THE
BUY NOW,
PAY LATER
STUDY

01 / Executive findings

The price is split. The obligation isn't.

Buy Now, Pay Later can be a genuinely useful form of short-term credit. It can also create a gap between what a checkout page makes feel affordable and the obligation that remains on the other side of the screen.

Manet Research examined Federal Reserve surveys, CFPB matched credit records, peer-reviewed behavioural experiments and ecommerce market data to understand what changes when a $200 purchase becomes "4 payments of $50." The evidence shows that BNPL can be an efficient financial tool, an interface that alters spending behaviour, a source of overlapping repayment schedules, and a potential trigger for bank fees — often in the same transaction.

Six findings, from six different sources, set the stage:

These six figures measure entirely different things: consumer perception, repayment difficulty, bank fees, borrowing patterns, market volume, and seasonal ecommerce activity. They do not combine into a single "BNPL cost."

02 / What BNPL actually is

"Pay in 4" is still a loan.

A standard pay-in-four BNPL transaction is straightforward: a $200 purchase is split into $50 at checkout and three additional $50 payments, usually biweekly, covering roughly six weeks. Most of the time, there is no interest if the schedule is met.

But "no interest" and "no cost" are different claims. The first describes a lender fee that may be zero. The second describes the total impact on the user's finances, which can include bank fees, overlapping obligations, and changes in spending behaviour — none of which appear on a lender's interest line.

BNPL is distinct from:

• Credit-card revolving debt — BNPL typically has a fixed repayment schedule and no compound interest.
• Installment loan — BNPL terms are shorter, often with softer underwriting.
• Layaway — BNPL delivers the product immediately.
• Debit card — BNPL moves the payment into the future.
• Longer-term point-of-sale financing — BNPL terms are measured in weeks, not years.

This distinction is not a critique of BNPL. It is a description of what BNPL does: it moves a payment through time and changes how that payment is presented. Both can be useful. Both can have consequences that the checkout page does not emphasize.

03 / How a niche payment became mainstream

From 19.8 million loans to 335.8 million in four years.

The scale of BNPL growth is difficult to overstate. In the CFPB's six-firm sample, loan originations rose from 19.8 million in 2019 to 335.8 million in 2023. Inflation-adjusted dollar originations rose from $2.7 billion to $45.2 billion over the same period. ^[4]

The UK tells a similar story. The FCA reported market growth from approximately £0.06 billion in 2017 to more than £13 billion in 2024. ^[12] Average loan size stayed near $131–$150 in inflation-adjusted terms throughout the CFPB period, meaning the growth came from volume — more people, more frequent use — rather than from larger individual transactions. ^[4]

BNPL is no longer an edge-case fintech product. It is a core part of the checkout experience for tens of millions of consumers across multiple markets. Understanding what happens after the first installment is not an academic exercise.

04 / The Installment Lens

What happens when $200 becomes "4 × $50"?

The Installment Lens is a Manet Research framework describing how a fixed total obligation can feel different when the payment schedule breaks it into smaller foreground numbers.

Peer-reviewed research published in the Journal of Retailing in 2025 found that BNPL installment-price presentation could reduce perceived purchase expensiveness and increase spending. ^[5] More installments and a smaller first installment further increased spending in the study. Importantly, installment-price display did not produce the same effect in other payment modes tested.

The mechanism is not a consumer flaw. It is a feature of how price information is processed. When a checkout page places "$50" at the top of the payment menu and "$200" in smaller type beneath it, the smaller number becomes more salient. The total obligation remains the same. The foreground number changes.

The Installment Lens connects to the creative-agency context of Manet Research. Pricing is not only arithmetic; it is also interface hierarchy, typography, proximity to the checkout button, and the order in which information is presented. The same $200 purchase can feel different depending on the design decisions made around it.

05 / Does BNPL make people spend more?

Three studies say yes. They do not say the same thing.

Three independent studies have found higher spending or sales under BNPL conditions. Each used a different methodology, a different outcome measure, and a different sample. They do not produce a single "BNPL effect size."

Central Bank of Ireland behavioural experiment: In a nationally representative Irish experiment, participants spent on average 4.39% more using BNPL compared with debit cards. Prior BNPL usage that inflated perceived available funds increased the likelihood of spending on a discretionary product by 22.2%. Salient risk disclosures improved understanding but did not significantly affect usage or spending. ^[6]

Journal of Retailing online retailer study: Using a synthetic difference-in-differences design around BNPL adoption at a focal online retailer, researchers found BNPL adopters increased online spending and order size by 6.42% relative to the comparison group. The effect was stronger among younger, lower-income consumers and for low-ticket items. ^[7]

NBER merchant experiment: In a randomized merchant experiment, offering BNPL increased sales approximately 20%. An NBER summary reported checkout conversion rising from roughly 70% to 80% in that setting. Much of the uplift came from lower-creditworthiness customers. Merchant benefits exceeded the financing and default costs in that setting. ^[8]

These three studies are not three estimates of one universal BNPL effect. They measure different things: consumer spending in an experimental setting, observed spending changes at a retailer, and merchant sales in a randomized experiment. They also come from different countries, different years, and different populations. The honest synthesis is that multiple credible designs have found a directional effect in the same general area — higher spending or sales under BNPL conditions — but that the specific size of that effect varies by context.

06 / Why people choose BNPL

Convenience is real. So is affordability pressure.

The Federal Reserve's 2025 SHED asked U.S. BNPL users why they chose BNPL for their most recent purchase. The answers show that BNPL serves multiple purposes simultaneously. ^[1]

BNPL can be a convenience product for one user, a cash-flow tool for another, and an affordability bridge for a third. The same interface serves different financial contexts. That is not a contradiction; it is a description of how a single product can fit multiple needs.

It also explains why BNPL is not a simple case of "people who cannot afford things buying things they should not buy." For many users, BNPL is a rational response to short-term liquidity constraints — a way to access a product today while earning the money to pay for it over the next several weeks.

07 / The Obligation Stack

Four payments are simple. Four loans are not.

The Obligation Stack is a Manet Research framework describing how multiple BNPL payment schedules can overlap, even when each individual schedule is straightforward.

In CFPB matched 2022 data, approximately 63% of BNPL borrowers originated multiple simultaneous BNPL loans at some point during the year. One-third (33%) did so across multiple BNPL firms. ^[3]

Consider the following illustrative scenario:

• Loan A: $28 / $28 / $28
• Loan B: $42 / $42 / $42
• Loan C: $18 / $18 / $18

Each checkout looks small. Combined next Tuesday, the obligation is $88. The interface for each merchant shows only its own schedule. The shopper is the only one who has to aggregate them.

The Obligation Stack does not mean BNPL is inherently dangerous. It means that a shopper who uses multiple BNPL plans across multiple merchants has a financial picture that no single checkout page shows. The interface is designed for a single transaction. The obligation can extend across many.

08 / Heavy users

For one in five borrowers, BNPL was more than an occasional checkout option.

The CFPB matched study defined "heavy users" as BNPL borrowers with more than 12 originations in a year. That group represented 20.3% of BNPL borrowers in 2022. ^[3]

Heavy users had a median of 23 BNPL originations per year, compared with 2 for occasional users. They had a median of 13.6 days between originations, compared with 41.4 days. And they had an average of 2.6 active BNPL loans on any given day, compared with 0.3 for occasional users. ^[3]

These figures describe behavior, not cause. They do not prove that heavy users are "addicted" to BNPL or that they cannot manage their finances. They describe a pattern of frequent use that is different from the occasional checkout convenience that many BNPL interfaces are designed to support.

09 / When 0% interest is not zero cost

Zero interest answers one question. It does not answer whether a payment will trigger another fee somewhere else.

The phrase "0% interest" is accurate for most pay-in-four BNPL transactions. It is also incomplete as a description of the total cost.

The Federal Reserve's 2025 SHED found that 26% of U.S. BNPL users said they had paid late on a BNPL payment in the prior year. Of those, 64% were charged extra for paying late. Overall, 17% of all BNPL users were charged extra for paying late. And 11% reported that a BNPL payment triggered an overdraft or NSF fee from their bank. ^[1]

Meanwhile, CFPB market-level data for the six large firms found that in 2023, 4.1% of loans assessed a late fee, with the average inflation-adjusted late fee assessed at $9.99. Late fee revenue represented 0.18% of origination volume. And 1.83% of loans were charged off or uncollectible. ^[4]

The gap between self-reported late payment (26%) and loan-level late fee incidence (4.1%) is not a contradiction. The Federal Reserve figure counts users who were late at least once in the prior year, not loans. The CFPB figure counts loans that assessed a late fee. A user can be late once in a year and still have paid late. A loan can be late and still not be charged off. The two figures describe different levels of the same system.

10 / Essentials

"Pay later" is no longer just for sneakers.

The Federal Reserve's August 2026 Consumer & Community Context report found that BNPL is being used for a broad range of purchases, including essentials like groceries and medical bills. ^[2]

The most common purchase categories for BNPL users were:

• Clothing and accessories — 49%
• Electronics — 32%
• Furniture and appliances — 26%
• Groceries and food delivery — 20%
• Travel — 19%

Among BNPL users with income below $50,000, 29% used BNPL for groceries or food delivery. ^[2]

The same Federal Reserve report found that among BNPL users who reported being charged extra for late payment or incurring a BNPL-triggered bank fee, the share was highest for those who used BNPL for groceries or food delivery (43%) and medical or veterinary expenses (34%). ^[2]

These numbers do not mean that buying groceries with BNPL is irresponsible. They mean that BNPL is being used for essential spending by people who may already be financially stretched, and that those users appear more likely to experience downstream costs from the repayment schedule.

11 / Who is most exposed?

BNPL use rises as financial cushions shrink.

The Federal Reserve's August 2026 analysis found that BNPL use was substantially higher among consumers with less emergency savings capacity. Among those who could cover less than $100 of an emergency expense from savings, 31% had used BNPL. Among those who could cover $2,000 or more, only 8% had used BNPL. ^[2]

The same pattern appeared for BNPL-triggered overdraft and NSF fees. Among users who could cover less than $100 from savings, 18% experienced a BNPL-triggered bank fee. Among those with $2,000 or more in savings capacity, only 4% did. ^[2]

Income and age also mattered in the Federal Reserve data. BNPL use was highest among adults aged 18–29 (22%) and 30–44 (21%), and among those with incomes between $25,000 and $49,999 (23%). Late payment rates were also higher among younger users (32% for 18–29, 31% for 30–44) and lower-income users (40% for under $25k, 33% for $25k–$49,999). ^[1]

This is a critical distinction. The data shows that people who are already financially stretched are more likely to use BNPL and more likely to experience downstream costs. It does not prove that BNPL caused their financial situation. The correlation is real. The causal story is more complex.

The FCA's Financial Lives Survey 2024 found that among the UK's 1.9 million frequent BNPL users (10+ uses per year), 76% were women, 67% had personal income below £30,000, 54% had low financial resilience, and 26% were in financial difficulty. ^[11] Again, this is a profile, not a causal claim.

12 / Why merchants offer BNPL

The merchant does not offer four payments as a public service.

BNPL is not only a consumer product; it is a merchant product. Merchants pay BNPL providers a fee, typically a percentage of the transaction, in exchange for offering the payment option. The incentive for merchants is clear: if BNPL increases conversion, order size, or customer acquisition, the fee can be worthwhile.

The NBER randomized merchant experiment found that offering BNPL increased sales by approximately 20% in that setting. Checkout conversion rose from roughly 70% to 80%. Much of the uplift came from lower-creditworthiness customers, and merchant benefits exceeded financing and default costs in the study. ^[8]

This is not cynical. It is a description of a rational commercial arrangement. Merchants offer BNPL because it can increase revenue. BNPL providers charge a fee and take the credit risk. Consumers get a useful payment option. The arrangement can be a win-win-win.

The merchant economics also explain why BNPL is promoted prominently at checkout. The installment price is not only consumer-friendly; it is also commercially effective. The interface design that makes $50 feel smaller than $200 is the same design that generates incremental sales for the merchant. The two are not in conflict; they are part of the same economic logic.

13 / Holiday & mobile checkout

In 2025, BNPL moved $20 billion through U.S. holiday ecommerce.

Adobe Analytics reported that U.S. online holiday spending using BNPL reached $20.0 billion in 2025, up from $18.2 billion in 2024 and $16.6 billion in 2023. ^[16]

82.2% of BNPL holiday purchase activity in 2025 occurred on smartphones. ^[16] Cyber Monday BNPL spending crossed $1 billion for the first time, reaching $1.03 billion. ^[16]

The mobile context matters for the Installment Lens. A narrow viewport, a thumb-scale checkout, and reduced information surface mean that the installment amount can be even more salient relative to the full price. The payment method is increasingly part of the merchandising, not just the transaction infrastructure.

14 / Credit reporting

The debt that was hard to see is becoming easier to see.

One reason BNPL has been distinct from traditional credit is that many BNPL loans were not reported to credit bureaus. In 2022, the CFPB found that the majority of BNPL loans in its matched sample did not appear in credit records. ^[3]

That is changing. Affirm announced in March 2025 that it would expand reporting of U.S. pay-over-time products — including Pay-in-4 — to Experian for products issued from April 1, 2025 onward. At the time of the announcement, this information was not yet intended to be incorporated into traditional credit scores in the near term. ^[18]

TransUnion's current guidance states that BNPL data furnished to TransUnion is visible to the consumer but is not currently available to scoring providers or lenders in the same way for credit decisions, though future use may change. ^[18]

The credit-reporting transition is important because it changes the visibility of BNPL obligations. A debt that was hard for other lenders to see is becoming easier to see. That can be good for responsible lending decisions. It can also affect consumers who have been using BNPL without thinking of it as "debt" in the credit-reporting sense.

15 / Regulation

Governments increasingly stopped treating BNPL as "just another payment button."

Across several jurisdictions, BNPL has moved from a lightly regulated payment option toward a more regulated form of consumer credit.

• Ireland: Regulated short-term credit, including BNPL, from 2022.
• New Zealand: BNPL brought under consumer credit law from September 2024. ^[14]
• Australia: BNPL providers required to hold a credit licence from June 2025. ^[13]
• United States: The CFPB's 2024 interpretive rule — which would have classified BNPL lenders as credit card issuers for certain purposes — was withdrawn on May 12, 2025. ^[15]
• United Kingdom: The FCA began regulating third-party Deferred Payment Credit from July 15, 2026. ^[12]

The regulatory direction is not uniform. Several jurisdictions have moved toward treating BNPL as a form of credit, with corresponding consumer protections. The U.S. federal approach shifted in 2025 with the withdrawal of the interpretive rule. The regulatory landscape remains mixed.

16 / What BNPL gets right

There is a reason people keep choosing it.

This section is mandatory. BNPL is not a predatory product. It has genuine benefits that explain its rapid adoption.

• Predictable fixed payments. Unlike credit cards with variable minimum payments and compounding interest, BNPL has a clear schedule.
• Commonly no interest under pay-in-four if paid on time. For many users, BNPL is cheaper than revolving credit-card debt.
• Can be cheaper than credit cards. The CFPB estimated that many BNPL borrowers would face credit-card APRs around 19–23% if they financed purchases on cards.
• Automatic payments can support repayment. Many BNPL providers use autopay, which can help users stay on schedule.
• Default rates lower than credit cards. The CFPB study found BNPL default rates were structurally lower than credit-card default rates among the same borrower population, though product structures differ. Even deep-subprime and no-FICO borrowers repaid BNPL loans approximately 96% of the time under the study's default definition.
• Can smooth cash flow. BNPL allows a purchase to be made today and paid for over the next several weeks.

BNPL is not a scam. It is not inherently dangerous. It is a financial product with genuine utility and genuine risks, both of which depend on how it is used and on the user's financial context.

17 / The Four Costs of Four Payments

So what does "4 easy payments" actually cost?

The Manet Research framework identifies four distinct categories of cost that can arise from a BNPL transaction. They are not additive; they are analytical.

01 — FINANCING COST: Interest and lender fees. Often zero for pay-in-four when paid on time.

02 — BANK-ACCOUNT COST: Overdraft and NSF fees triggered by automatic repayments. 11% of U.S. BNPL users reported this in the Federal Reserve survey, rising to 38% among those charged extra for lateness. ^[1]

03 — SPENDING COST: The possibility that installment framing changes purchase likelihood, order size, or total spending. Multiple studies find higher spending or sales under BNPL conditions, but the specific effect size varies by context. ^[5][6][7][8]

04 — OBLIGATION COST: The complexity and financial load created by overlapping payment schedules, multiple providers, and invisible debt. 63% of matched CFPB borrowers had multiple simultaneous BNPL loans at some point in 2022. ^[3]

These four categories are not additive. A transaction can involve one, several, or none of them. The framework is a way of asking: "Which of these costs apply here, and what evidence exists to assess them?"

18 / Interactive tool

The 6-Week Payment Stack

Enter the details of your BNPL plans to see how the payments stack across the next six weeks. Everything runs in your browser — nothing you type is sent anywhere or stored.

This tool does not provide financial advice. It is an information aid designed to help visualize payment schedules.

19 / Shopper playbook

Before you split the price, reconstruct the full obligation.

BNPL is not good or bad. It is a tool. Like any tool, its value depends on how and when it is used.

• Read the full price first. Do not let the installment amount be the only number you evaluate.
• Multiply installment × number of installments. The total is what you actually owe.
• Put all active BNPL plans on one six-week calendar. You are the only one who can see the full picture.
• Check whether autopay can hit a low-balance bank account. An overdraft fee can cancel out any interest savings.
• Check lender late-fee policy. Know what happens if a payment is missed.
• Compare with paying in full. Sometimes a discount is just a discount, not a reason to split the payment.
• Compare with a credit card only if you know whether you would carry the balance and incur interest. BNPL can be cheaper for some users.
• For essentials, consider whether next month's income is already committed. A grocery purchase today is still a grocery bill next month.
• Do not use the installment amount as the affordability test. The installment is the beginning of the obligation, not the end.
• Treat each pay-in-four plan as debt even if the interface calls it a payment option. A loan by any other name still has a repayment schedule.

20 / What the evidence cannot prove

What we cannot conclude

Credibility requires stating the limits of this evidence as plainly as its findings.

• BNPL does not universally increase spending by one fixed percentage. The three spending studies use different methods, outcomes, and populations.
• A correlation between financial vulnerability and BNPL use does not establish BNPL caused vulnerability. People who are already financially stretched may be more likely to use BNPL.
• Late-payment survey results and loan-level fee rates use different denominators. The Federal Reserve's 26% and the CFPB's 4.1% are not in conflict; they measure different things.
• The 63% simultaneous-loan figure is from six specific firms and matched 2022 data, not all users today. It describes a pattern in a specific dataset, not a universal law.
• Merchant sales uplift varies by merchant, product, and customer context. The NBER result is from one randomized experiment, not a universal law of BNPL economics.
• Regulation differs by country and product type. Not all BNPL products are regulated in the same way, even within the same jurisdiction.
• Interest-free BNPL can genuinely reduce financing cost compared with revolving card debt for some users. That is a real benefit, not a marketing illusion.
• Survey self-reports are different evidence from transaction records. The Federal Reserve and CFPB data use different methods.
• This report does not provide individual financial advice. It is an information resource, not a recommendation about whether to use BNPL.

21 / Methodology

How Manet Research conducted the analysis

Manet Research conducted a cross-market secondary-data investigation into Buy Now, Pay Later, synthesizing government datasets, regulatory research, longitudinal credit data, behavioural experiments, peer-reviewed retail studies, and ecommerce market data.

What Manet Research contributed

• bringing unlike datasets into one evidence architecture
• distinguishing direct fees from behavioural and balance-sheet costs
• comparing consumer-side and merchant-side evidence
• identifying where findings converge and where they conflict
• creating an original framework for understanding "the cost" of BNPL
• creating original visualizations from published values
• translating the evidence into a practical consumer decision tool

Source hierarchy

Evidence was prioritized in the following order:

• government, central bank, and regulator datasets
• matched administrative credit records
• peer-reviewed and academic empirical studies
• nationally representative surveys
• large ecommerce market datasets
• provider and bureau material only for provider-specific operational changes

Normalization

The source studies do not share a common definition of "use," "cost," or "spending." Rather than force them into one number, findings are grouped throughout this report by the question they answer, presented with their original sample and methodological context intact.

22 / Conclusion

Four small payments still add up to one full decision.

Buy Now, Pay Later solves a real problem. It lets people move a purchase through time without necessarily paying interest for the privilege.

But the checkout does more than move the payment. It changes which number is most visible, how the obligation is scheduled, and how easily another obligation can be placed beside it.

For some shoppers, BNPL is useful liquidity. For others, it is a stack of future payments that arrived one easy checkout at a time.