Case study: an electricity retailer, $100m of sales, and $109k of savings

As I have said in earlier articles, electricity and gas retailers are primed to earn savings from a couple of specific innovations. The longer they wait, the more they miss out. Don’t rely on your bank to raise it — you need to understand what is out there.

So let’s do a case study. Call the company Best NRG Co.

What this covers. These numbers are savings on scheme and interchange fees only. A business also pays payment gateway and merchant acquiring fees, so depending on those rates the final outcome will differ. This assumes no change in gateway cost or acquiring bank margin — then, at the end, I test what happens if both go up.

Figures are as at 2022. Interchange caps change on 1 October 2026, so treat the method as the durable part and the rates as historical.

Where Best NRG started

Their payment gateway and acquiring bank could not offer scheme (network) tokenisation or least cost routing to eftpos online. The CFO, John, felt like they were missing out. He was right.

The gateway was a bank’s own, running old XML APIs, with no real investment going into it. Changes arrived years after the more nimble players had shipped the same thing. John had been promised improvements by his relationship manager and his transactional banking specialist, and the deadlines kept slipping. It was starting to feel like the bank did not care as much about its clients as it used to.

John was open to change. That is a huge step toward getting a better deal.

The numbers going in

They are not a huge company, and not a small one. They have pushed customers toward card payments over the last few years and away from bank-to-bank direct debit, because they found fewer doubtful debts and less work for the collections team. Card is now 80% of payments — mostly monthly, some still quarterly.

Note the debit share. At 65% they have a more cost-conscious customer base than the market average — these are consumers who do not carry credit cards.

Which cards actually benefit from tokenisation

Best NRG provided a card mix and we worked out which cards would gain from scheme tokenisation. All cards are currently processed as recurring payments, so they already benefit from the interchange that comes with that.

It is no surprise that 100% of debit cards can benefit — Visa and Mastercard are actively competing against eftpos, which cannot process credit card transactions.

Because Best NRG’s customers are focused on price, they do not carry many premium credit cards, so the proportion of credit cards that gain is lower than a business with a different customer base would see. There are still good savings in it.

Step one: scheme tokenisation on its own

John was pleased — and what pleased him beyond the number were the other benefits of tokenisation: a higher transaction approval rate, and better security.

He was already reasonably happy. But I had more to show him.

Step two: routing eligible transactions to eftpos

Debit cards are 65% of all payments to Best NRG, and we assume 90% of those are dual-network, meaning they can be routed to eftpos. I asked John to check his own debit card — it had the eftpos logo on it, so I told him that was a card that could be routed.

Look at that carefully. Routing everything to eftpos saves less than routing only Mastercard debit.

Does that look strange? It shouldn’t, once you see why. Visa has a low interchange fee for tokenised recurring payments, which together with their scheme fees makes it cheaper to keep those transactions on Visa than to route them to eftpos online. Mastercard’s interchange for online tokenised is lower still than Visa’s.

I reminded John that the cost is not just interchange — there is also a scheme fee, and the two have to be looked at together. Companies that look at interchange in isolation make errors. He told me his corporate banker had never explained any of this. I was not surprised.

This is the part most people get wrong. Least cost routing is not “route everything to eftpos.” It is route what is genuinely cheaper, which requires knowing the interchange category each transaction lands in.

What recurring interchange was already worth

To show John what the recurring categories were saving him, here is the same analysis assuming Visa sat in tokenised online interchange instead of recurring:

The Visa saving is still fairly small, but it makes it worthwhile to route those transactions to eftpos online.

I also told him I keep on top of interchange changes, and that an ongoing relationship means I can tell a client when a change affects them.

And if there were no tokenisation at all

Everything above assumes credit card transactions are scheme tokenised. If a business does not tokenise — say it takes only ad-hoc payments, an online store or a business issuing invoices — then merchant routing on its own is very much worth having. You can route Visa and Mastercard dual-network debit to eftpos without tokenising the card at all.

What it cost to implement

John was keen to move. His KPIs were tied partly to net profit and gross margin, and payments sat inside both. He put me in touch with his billing platform, who quoted $20,000 to implement. He estimated 40 hours of his own staff’s time to update invoices and the website, plus directors completing KYC and providing information for a credit check.

The obvious objection: what if the better providers cost more?

Fair question. The providers who can actually offer these solutions are sometimes more expensive. Innovative companies often are. But they can also save you more — so test it rather than assuming.

Say the payment gateway charges $0.02 more per transaction and the acquirer takes 0.05% more:

Even paying more for the gateway and giving up five basis points of acquiring margin, it is still a 156% return in year one and $43,608 a year after that.

That is the point worth taking away. A business that negotiates hard on bank margin and ignores interchange and scheme fees can end up rejecting the arrangement that would have saved it more.

John was thrilled, took me to a no-expense-spared lunch, and promised to invite me to the Christmas party.

FAQ

Does least cost routing always mean routing everything to eftpos?

No, and this case study is the clearest illustration of why. Routing all of Best NRG’s Visa and Mastercard debit to eftpos saved $62,528, while routing only Mastercard debit saved $77,774. Visa’s tokenised recurring interchange was low enough that routing away from it cost money. You have to know which interchange category each transaction lands in before you decide.

Why isn’t interchange the whole picture?

Because scheme fees sit alongside it. A category with lower interchange can carry scheme fees that make the total worse. Looking at interchange in isolation is the single most common error I see, including inside very large companies.

What does scheme tokenisation actually do for a recurring biller?

It moves the stored token from the gateway to Visa or Mastercard, which shifts transactions into cheaper interchange categories, lifts approval rates because the token survives a card reissue, and reduces the PCI scope the business carries.

Is a $28,000 implementation cost typical?

It depends entirely on the billing platform. The $20,000 quote came from Best NRG’s own provider. The staff time — around 40 hours for invoice and website changes plus director KYC — is fairly representative for a business of that size.

Would these numbers still hold today?

The method holds. The rates do not: interchange caps change on 1 October 2026, so any current analysis has to be run against the new caps. What does not change is that the answer depends on your own card mix, and that nobody can tell you what it is worth without looking at your statement.