Buy Now, Pay Later: Impact of Installment Payments on Customer Purchases
Maesen · Ang — Journal of Marketing 89(3), 13–35, 2024 (accepted manuscript via White Rose)
Thesis
The Di Maggio result isn't a banking-data quirk — it replicates inside a single retailer. When a large US retailer switched on BNPL installments, its own customers started buying more often and spending more per period. The mechanism the authors test: installments shrink the perceived payment, easing felt financial constraints even when real budgets didn't change. Pay-in-4 reframes $60 as "$15 today," and reframing is enough to move behavior.
Method
- Study 1 (field): customer-level transaction data from a large US retailer, before/after it introduced BNPL; difference-in-differences comparing adopters to matched non-adopters, with parallel-trends checks and extensive robustness (control-group definitions, alternative outcomes, COVID sensitivity).
- Studies 2–4 (pre-registered experiments): manipulate installment framing in purchase scenarios; measure perceived financial constraints and spending choices to isolate the mechanism (temporal reframing → lower perceived constraint → more spending).
Key findings
- BNPL adoption at the retailer → purchase incidence up ≈9 percentage points and purchase amounts up ≈10% relative (β = .100, p < .01). Statistically and economically significant; robust to the specification battery.
- The spending lift is not confined to BNPL purchases — adopters' overall purchase behavior rises, matching the flypaper pattern in transaction-data studies.
- Experiments trace the channel: installment options reduce perceived financial constraints, which increases purchase intentions — the effect attenuates when the full payment total is made salient.
- Effects are stronger for customers with tighter budgets — the people most attracted to installments are the ones most moved by them.
Limitations
- One retailer, one BNPL provider, online-heavy sample — external validity to other verticals unproven.
- Adoption is voluntary; selection is handled by matching/DiD, but unobserved taste-for-spending shocks can't be fully ruled out (the paper stress-tests this).
- Journal of Marketing paper — the experiments use purchase-intention scenarios, not real money at real checkout.
- Study window overlaps COVID-era e-commerce distortions (dedicated robustness section, but still).
So what for Nick
- Retail-side confirmation: the same checkout button that shows up in bank data shows up in the merchant's till — incidence AND basket size. "It's just a payment method" is now hard to defend.
- The mechanism is reframing, and reframing is beatable: the effect weakens when the total price is salient. Read the full amount, then decide.
- For any budget conversation with the kids: the installment option reliably raises how often you buy, before it raises what you pay — frequency is the silent leak.