The most consequential shift in consumer payments this year is not happening online, where buy now, pay later already won, but at the physical checkout counter, where it has so far been a footnote. The pattern of the last 30 days suggests that gap is about to close fast. Three independent signals, a new terminal-network deal, an issuer’s in-store card data, and the wiring of pay-later into mobile wallets, point to the same outcome: in-store BNPL is likely to reach genuine mainstream scale at major US retailers before the 2026 holiday season, with the in-store mix becoming the primary growth vector for the largest providers by the fourth quarter. This piece lays out why that prediction is grounded rather than hopeful, and where it could still be wrong.
In short
- Prediction: in-store, physical-checkout BNPL likely crosses into mainstream availability at major US retailers before holiday 2026 (Black Friday and Cyber Monday), with in-store volume becoming the leading growth contributor for Affirm and Klarna by the December quarter.
- Signal 1: the Klarna and Worldline partnership announced on May 20 puts Klarna pay-later directly onto a global in-store terminal network, in the exact card-dominated markets where physical retail volume sits.
- Signal 2: Affirm’s most recent quarter, reported in May, shows the Affirm Card and its in-store usage growing several times faster than the legacy e-commerce checkout button, with management describing in-store Card spend as an order of magnitude above non-Card surfaces.
- Signal 3: the wiring of Affirm and Klarna into Google’s surfaces and Google Pay (announced May 12), alongside deeper Apple Pay ties, bridges the online button and the tap-to-pay terminal through the wallets consumers already carry.
- The caveat: regulatory reclassification, a softening US consumer, thin physical-checkout economics, and a decade of over-promised in-store BNPL all argue the timeline could slip past the holidays into 2027.
Why this matters now
BNPL grew up online. The format was built for the friction point of an e-commerce cart, where a shopper hesitates over a $180 pair of trainers and a “4 payments of $45” button converts the sale. For roughly six years that is where the volume lived, and the physical store remained the domain of cards, cash, and the occasional store-branded installment plan.
The structural problem is that the store is still where most US retail spend happens. Even after a decade of e-commerce growth, the large majority of US retail sales clear at a physical point of sale. A payment method that only works online is, by definition, capped at a minority of the addressable market. The providers know this, and the last month shows them moving on it from three different directions at once.
The timing question is not whether in-store BNPL happens, but when it stops being a rounding error. The signals below suggest the inflection is being engineered now, deliberately, to land before the highest-volume shopping window of the year. That is the difference between a press-release ambition and a falsifiable near-term call, and it is why the holiday quarter is the right clock to watch. Our earlier analysis of how BNPL is becoming a card network rather than a checkout button framed the rails; this is the next leg of the same trend, moving from the browser into the terminal.
Signal 1: Klarna puts pay-later onto a global terminal network
On May 20, Klarna and the payments acquirer Worldline announced a partnership that extends Klarna’s services from e-commerce into physical retail, making pay-later available at both online and in-store points of sale where Worldline operates. The phrasing in the announcement is precise: the deal targets “markets where traditional card-based payments have been dominant.” That is the tell.
Worldline is one of Europe’s largest merchant acquirers and terminal operators, with a footprint across hundreds of thousands of physical merchants. Plugging Klarna into that estate does something the company could not achieve merchant-by-merchant: it converts in-store BNPL from a bespoke integration into a feature an acquirer can switch on. The economics of distribution change when the rail already terminates at the checkout counter.
The backdrop matters too. Klarna reported first-quarter 2026 revenue of roughly $1.0 billion, up from around $701 million a year earlier, while trimming losses to near break-even, according to its quarterly disclosures. A provider scaling revenue while controlling credit costs has the balance-sheet room to pursue distribution rather than defend margin. The Worldline deal reads as a distribution move, not a survival one.
| Attribute | Online BNPL (2019-2024) | In-store BNPL (the Worldline model) |
|---|---|---|
| Integration unit | Per-merchant plugin or SDK | Acquirer-level terminal switch |
| Merchant onboarding | Months, developer effort | Configuration, no rebuild |
| Addressable spend | E-commerce checkout | Physical point of sale |
| Consumer surface | Web and app cart | Card terminal, QR, wallet |
The significance is not the single deal; it is the model. When a major acquirer can enable pay-later as a terminal feature, the rollout curve stops looking like software sales and starts looking like network activation. That is the kind of step-change that precedes a visible jump in availability within a couple of quarters.
Signal 2: Affirm’s in-store card data is already bending up
The second signal is not a forward-looking promise but a backward-looking number. In its most recent quarter, reported in May, Affirm disclosed gross merchandise volume growth of roughly 36% to around $13.8 billion, with the standout being its physical-world product. The Affirm Card drove direct-to-consumer volume up sharply, card-specific volume grew well over 100% year on year, and active cardholders more than doubled to roughly 4.4 million, with the card attach rate near 17%, per the company’s shareholder materials (see Affirm’s investor relations page for the underlying disclosures).
The qualitative detail is the more telling part. Management characterized in-store spend on the Affirm Card as running an order of magnitude higher than in-store spend on Affirm’s non-Card surfaces. In plain terms, the card is the mechanism that unlocks the physical store, and it is precisely the card that is compounding fastest. On a June 4 investor call, Affirm reiterated a path toward doubling the business to roughly $100 billion in GMV, with card growth positioned as the engine.
This is the cleanest evidence that the shift is real rather than aspirational. A checkout button cannot ride along in a physical wallet; a card can. Once a BNPL provider issues a card that lives in Apple Pay or Google Pay and taps at any terminal, the distinction between online and in-store collapses at the moment of payment.
The Walmart episode underlines the durability of the in-store appetite. When Klarna displaced Affirm as the BNPL provider inside Walmart’s OnePay app in 2025, a notable share of affected shoppers simply obtained Affirm Cards to keep pay-later access at the register, online and offline, according to Affirm management commentary in May. Demand for in-store installments is sticky enough to survive a provider switch, which is not how a fad behaves.
Signal 3: the wallets are doing the bridging
The third signal is the connective tissue. On May 12, Google announced that Affirm and Klarna would be available across Google Search, its AI Mode, the Gemini app, and Google Pay checkout. Around the same window, both providers deepened their integrations with Apple Pay. Read in isolation, these look like online distribution wins. Read alongside the first two signals, they are the bridge.
Google Pay and Apple Pay are not only online checkout tools; they are the tap-to-pay credentials hundreds of millions of consumers already use at physical terminals. Once a pay-later option is provisioned into those wallets, the same instrument that funds an online order can fund an in-store tap. The wallet is the place where the online button and the physical terminal converge, and that convergence is now being wired deliberately.
The agentic-commerce framing of the Google deal is a further tell about intent. BNPL is being positioned as a native funding source for AI-mediated purchases, which only deepens its integration into the core payment stack rather than leaving it as a merchant add-on. Our analysis of why card-network rails, not closed-loop checkout, will win agentic commerce traces the same logic from a different angle, and the two trends reinforce each other.
| Signal | Date | What it proves | Independent vector |
|---|---|---|---|
| Klarna and Worldline | May 20, 2026 | Acquirer-level in-store rails exist | Terminal and acquiring infrastructure |
| Affirm Card results | Reported May 2026 | In-store demand already compounding | Issuer and consumer behavior |
| Google and Apple wallet ties | May 12, 2026 | Online and tap-to-pay converge | Wallet distribution |
What the pattern suggests
Three signals from three different layers of the payment stack, infrastructure, issuance, and distribution, are rarely a coincidence. When an acquirer, an issuer, and the dominant mobile wallets all move toward the same surface inside a single month, the more likely reading is a coordinated industry push timed to a known commercial deadline.
That deadline is the holidays. The fourth quarter is where BNPL providers earn a disproportionate share of annual volume, because higher-ticket gift purchases are exactly where installments convert. Building the in-store rails in May and June, ahead of merchant onboarding cycles that typically run through late summer, is consistent with wanting them live and stable before the November peak.
The pattern also fits a precedent that has repeated reliably in payments: what ships online first migrates into the physical store once the wallet rail matures. Contactless and tokenized mobile payments themselves followed this arc, starting as an e-commerce and novelty behavior before becoming a default in-store tender. The prior precedent points to BNPL following the same path, and the wallet integrations are the specific mechanism that makes the migration mechanical rather than promotional.
None of this guarantees ubiquity by December. The signals point to availability and momentum, not saturation. The reasonable base case is that by the holiday quarter, in-store pay-later moves from “available at a handful of marquee merchants” to “routinely offered at major US retailers and visibly the fastest-growing slice of provider volume,” with full saturation a 2027 story.
It is worth being explicit about why three layers moving together is more informative than any single announcement. A terminal deal alone could be a partnership that never ships volume. An issuer’s strong card quarter alone could be a one-off cohort effect. A wallet integration alone could be a marketing tie-up. The reason the convergence carries weight is that each layer removes a different bottleneck: Worldline removes the merchant-integration bottleneck, the Affirm Card data removes the question of whether consumers actually want to pay in installments in a store, and the wallets remove the friction of provisioning credentials at the terminal. Solve all three in one month and the remaining obstacle is commercial will, not capability.
Prior precedents: how new tender types cross into the store
The strongest reason to take the timeline seriously is that the payments industry has run this play before. New tender types almost always prove themselves online, where the integration is software and the risk is contained, and then migrate into the physical store once a wallet or a card gives them a body. The migration is rarely linear, but the sequence is consistent.
Contactless and mobile wallets are the clearest example. Tap-to-pay began as a novelty and an e-commerce convenience before it became, over several years, a default in-store tender that now clears a large share of card-present volume. The enabling step was not consumer education; it was the moment the credential lived in a phone that worked at any terminal. BNPL is now reaching that same enabling step through Apple Pay and Google Pay.
Store-branded installment plans are the counter-example worth remembering. Retailer-specific financing has existed at the physical register for decades, but it stayed siloed because it was tied to one merchant and one application flow. The new model inverts that: the installment relationship belongs to the consumer and the provider, and travels across merchants. That portability is what turns a niche financing option into general-purpose tender.
| Tender type | Online-first phase | What enabled the in-store crossover | Outcome in the store |
|---|---|---|---|
| Mobile wallet (tap-to-pay) | App and e-commerce novelty | Credential provisioned into the phone | Default card-present tender over time |
| Store-branded installments | Tied to one merchant flow | Never made portable | Stayed siloed and niche |
| BNPL (the current move) | E-commerce checkout button | Provider card plus acquirer-enabled terminals | Prediction: routine in-store tender by holiday 2026 |
The precedent does not promise the speed; it explains the mechanism. Each time the credential became portable and the integration became a switch rather than a project, in-store adoption followed within a small number of quarters. The signals of the last month are precisely those two conditions being satisfied for BNPL at the same time.
Wider context: why physical retail is the prize
The strategic logic behind the in-store push is straightforward. Online BNPL, for all its growth, addresses the smaller share of total retail spend. The providers have largely captured the e-commerce opportunity and now face decelerating online growth rates off a large base. The next leg of compounding has to come from the physical store, or it does not come at all.
In-store also changes the product’s character in a way that favors the providers. A card that lives in a wallet and works everywhere shifts BNPL from a per-transaction merchant feature to a consumer’s default spending method, which is a far stickier and higher-frequency relationship. That is the same logic we explored in our piece on BNPL becoming a card network: the endgame is to be tender, not a button.
There is a competitive dimension as well. The incumbents Klarna, Affirm, and the Apple and PayPal pay-later products are racing to own the physical surface before any one of them establishes default status. Distribution land grabs of this kind tend to compress into short, intense windows, which is another reason the activity is clustering now rather than spreading evenly across the year. The capital-markets backdrop reinforces the urgency; as we noted in our look at the reopening fintech and commerce IPO window, providers eyeing public-market milestones have every incentive to show in-store traction in their numbers.
For retailers, the holiday calendar adds its own pressure. With the US discount season already stretching earlier, as we documented in the analysis of the summer sales peak moving to June, merchants are looking for any conversion lever that lifts basket size across a longer promotional runway. In-store BNPL is a credible one.
Implications for retailers, providers, and investors
For retailers, the practical question over the next two quarters is whether to enable pay-later at the physical register and on what terms. The acquirer-led model lowers the integration bar, but the merchant-fee question remains live: physical-checkout economics are thinner than online, and a BNPL fee that is tolerable on a discretionary e-commerce order may be harder to justify on an everyday in-store purchase. Expect larger retailers, especially in discretionary categories such as electronics, furniture, and apparel, to move first.
For the providers, the prize is frequency. An in-store card holder transacts far more often than an online-only BNPL user, which improves data, underwriting, and lifetime value. The risk is symmetrical: extending credit to everyday in-store spend pushes BNPL closer to general-purpose consumer credit, with the loss curves and regulatory attention that come with it.
There is also a defensive logic at work for the providers. If BNPL stays online-only, it remains a feature that card networks, banks, and wallet operators can replicate or absorb. Owning a physical-world card that consumers reach for at the register is what turns a checkout option into a durable banking relationship, which is why the in-store push is as much about competitive moats as it is about incremental volume. The provider that establishes default in-store status first is likely to be hardest to dislodge.
For merchants weighing the decision, the calculus is not only fees but conversion and basket size. The evidence from online BNPL is that installments lift average order value and reduce abandonment, particularly on higher-ticket discretionary goods. Whether that lift survives at the physical register, where the shopper has already committed to entering the store, is the open empirical question that the holiday quarter will begin to answer.
For investors, the signal to track is mix, not headline GMV. The thesis is confirmed if Affirm and Klarna report in-store or card volume as the leading growth contributor in their holiday-quarter results, and if Klarna-on-Worldline terminals go live in the US. It is weakened if in-store stays a single-digit share of volume into 2027. The crypto-and-wallet adjacency is worth watching too; our note on payments incumbents launching dollar stablecoins describes the same wallets that BNPL is now riding into the store.
| Scenario | What we would observe by Q4 2026 | Read on the prediction |
|---|---|---|
| Base case (likely) | In-store pay-later live at major US retailers; card and in-store volume the fastest-growing slice | Confirmed |
| Acceleration | Acquirer-enabled rollouts plus wallet defaults; in-store a double-digit share of new volume | Confirmed, ahead of schedule |
| Stall | Pilots only; merchant-fee or regulatory friction caps in-store at low single digits | Prediction slips to 2027 |
Caveats: what could go wrong
The prediction is falsifiable, and several forces could push the timeline past the holidays. The first is regulatory. US consumer-credit regulators have moved to treat certain BNPL products more like credit cards, with the disclosure and dispute obligations that implies. A tighter classification, or fresh enforcement attention, could make providers slow in-store rollouts precisely as everyday-spend volume grows, because everyday spend is where the regulatory and reputational risk concentrates.
The second is the consumer credit cycle. A softening US consumer with rising delinquencies would push providers to tighten underwriting, and in-store everyday spend is a more marginal credit population than the discretionary online shopper. If loss curves deteriorate, the rational move is to slow the in-store land grab, not accelerate it.
The third is merchant economics. Physical-checkout interchange and acceptance costs are thin, and merchants have historically resisted layering additional fees onto in-store transactions. If the BNPL take rate cannot be reconciled with physical-checkout margins, adoption could stall at the largest discretionary retailers and never reach the everyday register.
The fourth is history itself. In-store BNPL has been promised before. Provider cards launched years ago, and physical adoption lagged the online business for longer than the early roadmaps implied. The order-of-magnitude growth in Affirm’s in-store Card volume is impressive in percentage terms but is still compounding off a comparatively small base. A skeptic can reasonably argue the inflection is real but slower than a holiday-2026 timeline assumes, landing instead in 2027. That counter-case is the one to respect.
The honest synthesis: the direction is well supported by the signals; the speed is the genuine uncertainty. The most likely outcome is meaningful mainstream availability by the holidays with full saturation later, but a regulatory shock or a credit downturn could convert that into a 2027 story.
FAQ
What exactly is the prediction, and how is it falsifiable?
The prediction is that in-store, physical-checkout BNPL reaches mainstream availability at major US retailers before the 2026 holiday season, with in-store or card volume becoming the leading growth contributor for Affirm and Klarna by the December quarter. It is falsifiable: a future observer can check holiday-quarter disclosures, whether Klarna-on-Worldline terminals are live in the US, and whether pay-later appears as a tap-to-pay option in major wallets at physical terminals.
Why focus on in-store when BNPL is already huge online?
Because most US retail spend still clears at a physical point of sale, and online growth is decelerating off a large base. The next leg of provider growth has to come from the store, which is why the recent moves all target the physical surface rather than the browser.
How is this different from BNPL that already exists at some stores?
Earlier in-store BNPL relied on per-merchant integrations and was rare. The new model, an acquirer enabling pay-later as a terminal feature plus provider cards living in mobile wallets, converts availability from a bespoke project into a network switch, which is what changes the adoption curve.
Are these three signals really independent?
Yes. They come from different layers of the stack: terminal and acquiring infrastructure (Worldline), issuance and consumer behavior (Affirm Card data), and wallet distribution (Google and Apple). They are not three reports of one event, which is what makes the convergence meaningful.
Could regulation stop this?
It could slow it. US regulators have moved toward treating some BNPL products like credit cards. Tighter rules or enforcement would raise the cost of extending pay-later to everyday in-store spend, which is exactly where the regulatory risk concentrates, and could push the timeline into 2027.
Which retailers are likely to adopt first?
Larger merchants in discretionary, higher-ticket categories such as electronics, furniture, and apparel, where installments most clearly lift conversion and basket size, and where the BNPL fee is easiest to justify against margin.
What would prove the prediction wrong?
If in-store or card volume remains a low single-digit share of provider volume into 2027, if Klarna-on-Worldline terminals do not go live in the US, or if a credit downturn or regulatory shock causes providers to pull back from everyday in-store lending.
What is the single most important metric to watch?
Volume mix in the holiday-quarter results. If Affirm and Klarna report in-store or card volume as the fastest-growing contributor, the thesis is confirmed; if it stays marginal, the call slips.
How does this connect to agentic commerce and wallets more broadly?
The same wallet rails that bridge online and in-store BNPL are also the rails being wired for AI-mediated and tokenized payments. BNPL riding into the store on Apple Pay and Google Pay is one expression of a wider move to make pay-later a native part of the core payment stack rather than a merchant add-on.