How does the buy-now-pay-later (BNPL) industry and business model work in 2027?

What BNPL actually is and why the model matters

Buy-now-pay-later is a short-term, point-of-sale installment product embedded directly into a merchant's checkout. The shopper picks "pay in 4" or a longer monthly plan, passes a soft underwriting check that takes under a second, and walks away with the goods. The merchant is paid in full almost immediately — typically within one to three business days, net of the provider's fee. The provider now owns the receivable and the risk of the shopper not paying.

That last sentence is the whole business. BNPL is not a payment rail that passes money through; it is a lender that acquires customers at the moment of purchase intent and gets a retailer to subsidize the acquisition. Understanding the industry means holding three facts at once: the merchant funds the majority of revenue, the consumer funds a meaningful minority, and credit losses sit between the two as the variable that decides whether either stream produces profit.

The scale is real. Global BNPL gross merchandise volume reached roughly $560 billion in 2025, growing about 13.7 percent that year — a growth rate that marks the category's shift out of hypergrowth into something closer to a maturing payments vertical. Provider revenue across the industry ran near $44.9 billion in 2025 with projections toward roughly $54.6 billion in 2026. Those are revenue figures for the providers themselves, not the volume they move; the gap between $560 billion in GMV and $45 billion in revenue is the effective take rate of the whole industry, landing around 8 percent when you blend merchant fees, interest income, and consumer charges together.

The public leaders show the model at different points on the maturity curve. Klarna posted about $2.81 billion in revenue on roughly $105 billion in GMV — a take rate near 2.7 percent, reflecting its heavy weighting toward interest-free pay-in-4 where merchant fees do nearly all the work. Affirm crossed into profitability in fiscal 2025 with roughly $52 million in net profit on revenue of about $3.3 billion, a milestone that took the company years of growth-first operating to reach. Afterpay, now inside Block, processes billions in US volume under a strict zero-interest, late-fee-only structure. Three companies, three different answers to the question of who pays and how much.

Why any of this matters outside fintech: BNPL is the cleanest large-scale example of a business model where the party that captures the value funds the service. The shopper gets convenience they mostly do not pay for. The retailer gets incremental revenue and pays for it out of gross margin. Any operator designing pricing for a service that demonstrably lifts a partner's revenue is looking at the same structural question BNPL answered at $560 billion of scale.

The step-by-step process from checkout click to collected cash

Walk a single $200 order through the machine and every revenue and risk line becomes visible.

Step one — presentment. The BNPL option appears on the product page, in the cart, and at checkout. Placement is not cosmetic; providers push merchants to surface the installment price ("4 payments of $50") on the product detail page because the message has to land before the shopper decides whether the item is affordable. Merchants that only show BNPL at final checkout capture a fraction of the lift they paid for.

Step two — underwriting. The shopper enters name, date of birth, phone, and sometimes the last four digits of a government ID. The provider runs a soft credit inquiry that does not affect the shopper's score, then layers on its own signals: device fingerprint, email age, prior repayment history on the network, shipping-to-billing address match, order value relative to the shopper's historical basket. The decision returns in well under a second because a checkout that stalls kills the conversion the merchant is paying for. First-time users are typically capped at small limits — often in the $50 to $500 range — and earn higher limits through clean repayment.

Step three — settlement. On approval, the shopper pays the first installment ($50 of the $200) immediately. The provider advances the merchant the full $200 minus its fee. At a 5 percent merchant discount rate the retailer nets $190, usually within one to three business days.

Step four — the provider's exposure window. The provider has now paid out $190 and collected $50. It is $140 in the hole on this order and will remain so for six weeks while the remaining three installments come due every two weeks. Multiply that by millions of orders and you get the working-capital problem at the heart of the industry: BNPL companies must fund a continuously revolving receivables book through warehouse credit lines, forward-flow agreements with banks, or securitization. The cost of that funding is a direct input to unit economics and rises with interest rates.

Step five — collection. Auto-debit pulls each installment from the shopper's linked card or bank account. Most shoppers pay without incident. When a pull fails, the provider retries on a schedule, sends notifications, and after a grace period — commonly 7 to 10 days — applies a late fee where its model uses one, and often freezes the account against new purchases. Persistent nonpayment moves to hardship handling, then to collections or charge-off.

Step six — outcome. Roughly 98 percent of the time the provider collects the remaining $150 and books its $10 merchant fee as gross revenue, less funding cost and servicing cost. The other roughly 2 percent charges off, and the provider eats the outstanding principal.

The longer-tenor products change the arithmetic but not the shape. A 12-month plan on a $1,500 mattress puts the provider out of pocket far longer, so it either charges the shopper interest — Affirm's APRs run up to 36 percent on interest-bearing plans — or charges the merchant a much larger upfront discount to buy down the rate to zero. That second structure, the merchant-subsidized 0 percent APR offer, is common in furniture, fitness equipment, and elective healthcare, where the retailer's margin can absorb it.

Costs, fee structures, and the ranges that actually appear in contracts

Merchant pricing is the single most misunderstood part of the industry, because the headline "2 to 6 percent" hides enormous variance driven by tenor, vertical, and volume.

Short-tenor pay-in-4 sits at the low end of the range for large merchants and the high end for small ones. A national retailer moving nine figures of BNPL volume negotiates rates well below what a Shopify store on a standard plan pays off the rate card. The fee is quoted as a percentage plus a fixed per-transaction component — the fixed piece matters disproportionately on small baskets, which is why providers set minimum order values and why BNPL underperforms in low-ticket categories.

Longer-tenor installment financing costs the merchant substantially more, often in the mid-to-high single digits, because the merchant is effectively buying down the consumer's interest rate. The economics are straightforward: the provider needs a certain yield on a 12-month loan; if the consumer pays 0 percent, the merchant funds the difference upfront as a larger discount on the loan sale.

Vertical drives everything. A luxury goods retailer with 50 to 60 percent gross margins can hand over 5 or 6 points and still come out ahead if the offer lifts average order value. A grocery or essentials merchant working on 5 to 15 percent margins cannot — which is why BNPL penetration in those categories relies on capped per-order fees, split-cost structures where the consumer pays a small service charge, or simply does not happen.

Compare that to the alternative rails. Card processing typically lands somewhere in the 1.5 to 3.5 percent range depending on card type and interchange. BNPL is meaningfully more expensive per transaction. The merchant accepts the premium only if the incremental sales cover it. That is a testable claim, and it is where most merchant-side analysis goes wrong.

Running the math honestly. Suppose a merchant with a 40 percent gross margin pays 5 percent for BNPL. On a $200 order the fee is $10 against $80 of gross profit — the merchant keeps $70. For BNPL to be worth it, the offer must generate enough genuinely incremental orders to cover the fee paid on all the orders that would have happened anyway. If 80 percent of BNPL orders would have converted on a credit card regardless, the merchant is paying $10 on each of those for nothing, and the 20 percent that are truly incremental have to carry the whole cost. The break-even here is a real number a finance team can compute, and it is why serious merchants insist on holdout testing rather than accepting a provider's attributed-lift dashboard.

Consumer-side costs split by philosophy. Affirm charges interest on longer plans — up to 36 percent APR — and does not charge late fees. Klarna and Afterpay run interest-free on their core products and charge late fees instead, with Klarna's capped around $7 per missed payment and Afterpay's capped as a percentage of the order value, up to 25 percent. These are not arbitrary choices. Interest is a disclosed, regulated, APR-expressed price that scales with the size and length of the loan. Late fees are a penalty that falls entirely on the subset of customers who struggle — a structure that produces better optics on the headline offer and worse optics under regulatory scrutiny.

Timelines merchants should plan around. Integration through a hosted checkout plugin is a matter of days. A custom integration with product-page messaging, cart-level presentment, and order-management reconciliation runs weeks. Settlement lands in one to three business days. Refunds and returns flow back through the provider and can take a full billing cycle to unwind on the consumer's side, which generates support tickets the merchant did not anticipate. Chargeback and dispute handling differs from card rails and needs its own runbook.

Where the model breaks and where teams read it wrong

Mistake one: reading BNPL as a payment method rather than a lending business. Everything about BNPL's cost structure follows from the fact that the provider is warehousing consumer credit. Funding costs, loss reserves, servicing infrastructure, collections operations — none of that exists for a card processor. When rates rise, BNPL margins compress even if volume holds, because the receivables book gets more expensive to fund. Any analysis that models BNPL as pure take-rate-on-volume misses the largest cost line.

Mistake two: treating attributed lift as incremental lift. Providers report conversion improvements and average-order-value increases with real enthusiasm, and the underlying effect is genuine — installments do move behavior, especially on considered purchases where the sticker price is the objection. But the reported number almost always includes cannibalization: shoppers who would have bought anyway, choosing a different tender. The only honest measurement is a geographic or audience holdout with BNPL suppressed, run long enough to clear seasonality. Merchants that never run one are managing on a number they cannot defend.

Mistake three: assuming low default rates mean low risk. Charge-offs across the industry run in the range of roughly 1.5 to 2 percent, which is genuinely low and is the reason the model works at all. But 34 to 41 percent of BNPL users report having paid late at least once. That gap between "defaults" and "distress" is the interesting number. Low charge-offs on short-tenor loans reflect tight limits, short exposure windows, and the fact that most of the book was originated in benign conditions. The loss rate is the denominator the whole model rests on, and it is the metric most sensitive to a change in the macro environment. A book that looks profitable at 2 percent losses can invert at 4 percent, because the merchant fee has no room to absorb it.

Mistake four: ignoring loan stacking. A single provider sees its own repayment history. It does not automatically see that the same shopper has four open plans across three other providers, because BNPL reporting to credit bureaus has historically been inconsistent. That blind spot is the structural risk unique to this industry, and it is precisely what regulators have focused on.

Mistake five: launching without operational readiness. BNPL changes the return flow, the customer service script, the reconciliation process, and the fraud profile. First-party fraud — a real person who never intends to pay past the first installment — behaves differently from card fraud and is not covered by the same protections. Merchants that treat the integration as a checkout toggle discover the operational tail after launch.

Mistake six: over-indexing on the consumer-payer narrative. Public discussion of BNPL fixates on late fees and consumer harm, and there are legitimate concerns there. But the fee income is a minority of provider revenue and shrinking under regulation. A competitor, investor, or partner who models BNPL as a late-fee business will consistently misjudge how these companies behave, because their actual incentive is merchant retention and volume growth, not maximizing penalties on struggling borrowers.

Choosing a structure: a decision framework for merchants and operators

The right BNPL configuration depends on ticket size, margin, and what the offer is actually supposed to do.

If your average order is under about $75 and margins are thin, BNPL is usually a poor fit. The fixed per-transaction fee component eats the economics, the incremental lift on impulse-priced items is small because affordability was never the objection, and you will pay the fee on a large volume of orders that would have converted anyway.

If your average order sits in the $100 to $500 band with healthy margins, short-tenor pay-in-4 is the standard fit. Present it on the product page, not just at checkout. Measure with a holdout. Expect the lift to concentrate in categories where the price is a genuine hesitation point rather than in categories bought on routine.

If your average order exceeds roughly $800, the question changes from "does splitting help" to "who pays for the money." Longer-tenor financing with a merchant-subsidized 0 percent offer converts far better than an interest-bearing plan, but costs materially more per transaction. Model it as a promotional expense with a measurable payback, the way you would model a discount — because that is what it is.

If you sell across multiple price bands, run a multi-provider setup routed by cart value, and negotiate on total volume. Providers price aggressively for exclusivity; the trade is a better rate against losing the ability to route.

For operators outside retail, the transferable framework is a set of three questions. First, who captures the value your service creates — if a partner's revenue measurably rises because of what you do, that partner can fund it, and pricing to the end user leaves money on the table. Second, can you monetize more than one side, and what friction does each fee add — BNPL earns from merchants and consumers both, but every consumer-side charge invites regulatory and reputational cost that merchant-side pricing does not. Third, if your model carries risk on the balance sheet, what is the loss rate that flips it from profitable to underwater, and how far is today's number from that line. A RevOps team pricing a service on any of these three axes is solving the same problem BNPL solved at scale, and the discipline of stating the break-even loss rate explicitly is the part most teams skip.

The direction of travel in the industry reinforces all three. Providers have been building beyond the checkout — physical and virtual cards that carry the installment offer to any merchant, app-based shopping surfaces, loyalty and rewards, subscription tiers that waive fees. Every one of those moves does the same thing: it raises the number of transactions per user per month, which raises lifetime value, which lets the provider spend more to acquire and retain. It also diversifies revenue away from the two streams regulators are most focused on. Meanwhile regulatory attention has increased in the US, UK, EU, and Australia, converging on affordability checks, clearer disclosure, and consistent credit reporting. Consolidation follows regulation; compliance cost is fixed and favors scale. The category is moving from land-grab to discipline, and the operators who survive it will be the ones who controlled credit risk while everyone else was buying volume.

Related questions

Does BNPL show up on a credit report?

Increasingly yes, though inconsistently. Reporting practices vary by provider and market, and regulators have pushed toward consistent furnishing so that lenders can see a borrower's total installment exposure. Short interest-free plans have historically been the least likely to appear.

Why do merchants pay more for BNPL than for cards?

Because they are buying incremental sales, not just payment acceptance. The provider takes on underwriting, funding, and default risk that a card processor does not, and prices for it. The merchant accepts the premium when the conversion and basket-size lift covers the spread.

What happens to a BNPL plan if the customer returns the item?

The merchant issues the refund through the provider, which cancels or reduces the remaining installments and returns any amount already collected. Timing can lag a full billing cycle, so customers sometimes see a scheduled payment after the return — a common support ticket.

Is BNPL profitable for the providers?

It can be. Affirm reached profitability in fiscal 2025 with roughly $52 million in net profit, after years of prioritizing growth. Profitability depends on funding costs, credit losses staying near 2 percent, and enough scale to spread fixed servicing and compliance costs.

How do providers approve customers so fast without a hard credit check?

A soft inquiry plus proprietary signals — device data, email history, prior repayment on the network, order-to-history ratio — scored in milliseconds. Initial limits are deliberately small, so a wrong decision is cheap, and limits grow only after demonstrated repayment.

FAQ

How do BNPL providers make money if consumers pay no interest?

The merchant is the primary payer. Retailers pay a percentage of each transaction plus a fixed per-order component in exchange for the conversion and average-order-value lift installments produce. Merchant fees make up the bulk of industry revenue — the roughly $44.9 billion providers earned in 2025 against about $560 billion in GMV. Consumer charges, whether interest on longer plans or late fees on interest-free ones, are a real but secondary stream, and one that is shrinking as regulation tightens.

What happens when a customer misses a payment?

The auto-debit fails, the provider retries and notifies, and after a grace period of roughly 7 to 10 days a late fee may apply depending on the provider's model — Klarna caps around $7 per missed payment, Afterpay caps as a share of the order value up to 25 percent, and Affirm charges no late fees at all. The account is typically frozen against new purchases. Charge-offs across the industry run near 1.5 to 2 percent, but 34 to 41 percent of users report having paid late at least once, which is a better read on consumer strain than the default rate alone.

Is BNPL regulated the same way as credit cards?

Not identically, and the gap is narrowing. Regulators in the US, UK, EU, and Australia have moved to bring short-term installment credit under clearer disclosure, affordability-check, and fee-cap requirements, and to standardize credit reporting so lenders can see stacked plans. Historically short interest-free products faced lighter oversight than revolving credit, which is exactly the arbitrage regulators have been closing. Anyone building on BNPL rails should treat the compliance surface as expanding, not stable.

How do providers control credit risk on loans with no interest income?

Through tight limits and short exposure. First-time users get small caps, often in the $50 to $500 range, and earn increases only through clean repayment. The typical pay-in-4 plan is fully repaid in six weeks, so the provider's money is at risk for a short, bounded window. Underwriting combines a soft credit pull with behavioral and device signals scored in under a second. The combination is why default rates stay low despite minimal traditional credit screening.

Why don't merchants just offer their own installment plans?

Because doing it in-house means underwriting consumers, funding receivables, running collections, and carrying the default risk on the balance sheet — a lending operation most retailers have no interest in building. Outsourcing converts all of that into a per-transaction fee with immediate settlement and no credit exposure. The merchant pays a premium over card processing for the privilege, and for most retailers that trade is clearly worth it.

Is the BNPL market still growing?

Yes, but at a maturing pace. Roughly 13.7 percent growth in 2025 to about $560 billion in GMV is healthy, not explosive — the profile of a category moving past land-grab. Growth is slowing in the US and Western Europe where penetration is already high, and shifting toward emerging markets, new verticals like travel and healthcare, and non-checkout products that raise transactions per user rather than adding new users.

Sources

• Consumer Financial Protection Bureau — Buy Now, Pay Later research and market reports
• Federal Reserve Bank of New York — Liberty Street Economics on BNPL borrowing
• UK Financial Conduct Authority — Buy Now Pay Later regulation
• Affirm Holdings — Investor relations and quarterly results
• Klarna — Investor relations and financial reporting
• Block, Inc. — Investor relations (Afterpay results)
• Bank for International Settlements — Research on fintech credit and payments
• Australian Securities and Investments Commission — Buy now pay later reviews
• Business of Apps — Buy now, pay later app market data

Related on PULSE

• [How does the open-source and open-core business model work in 2027?](/knowledge/q13071)
• [How do you build pipeline in a regulated industry like banking?](/knowledge/q150)
• [What's the right cadence for benchmarking your sales metrics against industry peers (Pavilion, Bridge Group, OpenView)?](/knowledge/q258)
• [How do you correlate sales rep tenure and prior industry experience with product line success?](/knowledge/q9788)
• [What question would you ask to test if a salesperson truly understands their buyer's industry trends and challenges?](/knowledge/q14422)