. It is, increasingly, the table stakes of operating a global digital business. The boards asking the right questions have already started. The ones that are not will find themselves on the wrong side of a comparison they did not know they were running.

 

The lesson treasurers learned a generation ago applies here, too. Concentration is rarely the cheapest option. It is just the option whose cost takes the longest to surface.

---

To learn more, please visit solidgate.com

---

Solidgate is a payment orchestration platform for global online businesses and digital commerce. The platform provides connectivity to over 150 acquirers and payment methods, intelligent routing and a unified payment infrastructure across more than 200 markets in a single integration.