Refund Economics
A framework for quantifying what each refund actually costs your P&L — and why cutting refund rate by two points can be worth more than a comparable conversion-rate lift.
Refund Economics
The full P&L cost of a refund — lost contribution margin plus return shipping, restocking, and write-offs — modelled against the order's gross revenue.
Refund economics is the CFO-facing view of returns. Each refund doesn't just reverse a sale; it destroys the contribution margin you booked, then layers on outbound and return shipping, payment-processing fees you can't recover, restocking labour, and — for apparel, beauty, and food — often the unit itself.
Framed this way, refund rate stops being a customer-service KPI and becomes a margin lever. A two-point cut in refund rate frequently moves more contribution profit than a comparable lift in conversion rate, because every avoided refund saves the full cost stack while every incremental conversion only adds one margin unit.
Most stores track refund rate as a single percentage on a monthly dashboard and stop there. That number hides the actual damage: a 12% refund rate on an apparel SKU with 55% gross margin and €8 round-trip shipping is not a 12% revenue haircut — it's closer to a 20% contribution-margin haircut once you fully load the cost.
The point of refund economics is to put a euro figure on each refund so you can compare it to the cost of preventing it. Better size charts, sharper PDP photography, fit-prediction tooling, or stricter returns windows all cost money — but so does every refund you ship, restock, and write off.
The true cost stack of a single refund
Start with the order's contribution margin — gross revenue minus COGS, payment fees, outbound shipping, and pick-pack. When that order is refunded, you lose the margin and you keep the variable costs you already spent. Outbound shipping is gone. Card processing fees on most European acquirers are non-refundable on the percentage portion. Pick-pack labour was already paid.
Then add the return-side costs: prepaid return label, inbound handling, QC inspection, and restocking. For a €60 apparel order on Shopify, that return-side stack typically runs €9–€14. Finally, factor in disposition: in apparel and beauty, 15–30% of returned units can't be resold at full price and either get discounted, liquidated, or written off.
Why a refund cut often beats a CR lift
A conversion-rate improvement adds one new order at the marginal contribution rate. A refund-rate improvement preserves an existing order at the full contribution rate AND avoids the entire return cost stack. The two levers are not symmetric — and on most apparel and beauty stores the refund lever pays out roughly 1.5–2x per percentage point moved.
Concretely: a Shopify apparel store doing €5M in annual revenue with a 22% refund rate and 50% gross margin loses around €1.1M in gross refund revenue per year, but the loaded contribution hit is closer to €1.4M once shipping, fees, and unsellable-unit write-offs are stacked in. Cutting the refund rate to 20% recovers roughly €130k of contribution — more than what a comparable +0.2pp conversion-rate lift would deliver on the same traffic.
Rule of thumb
On apparel and beauty: one percentage point of refund rate is worth roughly 1.5–2 percentage points of equivalent margin gain from a conversion-rate lift, because the avoided refund preserves margin AND eliminates return-side costs. On electronics and home, the multiplier is lower (1.2–1.5x) because resale recovery is higher.
Levers that actually move refund rate
The biggest single driver in apparel is fit — size charts, model-height annotations on PDPs, and review filters that surface fit comments. In beauty it's shade matching and ingredient sensitivity flagging. In electronics and home it's expectation alignment: dimensions in context, true-to-life colour rendering, and unboxing video on the PDP. Each of these is a CRO test in disguise, but the success metric is refund rate at 30 days, not click-through on the PDP.
Operationally, segment refund rate by SKU, channel, and discount band before you act. Heavy-discount orders refund at 1.5–3x the baseline rate on most stores, and a handful of SKUs typically drive 40–60% of the return volume. Pair this view with the Refund Rate Calculator to size the prize, and benchmark your category against the Refund Rate Benchmarks by Industry before assuming you have a problem worth solving.
Annual contribution recovered: 2pp refund cut vs 0.2pp CR lift
2pp refund rate cut
0.2pp CR lift
Refund economics FAQ
Lost contribution margin (gross revenue minus COGS, outbound shipping, payment fees, pick-pack), plus return-side costs (return label, inbound handling, QC, restocking), plus the disposition haircut on units that can't be resold at full price. On apparel that often totals 110–125% of the original order's gross revenue once write-offs are included.
Start with refund rate × revenue to get gross refund volume, then apply your blended contribution margin and add the per-return operational cost (€8–€15 for most European stores) multiplied by return count. The Refund Rate Calculator does this end-to-end if you want a quick number to take into a finance meeting.
On apparel and beauty stores with refund rates above 15% and shipping-heavy unit economics, yes — the avoided refund preserves margin and eliminates the return cost stack, so the per-point payoff is 1.5–2x a comparable CR improvement. On electronics and home with strong resale recovery, the gap narrows.
Most European acquirers refund the percentage fee on refunded transactions but keep the fixed per-transaction component, and some keep both. Treating the full processing cost as sunk is the conservative default — check your acquirer's specific refund policy, but plan for 1.5–2.5% of order value as unrecoverable.
Refunds are a direct deduction in the Net Revenue Retention DTC formula — gross revenue minus refunds, plus repeat purchases, divided by gross revenue. A store with strong repeat behaviour but a 25% refund rate often has worse NRR than a store with weaker repeat and a 10% refund rate.
Apparel runs 20–35%, beauty 8–15%, home goods 8–12%, electronics 10–18%. See Refund Rate Benchmarks by Industry for vertical-specific cuts. Compare against your category — comparing a fashion store's refund rate to an electronics store's is meaningless.
Yes, typically by 3–8 percentage points in apparel and beauty. The question is whether the conversion-rate lift from offering free returns more than compensates. For most stores under €5M, a paid or threshold-based return policy out-performs free returns on contribution margin once the loaded cost is modelled.
Always. Discount-heavy orders typically refund at 1.5–3x the baseline rate, and aggressive promo campaigns can quietly destroy contribution margin even while topline revenue looks healthy. Run the refund-rate cut by order discount % and flag any segment more than 1.5x baseline.
Treat it as a refund-rate experiment: take the per-refund cost from your refund economics model, multiply by the refunds avoided over the test window, and compare to the build cost of the asset. Most apparel stores recoup a redesigned size chart inside 60–90 days at typical traffic volumes.
Bracketing (ordering multiple sizes to keep one) is the dominant driver of apparel refund rate above 20%. Identify it via per-customer return ratio over a rolling 90-day window; the top 5% of returners typically generate 25–35% of total return volume. Policy levers — restocking fees on third+ returns, return windows — usually outperform UX fixes for this segment.
See Metricuno on your data
Bring your stack — Google Analytics, Stripe, a CRM, anything — and we'll walk through the metric tree that turns your funnel into one number.