Walmart announced yesterday that tap to pay arrives at select Walmart and Sam’s Club locations on August 24, with all US stores and clubs by the end of 2026 and fuel stations by mid 2027. The largest retailer in the country, and the last major holdout on contactless, is turning it on. What a great “going away” present to Jennifer Bailey!.
This is a bigger deal than the press release suggests. Walmart did not resist contactless because it disliked the technology. It resisted because contactless threatened a payment cost structure it spent twenty years building. The news here is not that Walmart added a feature. It is that Walmart concluded the feature no longer costs it anything.
For readers who want the plumbing, I laid out the EMVCo and tokenization dependencies in Understanding ApplePay in PIN Debit (May 2026). This post is the business version.
What I told Apple, repeatedly
Jennifer Bailey at Apple would always ask me the same question: “How can we get Mike [Cook] to accept ApplePay?” My answer never changed, I told her “It was not about Apple.” It was about three things:
- Funding mix. A tap is frictionless in a way that a PIN entry is not, and frictionless favors credit. If contactless shifts even a few points of Walmart’s basket from debit to credit, the interchange bill goes up. Nothing Apple could offer would offset that.
- Debit routing. Walmart had to work out how to keep routing a tokenized debit transaction on PIN rails. This is the hard one, and I believe store operations pushed it harder than the payments organization did. They have now resolved how to manage PIN debit inside ApplePay. That deserves its own post, and it is the reason this announcement will really juice ApplePay volume.
- Network tokenization fees and value added services. Every tokenized transaction creates a new billable event that did not exist on a swipe. While the tokenized debit can have a Common US Debit AID, the DPAN in the phone is still a VTS token (given that Visa is 90+ of debit I’m not mentioning MDES but it works the same).
Solve all three and Walmart says yes. Solve two and Walmart says no. It took until 2026.
Credit where it is owed
One piece of history that keeps getting lost. Jennifer Bailey and Apple deserve 100% of the credit for making debit available in contactless at all. In 2014 the banks wanted tap to pay to be a premium credit card feature. Debit portfolios were to be excluded. Apple’s issuer agreement required every participating issuer to enable its debit portfolios at parity with credit as a condition of taking part in ApplePay. I was in those conversations and I am the first hand source for this.
Without that mandate there is no debit in the wallet, and therefore nothing for Walmart to route today. I made a related argument at the time in ApplePay and Merchants.
Walmart and debit, the long war
Walmart’s history with debit is legislative, regulatory and litigious in roughly equal measure. Walmart built point of sale infrastructure that forced PIN on effectively 100% of debit transactions and disabled signature debit on EMV cards through custom terminal modifications, which I covered in PIN Debit at the POS. Very few merchants had the engineering capability to do that.
That fight is over, and Walmart won it in a way most people missed. Durbin enabled Pinless PIN, so Walmart no longer needs to force the PIN prompt at all. It assumes the risk itself and processes debit far below Durbin rates, at roughly $0.08 to $0.10 per transaction (my estimate). This is far below the Durbin rate of $0.21 + 5bps.
Put that in context. Autonomous estimates regulated debit interchange runs about $0.24 per transaction, against ~$0.50 for unregulated debit (Kenneth Suchoski, Autonomous). Walmart is paying a third of the regulated rate. The PIN prompt was never the objective. The routing right was the objective, and Walmart kept the right while discarding the prompt.
How the PIN actually gets handled in a tap (tech stuff)
Here is the mechanic that makes this work, and it is worth understanding because it is not obvious. When a Visa branded debit card is tapped through Apple Pay, the terminal sees two application identifiers: the Global AID (Visa Debit) and the Common US Debit AID. If Walmart’s point of sale is configured to prefer the Common Debit AID, it can pull the Visa token transaction off Visa rails and route it to STAR, NYCE, Accel or another network. That configuration sits entirely in Walmart’s control. Apple Pay does not change it. Thus Durbin routing rights survive the tap.
But it is not a clean bypass, because Visa is still in the chain. The device token (the DPAN) is a Visa issued token living in Visa’s vault. No processor can detokenize without the network, whether that is Visa or Mastercard’s MDES. So STAR passes the transaction back through a shared gateway to Visa for detokenization, Visa hands the funding PAN back to STAR, and STAR routes to the issuing bank. Visa collects a technical processing fee for the token lookup even when STAR earns the interchange.
There is a second “possible” path. Walmart could recognize the token (or user or phone) and substitute its own debit credential from card on file. Less likely, but probably still permitted.
The practical effect is that Walmart keeps its routing economics, Visa keeps a toll, and the merchant loses the ancillary wallet metadata that comes with the tokenized transaction once it has been detokenized. I worked through the earlier version of this problem in Debit, Routing, Tokens and Liability Shift, where the conclusion was that tokenization structurally advantages Visa in US debit. That conclusion holds, but Walmart has found the seam in it.
My open question
While detokenizing is complex, my top question is whether Walmart can map the token, or other tap data such as the wallet identifier, back to its own cards on file. If it can, it makes the whole thing more efficient on Walmart’s side and avoids paying the network for value added services it does not need. Walmart has tens of millions of cards on file and no shortage of engineering talent. I would not bet against them trying.
Net net, who pays and who gains
Walmart. No incremental cost, assuming that consumer payment mix stays where it is (no shift toward credit). What Walmart gets is a better customer experience and faster checkout, which at Walmart’s transaction count is worth real money on labor and throughput alone.
Issuers. Potential upside on credit volume if behavior does shift, and no exposure on the debit side, since they bear no cost when Walmart routes debit over a PIN network.
Visa. Some gain on tokenization and value added service fees, though realistically that gets negotiated inside a larger package of services rather than billed at rack rate to a merchant of this size. Visa’s real win is that it stays in the transaction chain even on a route designed to avoid it.
The risk case. All of Walmart’s arithmetic depends on the funding mix holding. Tap is easier than PIN, and easier has historically meant credit. If the mix moves, the calculation I laid out for Jennifer Bailey a decade ago reasserts itself and Walmart’s cost line goes up.
Merchants pay for everything in payments, which is why Walmart spent ten years refusing to do the obvious thing until the mechanics worked in its favor. I made the same point about the industry’s incentive structure in Processors: Understanding Competitive Dynamics. Walmart did not adopt contactless because it changed its mind. It adopted contactless because it finally solved the routing problem, and now the tap is free.