Your payment stack is a P&L line, not plumbing

Most companies that move money treat the payment stack as plumbing: something that should work, stay quiet, and cost as little as possible. Reliability is the whole mandate. That framing is reasonable, and it is also exactly why so much value sits untouched inside it. A payment flow you only manage for uptime is a payment flow you are leaving money in.

Here is the reframe. Every transaction you process has a margin profile: the rails you route it over, the fees you absorb or pass through, the foreign-exchange spread, when the money actually lands, and the mix of payment methods your customers use. None of that is fixed. It is a set of choices, and most teams made those choices once, early, under deadline, and never went back. The result is a flow optimized for shipping, not for economics.

The places value hides are predictable. Processor and interchange fees negotiated at a fraction of today’s volume. Routing logic that sends every transaction down the same path regardless of cost. FX handled at whatever spread the provider first offered, never benchmarked. A payment-method mix that quietly steers customers toward the option that is most expensive for you. Settlement timing and float that no one has modeled as a line item. Each looks small. Together they are often a real share of gross margin, sitting in a system everyone has agreed not to touch because it is working.

I have run this from the inside. At Benevity, the payment engine was not just a cost center to keep stable, it was a place to build. We launched a global white-glove payment service at 94% margin, introduced foreign-exchange spread as a revenue line rather than a pass-through cost, and turned transaction processing into more than $2M in net-new annual revenue, while moving over $3B a year across more than 150 countries. The reliability mattered. But the revenue came from treating the engine as a product with economics, not as plumbing to be ignored.

If you want to find the value in your own stack, start with one exercise: put a fully loaded cost and a margin on every transaction type you process, then sort by volume. The expensive, high-volume combinations are your map. Some you renegotiate, some you re-route, some you reprice, and some you turn into a service customers will gladly pay for. The work is not glamorous, but it is among the highest-return operational work a scaling fintech can do, because it compounds on every transaction from the day you fix it.

One caution, because it matters more here than almost anywhere else. This is client money, and clients feel changes to how it moves faster and more personally than changes to almost anything else. A margin move that quietly raises a fee or reroutes a payment can save real money and cost you a relationship if it lands as a surprise. So the work is never just extraction. It is balancing value capture against trust and retention, sequencing the changes, and bringing clients along, so the economics improve without the relationship paying for it. Done well, the two reinforce each other. Done carelessly, you win the margin and lose the account.

The question is not whether your payment stack works. It is whether it is earning. Run it only for uptime and the answer is almost always no.

This note is part of Bloomera’s Payments & Disbursement Operations practice.

If you suspect there is margin trapped in how you process transactions, that is worth a look. Start with a 30-minute call.