Risks and Costs of BNPL
Imagine an ecommerce team presenting its BNPL launch after one quarter. The dashboard says: Checkout conversion: +8% Average order value: +15% The room likes the numbers. Then finance adds a second page. BNPL transactions cost more to process than the merchant's previous payment
Risks and Costs of BNPL: What Merchants Should Measure Beyond Conversion
Imagine an ecommerce team presenting its BNPL launch after one quarter.
The dashboard says:
Checkout conversion: +8%
Average order value: +15%
The room likes the numbers.
Then finance adds a second page.
BNPL transactions cost more to process than the merchant's previous payment mix. Some of the larger baskets came back as returns. Refunds created additional payment adjustments. Support tickets increased. Several customers who selected financing were declined. Some existing card and debit transactions simply migrated to the new method. Fraud and disputes required operational work.
The original dashboard was not wrong.
It was incomplete.
The +8% and +15% figures above are deliberately hypothetical. They illustrate the central problem with BNPL measurement: gross commercial metrics can improve while contribution deteriorates.
A merchant therefore needs to ask a different question:
How much incremental gross profit did BNPL create after every incremental cost required to produce, service, reverse, and protect those orders?
That number—not financed GMV—is the economic case for BNPL.
Start with incremental gross profit, not BNPL volume
BNPL volume contains several kinds of transactions.
Some are genuinely new orders from customers who otherwise would not have purchased.
Some represent larger baskets that would otherwise have been smaller.
Some accelerate purchases a customer would have made later.
Some simply move an existing purchaser from a lower-cost payment method into BNPL.
Those effects should not be treated as economically equivalent.
Suppose a merchant processes $10 million through BNPL. That tells finance how much financed merchandise passed through the method.
It does not tell finance:
how much revenue was genuinely incremental;
how much merchandise was later returned;
how much gross margin those orders generated;
how much payment cost increased;
how much card or debit volume was displaced;
how many financing attempts were declined;
how much fraud and support expense followed;
whether the method changed customer lifetime value.
The first denominator should therefore be incrementality:
Would this order have occurred, at this value and at this time, without BNPL?
Only after answering that question should a merchant build its profit model.
Direct BNPL costs are the visible part of the bill
The most obvious cost is the provider charge.
Depending on provider, product, country, contract, and financing structure, merchant economics can contain:
a percentage transaction fee;
a fixed transaction fee;
a merchant financing subsidy;
implementation or integration expense;
platform or operational fees where contractually applicable;
cross-border or currency-related cost where relevant.
There is no responsible universal “BNPL costs X%” figure.
Merchant pricing is product- and contract-specific. For due diligence, finance should obtain the actual commercial schedule rather than using an industry benchmark or another merchant's pricing.
More importantly, the provider fee is not the total payment cost.
That distinction is where many BNPL business cases become too optimistic.
Provider fee versus total payment cost
Define:
Direct provider cost
as the amount contractually charged for processing or financing the transaction.
Define:
Total BNPL payment cost
as:
**provider cost
implementation
refund-related payment cost
dispute/fraud cost
reconciliation
support
settlement/cash-management effects
attributable operating overhead**
Then add the commercial effects of returns and method substitution separately.
A method with an acceptable headline fee can still be economically weak if operating and commercial consequences are large.
Returns can rewrite the original economics
A BNPL order is not economically finished when checkout succeeds.
This is especially important in retail categories where customers routinely return part or all of a basket.
Suppose BNPL increases an order from $100 to $120.
If $40 of merchandise later comes back, the relevant retained revenue is $80—not the $120 captured by the original AOV report.
Then the merchant needs to examine:
gross margin on the retained merchandise;
fulfillment cost;
reverse-logistics cost;
restocking or markdown loss;
payment-fee treatment;
refund-processing cost;
support activity.
This is why a merchant should track both:
Original BNPL AOV
and
Post-return retained BNPL AOV
A merchant should not assume BNPL causes higher return rates without evidence. Return behavior must be tested against comparable customer, product, transaction-value, promotion, and acquisition cohorts.
But a business that already experiences high returns has less room for a payment method whose cost is incurred on transactions that are later substantially reversed.
AOV and retained margin answer different questions
AOV answers:
How large was the order at checkout?
Retained gross margin answers:
How much economic value survived the retail lifecycle?
For a CFO, the latter is far closer to the decision metric.
Refunds are part of payment economics
A return is a merchandising event.
A refund is also a payment event.
In BNPL, that payment event can involve several relationships:
Merchant approves refund
→ provider records merchant-side adjustment
→ customer's financing balance changes
→ future installments may be changed or cancelled
→ amounts already collected may need to be returned
→ settlement and reconciliation records change
The mechanics vary materially between providers and products.
Merchants therefore need to verify:
full-refund behavior;
partial-refund behavior;
multiple partial refunds;
provider-fee treatment;
settlement adjustments;
customer repayment adjustments;
refund timing;
dispute interaction;
cancellation before or after financing has begun.
Do not model a BNPL refund as though it were universally identical to a card refund.
The economic question is not simply, “Can we refund?”
It is:
What does a refund cost us, how much work does it create, and what portion of the original payment expense remains after the order is reversed?
Fraud risk is not credit risk
This distinction is essential.
In a provider-funded BNPL model, the provider may underwrite the customer and bear defined exposure to the customer's failure to repay.
That does not logically transfer every merchant risk to the provider.
Merchant exposure can still include, depending on contract and business model:
account takeover;
stolen or synthetic identity;
first-party misuse;
high-risk resale;
triangulation;
refund abuse;
non-delivery disputes;
merchant fulfillment mistakes;
evidence requirements;
restricted-product violations.
The appropriate question is not:
“Who carries the BNPL risk?”
It is:
“Which risk is allocated to which participant?”
A useful control map distinguishes:
Provider: customer eligibility, underwriting, financing decision, repayment collection.
Merchant: product legitimacy, fulfillment, return decision, merchant-side fraud controls, dispute evidence, customer service.
Shared/contractual: transaction fraud allocation, dispute liability, refund mechanics, settlement adjustments, loss allocation.
Credit exposure and merchant operating exposure are different categories.
Disputes remain a merchant cost even without conventional card economics
BNPL financing does not eliminate disputes.
A transaction can still produce questions or claims involving:
non-delivery;
merchandise not as described;
duplicate transactions;
cancellation;
customer identity;
fraud;
refunds;
service quality.
The merchant may need to investigate, produce evidence, coordinate with a provider, reconcile adjustments, and communicate with the customer.
Even where the provider ultimately bears the monetary loss, the merchant can still bear operating cost.
That should be measured.
A useful metric is:
Dispute operating cost per BNPL order
=
internal dispute labor + external dispute fees + attributable losses
÷ completed BNPL orders
Finance should also monitor outcomes rather than dispute counts alone.
A customer decline is not a payment failure
BNPL creates an approval layer that conventional payment reporting can obscure.
The customer journey may include:
BNPL visible
→ BNPL selected
→ financing flow started
→ required information completed
→ provider decision
→ approved or declined
→ terms accepted
→ payment confirmed
→ order completed
A customer declined by the financing provider is different from:
a technical checkout failure;
merchant gateway downtime;
a failed settlement;
an ordinary card authorization decline.
That distinction matters because BNPL's addressable customer-access benefit is constrained by provider eligibility and approval.
If the customers most in need of payment splitting have low approval rates, the theoretical affordability benefit may not become a merchant conversion benefit.
Do not compare provider approval rates without controlling for transaction value, country, product, merchant category, and customer cohort.
Operational cost is easy to hide in another department's budget
BNPL can add work outside the payment processing line item.
Customer support
Teams may receive questions about:
eligibility;
declines;
repayments;
refunds;
cancelled orders;
returns;
payment schedules;
provider-versus-merchant responsibility.
Finance and reconciliation
Finance may need to connect:
merchant order
→ BNPL provider transaction
→ provider fee
→ settlement
→ refund adjustment
→ dispute adjustment
→ ledger entry
Every exception that cannot be automatically reconciled has a cost.
Engineering
Integration expense does not end at launch.
Teams may need to maintain:
APIs;
provider updates;
SDKs;
product messaging;
callbacks/webhooks;
reporting;
regional configurations.
Legal and compliance
BNPL increasingly sits inside formal consumer-credit regulation in major markets. That can affect provider selection, contracts, checkout messaging, product governance, and launch review.
These are real payment-method costs even if the payment team's processor-fee report does not contain them.
Settlement timing belongs in the model
Customer repayment and merchant settlement are separate financial events.
A shopper may repay a purchase over weeks or months while a provider settles with the merchant according to a different contractual timetable.
Finance therefore needs to model:
authorization date;
capture/merchant confirmation;
provider settlement date;
payout date;
refund deduction;
reserve or adjustment mechanisms where applicable;
currency conversion where relevant.
Two providers with similar transaction fees can produce different treasury or reconciliation effects if their settlement structures differ.
Do not infer merchant settlement timing from the customer's installment schedule.
Substitution is one of BNPL's least visible costs
Suppose a customer would have purchased using a debit card.
After BNPL is introduced, the same customer buys the same item, at the same price, at the same time—but chooses BNPL.
Revenue has not increased.
Conversion has not increased.
Customer acquisition has not increased.
The payment mix has changed.
If BNPL has a higher merchant cost, the difference is an incremental expense.
This creates the substitution cost:
Substitution cost
=
BNPL cost for non-incremental migrated transactions
− cost of the payment method those orders would otherwise have used
That number should be separated from the economics of genuinely incremental BNPL customers.
A high BNPL selection rate can therefore be economically ambiguous.
It may demonstrate access.
Or it may demonstrate cannibalization.
Incrementality is the bridge between conversion and profit
The correct causal question is:
What changed because BNPL existed?
Possible positive effects include:
previously abandoned purchases become completed;
a customer purchases sooner;
a customer buys a more valuable product;
a new customer becomes reachable;
retained basket size increases.
Possible neutral or negative effects include:
customers switch from cheaper existing methods;
customers start BNPL but are declined;
larger baskets produce more returned merchandise;
payment cost increases without incremental gross profit;
support and dispute expense rise.
The conversion article in this cluster should own detailed experiment design.
For this economics article, the important point is that every cost should be compared with causal, incremental benefit—not with total BNPL volume.
Regulatory cost is now part of the ownership model
BNPL regulation is not converging on one uniform global regime.
Australia extended its consumer-credit framework to BNPL contracts, with relevant credit-licensing requirements applying from June 10, 2025. ASIC describes third-party BNPL as an arrangement where a merchant supplies the goods or services, a BNPL provider pays some or all of the price, and the provider supplies credit to the consumer.
In the United Kingdom, the FCA began regulating third-party Deferred Payment Credit on July 15, 2026. The FCA says qualifying agreements supplied through third-party lenders are now regulated credit agreements, while the broking of those agreements is exempt; its regime includes requirements aimed at informed consumer decisions and responsible, affordable lending.
In the European Union, the revised Consumer Credit Directive was required to be transposed into national law by November 20, 2025 and is due to apply from November 20, 2026. That makes local implementation and product scope an immediate pre-publication and pre-launch verification item for European merchants.
The United States illustrates another form of regulatory uncertainty. The CFPB's current BNPL resources note that it withdrew its 2024 BNPL interpretive rule on May 12, 2025.
For merchants, these differences create several costs:
legal review;
contract changes;
financing-message review;
provider due diligence;
checkout/disclosure changes;
country-specific enablement;
policy monitoring;
potential product or eligibility changes.
The provider will often carry the primary lender obligations, but merchants should not conclude that provider compliance eliminates merchant implementation or reputational consequences.
This article provides general information and does not constitute legal, regulatory, financial, credit, tax, or compliance advice. Merchants should verify current requirements with qualified advisers, BNPL providers, payment partners, and relevant authorities.
Consumer risk can become merchant risk indirectly
A merchant may not be the lender but still benefits from an economically sustainable financing experience.
The CFPB has specifically studied persistence of BNPL use, loan stacking across providers, and BNPL users' broader unsecured-debt positions. Its 2025 research used matched application/origination data from six large U.S. pay-in-four firms and de-identified credit records.
The merchant implication is not that every BNPL customer is overextended.
It is that customer-credit outcomes are legitimate considerations rather than irrelevant lender-only issues.
Problems with affordability can become:
regulatory attention;
customer complaints;
reputational risk;
stricter approval policies;
changes in provider economics;
changes in checkout communication.
A financing method that creates short-term merchant volume while generating poor long-term customer outcomes may not remain commercially attractive.
Provider concentration is a strategic payment risk
A merchant can become dependent on a BNPL provider for:
approval coverage;
customer recognition;
checkout messaging;
settlement;
refund workflows;
dispute handling;
reporting;
market access.
That dependency creates concentration risk.
Possible changes include:
pricing;
merchant-category eligibility;
customer approval policy;
transaction limits;
country availability;
contractual risk allocation;
regulatory status;
product branding or structure.
The merchant does not need to integrate every provider to avoid concentration.
It does need to know:
How much profitable demand would disappear if this provider became more expensive, more restrictive, or unavailable?
That is a treasury and payment-strategy question, not only an engineering question.
The BNPL Total Cost Model
The purpose of the model is to move from gross financed performance to incremental merchant contribution.
Layer 1: Incremental commercial benefit
Measure:
Incremental completed orders
Orders that would not have occurred without BNPL.
Incremental retained basket value
Additional merchandise revenue that survives returns and refunds.
Incremental gross profit
Gross profit attributable to the genuinely incremental portion of BNPL.
Do not start with total BNPL GMV.
Layer 2: Direct payment costs
Include:
variable provider fee;
fixed fee;
financing subsidy where applicable;
implementation cost;
ongoing integration cost;
relevant cross-border/currency cost.
Layer 3: Operational costs
Include:
refund processing;
reconciliation;
dispute handling;
customer support;
reporting;
finance operations;
engineering maintenance;
manual exception handling.
Layer 4: Risk losses
Include where attributable:
fraud;
dispute losses;
refund abuse;
fulfillment-related losses;
unrecoverable provider/settlement adjustments.
Separate these from the provider's consumer credit losses.
Layer 5: Commercial offsets
Subtract:
return-driven lost gross profit;
substitution from cheaper payment methods;
promotion/discount expense associated with BNPL;
incremental customer-acquisition cost if required to produce the financed orders.
Layer 6: Strategic overhead
Monitor separately:
regulatory implementation;
provider concentration;
merchant eligibility;
reputational effects;
policy-change costs.
Some of these are difficult to allocate per transaction. That does not make them zero.
Core formula: incremental BNPL contribution
A useful finance model is:
Incremental BNPL contribution
=
Incremental retained gross profit
− Incremental provider/payment costs
− Incremental returns/refund operating cost
− Incremental fraud/dispute loss
− Incremental support/reconciliation cost
− Substitution cost
− Other attributable BNPL operating cost
The word incremental is doing most of the work.
If an expense would have occurred anyway, do not charge all of it to BNPL.
If an order would have occurred anyway, do not count all of its gross profit as a BNPL benefit.
That symmetry keeps the model credible.
Supporting visual: BNPL Total Cost Model
BNPL ENABLED
│
▼
GROSS COMMERCIAL EFFECT
┌────────────────────────┐
│ Orders / conversion │
│ Original AOV │
│ Financed payment volume│
└────────────┬───────────┘
│
▼
REMOVE NON-INCREMENTALITY
┌────────────────────────┐
│ Existing card/debit │
│ orders that migrated │
│ Purchases merely moved │
│ forward in time │
└────────────┬───────────┘
│
▼
RETAINED GROSS PROFIT
┌────────────────────────┐
│ Returns │
│ Partial refunds │
│ Product mix / COGS │
└────────────┬───────────┘
│
▼
DIRECT COST
┌────────────────────────┐
│ Provider fee │
│ Fixed fee │
│ Financing subsidy │
│ Integration │
└────────────┬───────────┘
│
▼
OPERATIONS + RISK
┌────────────────────────┐
│ Support │
│ Reconciliation │
│ Refund operations │
│ Fraud / disputes │
│ Engineering │
└────────────┬───────────┘
│
▼
STRATEGIC OVERHEAD
┌────────────────────────┐
│ Regulation │
│ Concentration │
│ Eligibility changes │
│ Reputation │
└────────────┬───────────┘
│
▼
INCREMENTAL BNPL CONTRIBUTION
│
┌───────────┴───────────┐
▼ ▼
POSITIVE NEGATIVE
Scale selectively Redesign, narrow,
and keep testing renegotiate or exit
Finance metrics that matter
Incremental completed orders
BNPL-attributable additional completed orders
÷ eligible BNPL customer opportunities
Requires a defensible counterfactual.
BNPL selection rate
BNPL selections
÷ eligible sessions where BNPL was visible
Useful operationally, but not a profitability metric.
Application completion rate
Completed BNPL applications
÷ started BNPL applications
Separates financing-flow friction from merchant checkout demand.
Approval rate
Approved applications
÷ completed applications submitted for decision
Segment by market, provider, transaction value, and customer cohort.
Payment completion rate
Confirmed BNPL payments
÷ approved BNPL transactions
Do not combine approval and successful order completion.
Original BNPL AOV
Gross value of BNPL orders
÷ completed BNPL orders
Useful, but incomplete.
Retained BNPL AOV
Net retained merchandise revenue after returns/refunds
÷ completed BNPL orders
Often much more meaningful for retail.
Return value rate
Returned BNPL merchandise value
÷ fulfilled BNPL merchandise value
Use a comparable non-BNPL cohort before attributing causation.
Total BNPL cost per completed order
Direct + attributable operating/risk costs
÷ completed BNPL orders
This tells finance what an average financed order costs to operate.
BNPL cost per incremental order
Incremental BNPL cost
÷ genuinely incremental completed BNPL orders
This denominator can expose a method that looks cheap against total volume but expensive against actual incremental demand.
Substitution rate
Estimate:
BNPL orders that would otherwise have used an existing successful payment method
÷ total BNPL orders
This is difficult but strategically important.
Fraud/dispute loss rate
Net BNPL fraud and dispute loss
÷ completed BNPL payment value
Separate monetary loss from operating effort.
Support cost per BNPL order
BNPL-attributable support cost
÷ completed BNPL orders
Track both contact rate and handling time.
Incremental contribution
The primary metric.
Report it:
in absolute currency;
per eligible session;
per incremental completed order;
by country;
by transaction-value band;
by product category;
by new/returning customer.
A single blended global BNPL contribution number can conceal unprofitable segments.
A hypothetical BNPL economics example
Consider a hypothetical merchant that produces:
+8% conversion
and
+15% AOV
after BNPL rollout.
The payment team calls the launch a success.
Finance should now ask:
What portion of the additional orders was causal?
What portion merely shifted from another method?
What gross margin did the incremental products carry?
How much additional merchandise was returned?
Which provider fees remained after refunds?
Did fraud/disputes change?
How much incremental support was required?
How much did reconciliation cost?
Were financing declines concentrated among the target affordability cohort?
What contribution remained?
Only if the answer to number ten is positive does the headline conversion result become economically convincing.
When BNPL economics can fail
Low incremental demand
Customers use BNPL, but most would have purchased anyway.
Result:
payment-share change without enough new gross profit.
Thin merchant margin
The method produces incremental sales, but provider and operating cost absorb much of the extra gross profit.
Result:
revenue growth with weak contribution.
High-return categories
Checkout basket expansion looks attractive, but post-return retained merchandise value is substantially smaller.
Result:
AOV overstates value.
Low approval in the intended cohort
The customers facing the strongest affordability constraint are frequently not approved.
Result:
theoretical access does not translate into completed orders.
Expensive substitution
Customers move from inexpensive successful payment methods into a more expensive financing method without changing purchasing behavior.
Result:
unchanged demand at higher payment cost.
Complex refund or dispute operations
A business with frequent cancellations, partial refunds, or service disputes can incur material finance and support overhead.
Result:
operating cost grows faster than the payment team's fee model predicts.
Rapid regulatory change
Merchant checkout, provider contracts, product eligibility, or disclosure requirements need repeated updates.
Result:
the payment method carries ongoing compliance and implementation overhead.
Provider dependence
A large portion of financed contribution relies on a single provider.
Result:
commercial or eligibility changes can affect a meaningful part of profitable checkout demand.
The merchant should not optimize all BNPL customers equally
Segment the economics.
A BNPL method may create strong value for:
higher-ticket purchases;
new customers;
a particular country;
specific product categories;
customers without sufficient conventional credit;
planned purchases with healthy gross margin.
It may perform poorly for:
very low-ticket orders;
customers already converting successfully;
thin-margin products;
return-heavy categories;
high-fraud merchandise;
complicated cancellation models.
The correct decision can therefore be:
Offer BNPL, but not everywhere.
Eligibility and presentation can be treated as commercial controls, subject to provider and legal requirements.
A stronger experiment: measure profit, not the BNPL button
A useful BNPL experiment should ultimately compare:
incremental contribution per eligible customer opportunity
between treatment and control.
Track through the full lifecycle:
Eligible customer
→ BNPL visible
→ selected
→ application completed
→ decision
→ approved
→ payment completed
→ order fulfilled
→ return/refund window
→ dispute window where practical
→ retained contribution
A checkout-only experiment can end too early.
The method could improve conversion today while producing costly returns tomorrow.
Finance needs the mature cohort.
Evidence that would support continued expansion
incremental completed orders;
positive retained basket growth;
positive incremental gross profit;
acceptable approval;
manageable returns/refunds;
acceptable fraud/disputes;
limited substitution;
stable operating burden;
positive incremental contribution.
Evidence that would weaken the case
high selection but little incrementality;
high financed GMV but weak gross profit;
AOV lift erased after returns;
meaningful substitution from cheaper methods;
low approval in the target customer group;
high dispute/support cost;
negative incremental contribution.
In that case, the merchant should consider:
narrowing availability;
changing placement;
targeting higher-value cohorts;
renegotiating economics;
testing another installment structure;
reducing provider dependency;
removing the method where the business case is persistently negative.
Finance diligence questions
Before approving or renewing a BNPL program, the CFO, payments team, risk team, and ecommerce organization should answer these together.
Commercial value
Which customers become newly reachable?
How many orders are genuinely incremental?
How much retained basket value is incremental?
What gross margin do those orders carry?
Does BNPL improve customer acquisition or only payment mix?
Direct cost
What percentage and fixed fees apply?
Is financing merchant-subsidized?
What additional platform or integration cost exists?
Are costs different by market or product?
Returns and refunds
How do BNPL customers' returns compare with matched cohorts?
Which provider fees remain after refunds?
How do partial refunds work?
How quickly do merchant systems reconcile adjustments?
How much support does the refund flow generate?
Risk
Which party bears customer credit risk?
Which fraud risks remain with the merchant?
Which disputes can reverse merchant funds?
What evidence must the merchant provide?
What losses and labor should finance attribute to BNPL?
Approval
What share of intended customers actually reach a financing decision?
What share are approved?
Do approval rates change by market, transaction size, and cohort?
What happens to declined customers?
Operations
Can provider settlements reconcile automatically to orders?
How are fees, refunds, and disputes represented?
How much manual work remains?
Who owns customer support?
How often does engineering need provider-specific maintenance?
Regulation
What legal entity provides the credit?
Which country's consumer-credit regime applies?
What checkout or promotional responsibilities fall on the merchant?
What changes are currently scheduled?
Who monitors them?
Strategy
What percentage of profitable orders depends on one provider?
Can another payment method solve the same affordability problem?
How would a pricing or eligibility change affect contribution?
What evidence would cause the merchant to reduce or remove BNPL?
If the business cannot answer these questions, it does not yet know the total cost of BNPL.
Final recommendation
BNPL should not be judged by the amount financed through it.
That amount is an activity metric.
Nor should it be judged by AOV alone. AOV is measured before returns, payment cost, disputes, support, and substitution have finished changing the economics.
Conversion is more meaningful, but even conversion is incomplete if financed orders replace existing successful transactions or generate insufficient margin.
The merchant decision should be made from:
Incremental retained gross profit
− incremental BNPL payment cost
− incremental operating and risk cost
− substitution cost
with regulation and provider concentration considered alongside the transaction economics.
That calculation can produce a strong positive answer.
For a high-margin merchant serving customers whose purchase is genuinely constrained by payment timing, BNPL can create commercially valuable access.
For another business, the same financing method can increase sales while compressing contribution.
The difference cannot be found in BNPL GMV.
It is found in incremental contribution margin.
I treated the uploaded brief as the controlling assignment. It explicitly makes this article the cluster’s total-downside and merchant-economics piece, with incremental contribution margin as the core output rather than financed volume, conversion, or AOV. It also requires the BNPL Total Cost Model to cover direct costs, variable operating costs, commercial effects, and strategic risks.
Risks and Costs of BNPL: What Merchants Should Measure Beyond Conversion
Editorial interpretation
Customer affordability problem: A customer may have purchasing intent and a usable payment instrument, yet find a purchase difficult or undesirable to fund in one payment. BNPL can remove that timing constraint—but only for customers who are eligible and approved.
Unique evidence anchor: A payment method can increase completed sales while reducing merchant contribution if revenue and conversion improve less than payment, return, refund, fraud, support, and operating costs increase.
Editorial position: BNPL is economically attractive only when the incremental gross profit it creates exceeds the full incremental cost of offering and operating it.
Common assumption challenged: “If BNPL raises sales and AOV, it must be good for the merchant.”
Merchant decision: Determine whether BNPL creates positive incremental contribution margin after accounting for direct payment cost, method substitution, declines, returns, refunds, disputes, fraud, operations, regulatory burden, and concentration risk.
Selected format: Merchant unit-economics model.
Structural rationale: Start with an apparently successful BNPL dashboard, then progressively subtract everything that the gross metrics omit.
Proposed architecture
Conversion can rise while profit falls
Start with incremental gross profit, not BNPL GMV
Direct BNPL costs
Returns: the cost that rewrites the original order
Refunds: where payment economics continue after the sale
Fraud risk is not credit risk
Disputes and merchant operating exposure
Customer declines reduce the reachable audience
Support, reconciliation, and settlement cost
Method substitution
Regulatory cost is becoming more material
Provider concentration and eligibility risk
Consumer risk becomes merchant risk indirectly
The BNPL Total Cost Model
Metrics finance should actually monitor
When BNPL economics fail
Finance diligence questions
Final recommendation
Original practical asset: BNPL Total Cost Model
Layer
What to measure
Merchant question
Incremental benefit
Incremental orders, retained basket growth, incremental gross profit
What would not have happened without BNPL?
Direct cost
Provider fee, fixed fee, subsidy, integration
What do we pay because BNPL exists?
Returns/refunds
Returned value, retained fees, reverse operations
What survives after the order changes?
Risk
Fraud, disputes, loss allocation
Which losses remain ours?
Operations
Support, reconciliation, engineering, finance
Which costs are hiding outside payment fees?
Approval
Eligibility, completion, approval, acceptance
Can the intended customer actually use it?
Substitution
Card/debit/wallet orders shifted into BNPL
Are we creating demand or repricing existing demand?
Strategic risk
Regulation, concentration, eligibility, reputation
How durable is the economics model?
Output
Incremental BNPL contribution
Does incremental value exceed incremental cost?
SEO metadata
H1: Risks and Costs of BNPL: What Merchants Should Measure Beyond Conversion
Meta description: Understand BNPL merchant fees, returns, refunds, fraud, disputes, approval limits, regulation, and how to calculate incremental BNPL contribution margin.
URL slug: /risks-costs-bnpl
Secondary keywords: BNPL merchant fees; BNPL risks merchants; buy now pay later costs; BNPL disadvantages merchant; BNPL fraud risk; BNPL returns; BNPL merchant economics; BNPL payment costs.
Related entities: Buy Now Pay Later; installment credit; contribution margin; merchant fees; returns; refunds; fraud; disputes; deferred payment; consumer credit; payment reconciliation; provider concentration.
Article excerpt: BNPL can increase sales and AOV while still reducing merchant contribution. The right business case measures incremental gross profit against provider fees, returns, refunds, fraud, disputes, support, substitution, regulation, and operating cost.
Suggested structured data: Article and BreadcrumbList. No FAQ schema—the article ends with finance diligence questions as required by the editorial brief.
Supporting visual: BNPL Total Cost Model.
Image alt text: “BNPL Total Cost Model showing incremental gross profit reduced by direct fees, returns, refunds, fraud, support, substitution, operations, regulation, and provider concentration.”
Internal-link plan
How BNPL Affects Checkout Conversion — Link when discussing experiments, incrementality, selection, approval, and conversion. That article should own detailed funnel experimentation.
BNPL vs Credit Cards — Link where the reader needs a deeper comparison of revolving credit, transaction-specific financing, merchant acceptance, disputes, and settlement.
Klarna for Merchants / Affirm for Merchants / Afterpay for Merchants — Link from provider-cost and refund sections when the reader needs provider-specific mechanics.
BNPL Adoption by Market — Link when market prevalence is relevant. Adoption should not be presented as evidence of merchant profitability.
BNPL for Travel / BNPL for Digital Goods — Link where cancellation, instant fulfillment, fraud, or refund structures become vertical-specific.
Authoritative source recommendations
FCA — Deferred Payment Credit regulation. Use for the current U.K. regulatory framework, which began on July 15, 2026, and for distinctions around third-party lending and regulated DPC.
ASIC — BNPL credit licensing. Use for Australia's treatment of BNPL under the National Consumer Credit Code and licensing obligations effective from June 10, 2025.
EUR-Lex — Consumer Credit Directive 2023/2225. Use for the EU's revised consumer-credit framework and November 20, 2026 application date, while checking country-level transposition before publication.
CFPB — BNPL market and consumer research. Use for U.S. pay-in-four market structure and research into persistence, loan stacking, and unsecured debt. The CFPB's December 2025 market report covers six large providers and calendar years 2019–2023.
Provider merchant agreements and current documentation. Required for actual merchant fees, settlements, refund treatment, dispute allocation, merchant eligibility, and product restrictions. Generic market averages should not substitute for contractual data.
Approved merchant analytics. Required to substantiate conversion, AOV, return, fraud, approval, substitution, and incremental contribution claims.
Proposed author and reviewers
Primary reviewer: Merchant finance lead.
Risk reviewer: Payments fraud/risk specialist.
Payments reviewer: BNPL/product or payment-method specialist.
Operations reviewer: Ecommerce operations/returns specialist.
Regulatory reviewer: Credit/payments compliance specialist if country-specific regulation remains in publication.
Do not state that review has occurred until each reviewer has actually approved the piece.
Provider fact-check items
For each provider mentioned in the final publication, verify:
current product names;
countries;
merchant eligibility;
customer eligibility;
financing structure;
merchant fee;
fixed fee;
financing subsidy;
transaction limits;
merchant settlement;
refund behavior;
partial refunds;
fee treatment after refunds;
disputes;
fraud allocation;
reporting;
reconciliation;
merchant-category restrictions;
current regulatory status.
Do not generalize one provider's terms to BNPL as a category.
Country fact-check items
For every country discussed:
applicable consumer-credit law;
latest effective date;
regulator;
merchant versus lender obligations;
advertising/disclosure requirements;
product scope;
broking/intermediation treatment;
licensing;
complaints/dispute rights;
transitional arrangements.
Current examples require special date discipline: U.K. third-party DPC regulation began July 15, 2026; Australian BNPL credit reforms apply from June 10, 2025; EU revised consumer-credit rules are due to apply from November 20, 2026.
Customer-eligibility fact-check items
For every provider/product:
age;
residence;
identity requirements;
required funding source;
transaction-value boundaries;
affordability/credit assessment requirements;
approval mechanism at a publishable level;
repayment conditions;
late fees/interest where applicable.
Do not speculate about proprietary scoring.
Merchant-eligibility fact-check items
Verify:
merchant country;
merchant category;
restricted goods/services;
transaction type;
average-ticket restrictions;
fulfillment restrictions;
onboarding requirements;
contract entity;
cross-border eligibility;
product-specific limitations.
Fee and interest fact-check items
Merchant fee: Use the merchant's signed/current commercial agreement whenever possible.
Customer interest/fees: Specify product and jurisdiction. Do not describe BNPL universally as interest-free or free.
Financing subsidy: Identify whether the provider, merchant, or customer bears the financing cost.
Refund fee treatment: Verify separately. Do not assume provider fees are returned when merchandise is refunded.
Refund/dispute fact-check items
Verify:
full refunds;
partial refunds;
repeated partial refunds;
settlement adjustments;
remaining customer installments;
refund-processing timing;
merchant fee reimbursement;
dispute submission;
response deadlines;
evidence requirements;
fraud allocation;
non-delivery cases.
Conversion and incrementality fact-check items
Require merchant-level data for:
eligible sessions;
method visibility;
selection;
application completion;
provider approval;
terms acceptance;
payment completion;
completed orders;
original AOV;
retained AOV;
return value;
incremental orders;
method substitution;
gross margin;
incremental contribution.
Never publish a conversion uplift or AOV increase without population, period, market, methodology, and denominator.
Regulatory fact-check items
Immediately before publication:
recheck FCA DPC requirements;
recheck ASIC BNPL/LCCC guidance;
check EU Member State implementation of Directive 2023/2225;
check current U.S. federal and applicable state treatment;
verify any provider authorization/licensing claim;
review checkout financing messages;
distinguish law from provider policy and merchant risk policy.
Amber-flag claims
“BNPL increases conversion.”
Requires causal merchant evidence.
“BNPL increases AOV.”
Requires merchant-specific measurement and should be paired with retained AOV.
“BNPL is more expensive than cards.”
Cannot be universalized without comparable contracts and full cost methodology.
“BNPL causes more returns.”
Requires matched cohort evidence.
“BNPL providers take all credit and fraud risk.”
Credit, fraud, dispute, and merchant operating risk are different.
“BNPL has no chargebacks.”
Do not use this as shorthand for no disputes or merchant loss.
“BNPL is interest-free.”
Product- and market-specific.
“Customers are instantly approved.”
Avoid. Approval processes and decisions vary.
“Higher financed volume means higher profit.”
The core thesis specifically rejects this.
“Provider X approval rate is better.”
Requires comparable market, product, cohort, transaction size, and risk policy.
“BNPL regulation is now harmonized.”
False framing. Major jurisdictions currently have materially different regimes and timelines.
Those sources do not, by themselves, establish:
current BNPL provider relationships;
BNPL merchant countries;
BNPL fees;
financing terms;
refund treatment;
settlement;
merchant eligibility;
performance uplift.
Current BNPL/installment providers.
Merchant countries.
Customer countries.
Provider/product names.
Merchant categories.
Integration path.
Merchant contracting entity.
Merchant pricing.
Customer financing products.
Payment confirmation.
Merchant settlement.
Refund and partial-refund support.
Dispute ownership.
Fraud allocation.
Reporting and reconciliation fields.
Current regulatory/product restrictions.
Approved conversion/AOV evidence.
Approved return/refund evidence.
Approved contribution economics.
Named internal reviewer.
Provider-specific verification requirements
For each BNPL provider considered operationally, obtain:
live merchant agreement;
pricing schedule;
product-country matrix;
settlement schedule;
refund-fee schedule;
dispute terms;
fraud allocation;
merchant restrictions;
transaction limits;
approval/eligibility documentation;
integration requirements;
reporting schema;
reconciliation process;
support escalation;
regulatory entity/license;
notice provisions for pricing or eligibility changes.
The commercial model should use those documents rather than provider marketing claims.
Cluster sameness-audit fields
Title: Risks and Costs of BNPL: What Merchants Should Measure Beyond Conversion
First two sentences: “Imagine an ecommerce team presenting its BNPL launch after one quarter. The dashboard says: Checkout conversion +8%; Average order value +15%.”
Opening device: A successful-looking dashboard whose economics deteriorate as hidden costs are added.
Customer affordability problem: Payment timing prevents or discourages an otherwise viable purchase.
Provider/market: Provider-neutral and international.
Primary format: Merchant unit-economics model.
Unique evidence anchor: Sales can rise while contribution margin falls.
Editorial position: BNPL is attractive only when incremental gross profit exceeds all incremental payment and operating costs.
Merchant decision: Retain, scale, narrow, renegotiate, or remove BNPL based on incremental contribution.
Main transaction-value context: Variable; transaction value is an input to economics rather than one global “ideal BNPL ticket.”
Original framework: BNPL Total Cost Model.
Primary metric: Incremental BNPL contribution.
Secondary metrics: retained AOV, cost per incremental completed order, return-value rate, approval rate, substitution rate, fraud/dispute loss, support cost.
Conclusion: Gross BNPL volume is activity; incremental contribution is the decision.
Closing element: Finance diligence questions rather than generic FAQ.
Cannibalization safeguards:
Buy Now, Pay Later Explained owns BNPL definitions and basic mechanics.
BNPL vs Credit Cards owns the detailed credit-model comparison.
Provider guides own specific provider mechanics.
BNPL Adoption by Market owns adoption data.
BNPL for Digital Goods owns digital-goods eligibility and risk.
BNPL for Travel owns travel-specific cancellation/refund economics.
How BNPL Affects Checkout Conversion owns detailed experiment design and conversion incrementality.
This article alone owns total merchant downside and BNPL unit economics.
Final editorial recommendation
Publish this as the cluster's CFO article, not its anti-BNPL article.
Its strongest argument is not that BNPL is costly.
Every payment method is costly.
The stronger argument is that BNPL changes enough parts of a merchant's commercial system that the processor fee alone cannot describe its economics.
The method can affect:
who buys
→ whether they are approved
→ what they buy
→ how much they buy
→ which payment method they abandon
→ what they return
→ how refunds are processed
→ what fraud/disputes remain
→ how finance reconciles the transaction
→ what regulatory obligations evolve
The resulting decision can absolutely be positive.
But the calculation must be:
incremental retained gross profit minus incremental total cost.
If that value is positive and durable, BNPL has earned its place in checkout.
If conversion and AOV rise while that value falls, the merchant has not discovered a profitable payment method.
It has discovered a more expensive way to report higher sales.
Research and current-fact review date: August 9, 2026.
Need help narrowing down payment options?
Answer a few questions about your business and compare the capabilities that matter to you.
Explore matching options ↗