One key problem: tracking interchange rates for each network available for each card. That is hard enough for a retail IT shop to do now, before the Fed's rules kick in. And although your CFO will probably want you to minimize interchange fees for each card as it's processed, automating that won't be cheap—whether you're paying your processor to pick the best deal or doing it yourself—and may not be worth the trouble.
The new Fed rules, which were announced on June 30, cap debit interchange fees at 21 cents per transaction plus 0.05 percent of the transaction value. That will cut the average debit interchange charge in half starting in October, and it's what has gotten all the press. The exceptions—cards from small banks and credit unions with less than $10 billion in assets, and cards for government programs like welfare and unemployment compensation—aren't capped, and that will take a bite out of the savings CFOs are hoping for.
Then there's processing-network competition. Most debit cards currently have only two options for processing: one for signature transactions and the other for PIN transactions, and both are controlled by the company whose brand is on the card—Visa, MasterCard, Discover, etc. Retailers lobbied the Fed heavily to generate competition in payment networks, and the new rules require that, starting in April 2012, each card must offer at least two networks that are unaffiliated with each other—and it's the retailer that can choose which network to use.
That means a Visa-branded debit card could use Visa's network for signature and a non-Visa network for PIN. Or the card could offer Visa's network for both signature and PIN, but also an alternate non-Visa network for PIN. Or the extra non-Visa option could be for signature. In other words, for any particular debit card, there might be a choice of payment networks for a PIN transaction, but there's no guarantee that's the case. All that's guaranteed is that for some types of transactions, a non-Visa network is an option.
Where there is a choice, it just keeps getting more complicated.Where there is a choice, it just keeps getting more complicated. Card issuers and networks are explicitly barred by the Fed rules from doing anything to force retailers to choose one particular network or to punish retailers if they choose a competing network. However, networks are explicitly allowed to offer "payments or other incentives to encourage the merchant to route electronic debit-card transactions to the network for processing." Issuers can also specify a default network to use if a merchant or processor doesn't want to make a choice.
For retailers that want to minimize interchange fees, all that's left to do is keep track of every different card's combination of networks and what the current interchange terms are for them, and then choose the cheapest network for the transaction as soon as the card is swiped. Yeah, that should be easy.
Building a secure system to identify what category of card the transaction is using, based on the card number? Expensive and a whole new can of worms for PCI, because it will require handling the unencrypted card number. Paying for a service to keep the card-type database updated, so it will actually pick the cheapest network for each transaction? An ongoing expense, and it won't be cheap.
It probably makes more sense for that pick-the-network system to reside at your processor. Then you're dodging the potential PCI issues but still paying an extra fee for finding the cheapest way to process the card. You're also giving up any sort of fine-tuned control of the process. In addition, it means every transaction will take just a little longer, and jumping among networks could break (or at least complicate) some existing security measures such as tokenization.
That's without any guarantee there will be noticeable competition based on interchange fees. Card issuers aren't required to pick their networks for each card until April 1, 2012. And remember, they're the ones who pick the competing networks, and they're also the ones who get much of the interchange money. How likely will they be to pick networks that have deeply cut interchange rates?
So will trying to squeeze every last nickel out of interchange fees be worth it? It's likely to be costly and risky (when it comes to PCI), and the payoff is largely theoretical. Starting on the project early—before you know for sure whether there will be any advantage at all to being picky about payment networks—means your ROI calculation will be even more of a crapshoot than usual.
But good luck explaining that to your CFO.