I've written in earlier blog posts about how I don't think that credit card processing is a particularly good business (Credit Card Processing is a Hard Business and Credit Card Processing as a Commodity Business).
The gist of my argument is that a company in the card processing business has to reach truly massive scale before starting to profit. That is very hard to do. Square, Braintree, Stripe, and WePay prove that one can build valuable businesses but they have all achieved massive scale. Braintree after all was purchased mostly for their consumer product and its network while Stripe is working hard to build out Stripe Checkout so it can store consumer card data and do the same.
I was chatting with Jareau Wade, one of the co-founders of Balanced Payments, to get some idea of comparables on the scale. He pointed out that Vantiv processed roughly $530B last year and currently has a market cap of $5.7B. Heartland Payments processed roughly $120B last year and currently has a $1.6B market cap. At the time of Braintree's acquisition it had a processing run rate of $12B a year and was worth $0.8B. PayPal did not perceive Braintree's value as coming from their processing business, no matter how fast it was growing.
My comments on card processing led me to a thought experiment: maybe a processor could give card processing away and make money some other, higher margin way.
No independent sales organizations (ISO) - such as Stripe or WePay - could make the marginal cost of a charge zero because a huge percentage of the fee they collect goes to the card networks and their member banks (often called, not quite accurately, the interchange fee). In fact, big businesses can often get a card processing fee structure of interchange + some margin (known as "interchange plus" pricing). Maybe these processing business could charge everyone that way and find some other way to make most of their money. They can't quite make it zero margin because the processor does take on some risk but they could run this part of their business at break even.
A business that comes immediately to mind that could do this is Sift Science. They could just give processing away on top of their already profitable product. Sift Science is proving that a ton of businesses will pay them for better credit card fraud protection. They've cracked that nut but their system get better and better as they see more transaction. More data improves their algorithms and their blacklists. One way they could see a lot more transactions is if they just gave the transactional service away for free.
I'm not actually advocating that Sift Science get in to the processing business. They're running a great business and likely don't need a different path to user acquisition. It would certainly burden them to become an ISO with someone like Wells Fargo or Paymenttech. I am using them as an example though of how one could build a business that gives processing away while helping build a profitable core business.
As a contrast take Square, for example. They keep releasing (not particularly successful) products to try to get at bigger margins: Cash, Wallet, Register etc. Rather than running a processing business and then trying to find one's high-margin product it would be better to figure out a high-margin product where dirt cheap processing would be a way to get a lot of people to use it.