Costs and pricing

How the money moves

Five streams decide whether a card program makes money: interchange, the company's rebate, funding the balance until the company pays, losses, and processing receipts. A sixth sits off the books: the float employees lend when they pay out of pocket.

These are rough rates from Visa's US and Canadian tables as of October 2026. Merchants also pay network fees and acquirer markup. The debit cap is why fintech prepaid programs favor banks under $10B.

Tax is money too. A $10,000 car allowance under a nonaccountable plan carries 22% withholding plus 7.65% employee and 7.65% employer FICA. In Canada, GST/HST recovery on a reasonable allowance is worth 4.8% (5/105) to 13% (15/115) of it, plus about 9% QST in Quebec.

Who gets the $27 on a $1,000 purchase

A fintech-run Visa Corporate charge card, online, monthly statement. It's illustrative: filings and analysts disagree tenfold on the bank, network and processor shares.

With validated Level 3 data, interchange drops to about $18, the pool loses about $9.50, and what's left goes negative unless rewards fall. With a 3% merchant surcharge, the company pays $1,030 to get about $13 back. As a cross-check, the largest closed-loop issuer's commercial business earned about 0.68% of billed business before tax in 2025.

How companies of different sizes use it

Rebates, settlement terms and underwriting all scale with card spend, but nobody publishes card spend by company size. So the steps below go by headcount, with my rough estimate of card spend for each, worked out from two public platforms' average customer. Treat it loosely: a 300-person software company and a 300-person field-services firm can differ tenfold. From a few million dollars of card spend up, the rebate (about 1-1.5% of spend, rising with size and payment speed) is worth more than the software, so the decisions that matter move from the price list to the card contract.

A startup (under 50 people, and by my rough estimate under $1 million a year on cards) spends mostly on software, ads, cloud and a little travel. It usually takes a fintech charge card with free software (Ramp, Brex, BILL's spend product, or Float in Canada), underwritten on its bank balance with no personal guarantee, so the limit falls when the cash does. Others use a bank or American Express small-business card, often revolving and personally guaranteed, and Intuit's new card goes after the same QuickBooks customers. Founders still mix in personal cards: in the Fed's 2026 small-business survey, almost half of the smallest firms that use credit cards put business on a personal one. Moving that onto company cards pays: out of pocket, employees lend the company money interest-free, keep the rewards, and someone has to check and pay every claim, which the comparison table above prices at tens of dollars a report. On my arithmetic the card wins for the company even if that old benchmark is several times too high. The founder or CEO owns it, often with an outside bookkeeper. More hires, an audit that wants clean books, a move to NetSuite or a second entity push it to the next step.

A scale-up (about 50 to 500 people, roughly $1-10 million a year on cards by my estimate) gives cards to most managers and travelers, writes an expense policy, and gets its first ERP and often a second entity. That's where per-user fees start: Ramp and Brex both keep richer ERP integrations and multi-entity support behind paid tiers priced per user, and Float has a similar paid tier in Canada. Banks only recently built cards for this size: U.S. Bank launched one for the "emerging middle market" in 2023. The controller or head of finance owns the program, and an office or people-ops admin hands out the cards. Settlement terms start to matter here. At the top of this band, a company paying its card monthly leaves the platform close to break-even on my arithmetic, while paying daily is worth about as much to the platform as a half-point rise in interchange; that's why rebate tiers reward fast payment, and it's the term I'd negotiate first. More entities and currencies, a travel program and IPO preparation push it on.

A mid-market company (about 500 to 2,000 people, roughly $10-50 million a year on cards by my estimate) runs a travel program with a travel management company or booking tool, purchasing cards for supplies, and a month-end close that can't wait on receipts. It buys a spend platform's enterprise tier or a travel and expense suite such as SAP Concur or Navan, and the card often comes from a bank (U.S. Bank, JPMorgan Chase or American Express) or from a fintech at scale. My guess is that the card now becomes part of the bank relationship, sold alongside the credit line. The controller and AP manager run the card program, and a travel manager, often in procurement or HR, owns travel. Software now costs real money (my rough estimate is in the table below), and the rebate is still worth more, even though mid-size buyers don't get the rates large buyers do. Going public, acquisitions that bring other card programs, and countries that need local-currency cards push it on.

An enterprise (about 2,000 to 10,000 people, roughly $50-500 million a year on cards by my estimate) runs separate programs: corporate cards for travel, often on individual liability in bank programs, purchasing cards for low-value supplies, a central travel account at the agency, and virtual cards for suppliers (those are in procure to pay). The cards come from bank programs (JPMorgan Chase, U.S. Bank, American Express Corporate) chosen by RFP, and expenses run on SAP Concur or Emburse, tied to the ERP; SAP says Concur holds about half the market. Ramp and Brex now sell here too, with enterprise tiers that add Workday and Oracle integrations and local cards in dozens of countries. Treasury or AP owns the card program, procurement owns travel and HR feeds the employee list. The rebate is worth millions a year, far more than the software, so the real decisions are the rebate terms, validated Level 3 data (which lowers interchange and with it the rebate) and the liability model. Global consolidation after acquisitions pushes it on.

A Fortune 500 company (the largest US companies by revenue, and by my rough estimate $500 million or more a year on cards) runs card programs in many countries, a global travel agency and several expense systems inherited from acquisitions. The bank and American Express programs are re-tendered by RFP every few years, often with a consultant; I couldn't find a reliable share for any issuer or expense vendor at this size. The treasurer or global head of procurement owns the card contract, and a global travel director owns travel. The rebate dominates the economics, priced by type of spend (large tickets, Level 3 data, lower rates at the very largest merchants) and by payment speed. The nearest public benchmark I found is the US federal card program, which got back about 1.2% of its spend in refunds in 2025. There's no next size up; the pressure comes from the board, auditors and acquisitions, and the Credit Card Competition Act, if it ever passes, would hit the big banks that serve this size.

What changes as a company grows (spend is my rough estimate):