The Traitors in your payment stack

Is it the ambiance, the contestants? We know for sure it isn't the tasks. So what is it?

It's that you, the viewer, know exactly who the traitors are, and you get to sit back and watch everyone else fail to work it out; the guilty pleasure we simply can’t deny. 

In the game, nobody at that table has any evidence. They vote on who looked shifty over breakfast, while the traitors nod along and sometimes lead the charge against someone innocent.

So, what has that got to do with payments?

Well, if your costs went up last year, and you want to know why, this might help explain why you’re struggling to identify the culprit. 

If you sell payments rather than buy them, you already know this game from the other side of the table, and you've watched a merchant vote against you for a decision you didn't make. 

Before we sit down…

To work out who's really at that table, Aevi's in-house payments experts - Eddie Johnson, Charles Leeming, and Adam Myers talked through what they see in the field, and where the money goes once it leaves a merchant's account.

You don't need to have watched a single episode to follow this, by the way. If you've ever looked at a payments invoice with no way of checking whether the number was fair, you already know how this game plays, but here's the key:

• The faithfuls = you, and everyone else trying to work out where the money actually went 

• The round table = the meeting where rising payments costs get discussed

• The traitors = the costs quietly taking a cut that nobody can account for

• The vote = blaming whichever part of the setup has a visible price tag

• The prize pot = your margin

So, if you’re 100% faithful, this one’s for you… 

…Somewhere on the statement is a percentage, and that percentage covers interchange, scheme fees, and the markup of everyone standing between the terminal and the account. 

The industry calls it a blended rate, which is a polite way of saying that everyone's costs have been stirred together until they can't be separated.

The problem is, say in the first quarter the statement reads 1.4%. Inside the blended rate though sits interchange, which the card scheme sets and nobody in the chain controls, the scheme's own fees, the acquirer's margin, a gateway fee, and a cut for whoever sold the merchant the setup in the first place. 

Then in the next quarter it reads 1.5%. Every one of those parties can look you in the eye and say it wasn't them, and every one of them might be telling the truth.

Could the blended rate be the traitor? 

When that number moves, there's no way to ask which part moved. Interchange might have shifted. A scheme might have added a fee. Someone in the chain might have adjusted their margin, and the average absorbs it without a ripple. 

It's an alibi that doesn't require anyone to lie.

And the instinct is to assume the vagueness is deliberate. Sometimes it is, but often it's what happens when you try to summarize something genuinely enormous.

The card is read, the transaction routed, checked, authorized and settled, with hundreds of small decisions made between the tap and the money landing. Each can carry a cost, and that cost varies by card type, by transaction value, by how the card was presented. An invoice reflecting all of it would be accurate and unreadable.

So, is anyone actually being cheated here, or is this just what a complicated system looks like from the outside?

That's the question, and from where a business is sitting there's no way to answer it. The genuine reason not to break the number down and the self-interested reason not to break it down point the same way, so nobody has to conspire for this to work.

But itemization was never really the issue. Nobody wants a three-hundred-line invoice. What a business wants is to know whether the number is fair, and to be able to move if it isn't.

A blended rate takes away both. You can't challenge a component you can't see, which means you can't negotiate one either - the only lever left is to replace the entire arrangement. That's a serious thing to threaten and an expensive thing to carry out, and everyone at the table knows it.

And it isn't only the business that loses here. When you can't tell one party from another, they all get treated the same way, so the gateway charging exactly what it quoted sits under the same suspicion as the one quietly widening its margin. Neither can prove anything either way. 

That's what a room does when it has no evidence and still has to decide something. It picks the suspect in front of it.

In payments, that suspect is almost always sitting on the counter - the terminal.

Walk into enough shops in the UK and you'll find one. A handwritten note taped to the card reader: no Amex.

The terminal didn't decide that.

There's your wrongful banishment, right there. 

The terminal is the thing customers complain about, so the terminal receives the vote, and the decision that caused it was made in a conversation the merchant was never part of.

…and those two things have almost nothing to do with each other.

The terminal sits on the counter. Everything else happens somewhere nobody can picture, and a business can't suspect what it can't picture. So when the network is slow, the terminal gets blamed. When the fees climb, the terminal gets blamed for that too.

Which produces the outcome you'd expect. The provider gets replaced, the business feels decisive about it, and the next statement reads the same as the last one, because the number was never coming from there. 

In short, a faithful goes home, the round table reconvenes next quarter, and the prize pot is smaller than it was. 

If you're on the other side of this, that faithful is you. 

Anyone who has sold payments has been on the receiving end of that vote - banished over a pricing decision made three seats up the chain, with no way to prove otherwise, because the statement doesn't show whose money is whose. 

Here's the one advantage the faithfuls have over the traitors in the show - The list of who it could be is finite, and the answer is somewhere on that list. 

For payments though, it doesn’t quite work that way. 

Everything up to this point has assumed the cost eventually shows up on a statement, and that the problem is working out which line it's hiding in. 

It’s just that some of the largest costs in payments never appear on any statement at all. 

Start with what a business isn't selling. An estate that can't take the payment method someone wants turns customers away, and no statement records that. The invoice covers what you paid on what you accepted, and stays silent on everything you didn't.

That silence has become more expensive. Not long ago people carried a card and cash as backup. Now they leave the house with a phone, and if a shop can't take how they want to pay, there's another shop somewhere close by. 

…when systems don't reconcile, someone has to make the numbers agree. That means people, hours, and a headcount line that lands in payroll. It never enters a conversation about payments costs, because on paper it isn't a payments cost.

Ask most businesses what their payments setup costs them and they'll read you the rate. Ask what it costs to make the numbers agree at the end of every month and you'll usually get a pause.

The subtler one is what all those workarounds do to your judgment…

Years of adapting to a system's shortcomings become the standard every alternative gets measured against. A better setup looks wrong because it doesn't accommodate a problem you shouldn't have had.

By this point a reasonable person would settle on a working theory: trust the parties you chose deliberately, and treat the invisible ones as the problem.

The show has a mechanic for exactly that mistake. Traitors recruit. Someone spends weeks genuinely on your side, gets a tap on the shoulder, and carries on behaving exactly as before.

Payments does this constantly, and it rarely involves bad intentions.

Nothing changed character. The business outgrew the terms, and terms that were generous at one volume become a ceiling at another.

The same thing happens at industry level, well out of sight. Hundreds of mandates come down from the card schemes every year, all of them compulsory, most of them invisible to the merchant. Complying with one can mean six months of work, and that work gets paid for somewhere, usually in a transaction fee with no visible cause. The party passing it on is a faithful doing exactly what the schemes require, and it still arrives looking like a markup.

Say a business beats all of it. It works out which part of the setup is expensive, builds the case, gets to a conclusion it can defend.

It has now done the hard part, and it still can't act, because in payments the vote is the cheap bit. Removing someone is where the cost is.

An hour of terminals not working is money that doesn't come back. And plenty of migrations aren't chosen at all: a security standard expires, hardware falls out of compliance, and a business whose setup works fine is told to replace it and carry the risk of doing so.

On top of that sits a commercial structure that makes staying put the rational choice. Terminals tied to a provider, a provider tied to an acquirer, every renewal starting the clock again. Which produces one of the more honest sentences anyone said on the call.

That's the whole thing. Bad arrangements seem to survive because everyone has worked out the cost of leaving and concluded it isn't worth it. 

In the show, banishment costs a vote. Here it costs your estate, and the traitors are pricing accordingly.

Which reframes the problem the article started with. Identifying the traitor was never really the point…being able to remove one is.

Decoupled layers mean the terminal isn't married to the acquirer and the software isn't married to the hardware, so changing one part doesn't force changing all of them. That's where an in-person payment orchestration approach becomes valuable, and it's a different job from the e-commerce version people usually picture, because here the terminals, the certifications and the keys all have to move independently of each other before anything else can.

You wouldn't use it often. That's fine. Being able to is what changes every conversation you have with everyone at that table.

It also makes the changes you didn't choose survivable.

A merchant keeps the legacy system running, gets a new terminal set up alongside it, and swaps over when ready - no gap in trading, no weekend spent hoping. A setup you can change one piece of is a setup where a price rise is a decision, rather than something that arrives out of the blue.

There's one more player worth naming, and it's the one this whole article has walked past without suspecting.

In the show, every series has someone who reaches the final without a single vote against them. In payments, it isn't a card fee at all, it’s cash. 

Card fees get scrutinized because they're legible. Cash gets a pass because its costs are scattered across budgets nobody connects to payments, and because it never sends you an invoice.

Go back through the suspects…

• The blended rate that nobody can be pinned to. 

• The terminal that took the blame for a decision made three seats away. 

• The lost sale, the reconciliation headcount, the partner whose terms slowly became a ceiling. 

• The cash drawer that everyone treats as free money.

No single one of them is the traitor. 

What they have in common is that none of them can be identified from where you're sitting, and none of them can be removed without taking half the setup with them. And every quarter they go unnamed, the pot gets smaller. 

That's the real traitor. Not a single provider, fee or platform, but the combination of opacity and dependency that keeps every party in the chain unaccountable.

Which is why the fix was never going to be better detective work. Every faithful at that table is reasoning perfectly well on the information available. The information is the problem, and so is the price of acting on it. 

Because knowing who the traitor is only matters if you can banish them, right?