Refund Rate Impact on Break-Even ROAS
Refunds eat contribution margin and add return-shipping cost — both lift the break-even ROAS your paid campaigns must clear. Here's the corrected formula and the DTC examples.
Quick answer
A 5-point rise in refund rate typically lifts your break-even ROAS by 15-25%. The fix is to divide your standard break-even ROAS by (1 − refund rate) and add return-shipping cost back into the numerator. Bidding against the pre-refund floor is why so many paid campaigns look profitable on the dashboard and lose money in the P&L.
Refund Rate Impact on Break-Even ROAS
The amount your break-even ROAS rises once refunds and return-shipping costs are subtracted from gross contribution margin.
Break-even ROAS is the revenue multiple a paid campaign must hit to cover variable costs. The standard formula uses gross margin — but every refund returns the revenue while keeping most of the cost (pick, pack, outbound shipping, payment fees, and usually return shipping too). That means your effective margin is lower than your gross margin, and your true break-even ROAS is higher than the number on your dashboard.
For apparel and beauty stores running 15-30% refund rates, the gap between reported and refund-adjusted break-even ROAS can be 0.4-0.8x — enough to turn a 'profitable' paid social campaign into a loss.
Most Performance Managers calculate break-even ROAS as 1 ÷ contribution margin. On a 40% margin product, that's a 2.5x floor. Then they bid Meta and Google campaigns against it, hit 2.8x blended ROAS, and call the quarter a win.
Why the standard break-even ROAS formula misses refunds
The standard formula treats every sale as final. In reality, on a 20% refund rate, one in five orders comes back. The revenue reverses, but the variable costs mostly don't.
Outbound shipping is gone. Payment processing fees are partially refunded but the fixed component stays. Pick-pack labor is sunk. And return shipping — if you offer free returns, which most apparel and beauty stores do — is a fresh cost the original order never carried.
The compounding cost
A refunded €80 apparel order doesn't just remove €80 of revenue. It also burns roughly €6 outbound shipping, €6 return shipping, €1.50 in non-refundable payment fees, and €3 of pick-pack — about €16 of pure loss on top of the reversed margin. That's why refund rate hits the ROAS floor harder than people expect.
The refund-adjusted break-even ROAS formula
The corrected formula is: refund-adjusted break-even ROAS = 1 ÷ (contribution_margin × (1 − refund_rate) − return_cost_per_order ÷ AOV). The (1 − refund_rate) term shrinks the effective margin pool. The return_cost_per_order term handles the asymmetric shipping/handling cost on returned orders.
Apply it to a beauty SKU with 40% gross margin, €50 AOV, 12% refund rate, and €4 return cost per refunded order. Effective margin = 0.40 × 0.88 − (0.12 × 4 ÷ 50) = 0.352 − 0.0096 = 0.342. Break-even ROAS = 1 ÷ 0.342 = 2.92x. Versus 2.50x naive — a 17% higher floor.
How to detect that refunds are breaking your ROAS targets
Three signals show up before the P&L does. First: blended ROAS sits above your target but contribution margin trends flat or down month-over-month. Second: paid social campaigns with shorter purchase windows (impulse buys) refund faster than search. Third: new-customer cohorts refund 1.5-2x more than repeat buyers, so scaling acquisition lifts the average.
The cleanest detection is a refund-adjusted ROAS column in your campaign reporting. Pull refunds by order, attribute them back to the acquiring campaign with a 30-60 day lag, and recompute. Campaigns ranked by gross ROAS rarely match campaigns ranked by refund-adjusted ROAS — the reorder is where the money is.
Refund-rate ranges that change the math
Electronics: 8-15%. Beauty and skincare: 5-12%. Apparel: 20-35% (size and fit returns dominate). Home and furniture: 10-18%. Anything above 15% means the refund-adjusted floor is materially different from the naive one — and you should be bidding against the adjusted number, not the dashboard number.
How to fix it and test against the right floor
Three operational fixes. One: rebuild your target ROAS in the ad platforms using the refund-adjusted formula above — not the gross-margin shortcut. Two: segment refund rate by product category and run separate ROAS floors per catalog feed in Advantage+ or Performance Max. Three: treat return-shipping cost as a campaign-level variable, not a fixed cost — paid social with impulse-heavy AOVs needs a higher floor than branded search.
On the experimentation side, test charged-return policies on high-refund categories (apparel sizes outside your core range), test size guides and fit-finder widgets on PDPs to attack the cause, and A/B test post-purchase upsells that lift AOV without lifting refund rate. Each of these moves the floor down rather than asking your media buying to clear a higher one.
Refund rate and break-even ROAS, answered
On a 40% margin product, going from 10% to 15% refund rate lifts the break-even ROAS from roughly 2.78x to 2.94x — about a 6% increase. From 20% to 25% it jumps from 3.13x to 3.33x. The higher your baseline refund rate, the more each additional point hurts because you're dividing by a smaller effective-margin pool.
Contribution margin — revenue minus all variable costs (COGS, payment fees, outbound shipping, pick-pack, payment processor). Gross margin only nets out COGS and overstates your true profitability floor. The whole point of break-even ROAS is to capture every variable cost the marginal sale carries.
Only the portion you absorb. If you charge a flat €4.95 return fee that fully covers your return shipping cost, you can drop the return_cost term. If you charge €4.95 but actual return shipping is €7, include the €2.05 delta. Most apparel and beauty stores still subsidise returns and need the full term in the formula.
Standard break-even ROAS is 1 ÷ contribution margin and assumes no refunds. The refund-adjusted version multiplies the margin by (1 − refund rate) and subtracts per-order return costs. For a sub-5% refund rate the two values are within 5% of each other; above 15% refund rate the gap is material and you should switch.
No. Both platforms report revenue at the purchase event, before any refund. Even if you send refund events back, they don't subtract from the historical ROAS the optimizer uses for bidding. You need to bid against a refund-adjusted target ROAS that's higher than your gross break-even, then let the platform optimize toward that ceiling.
Use a 60-day window for apparel and beauty (most returns land in 14-30 days but a long tail goes to 60). Electronics and home are usually 30-45 days. Anything beyond 90 days is rare enough to ignore for media-buying decisions, though it still matters for accounting.
Yes — this is one of the highest-leverage fixes. A category with 25% refund rate needs a meaningfully higher ROAS floor than one with 8%. Split your shopping feeds, set per-feed target ROAS in Advantage+ Shopping or Performance Max, and stop letting low-refund categories subsidize high-refund ones in the blended number.
Treat them as fractional refunds. If 5% of your orders get a 30% partial refund (damage, missing item discount), that's equivalent to a 1.5-point full-refund equivalent. Add it to your refund_rate input. Most order management systems can export partial-refund value alongside full refunds for this.
Yes, through the return-shipping term. On a €30 AOV beauty order, €6 of return shipping is 20% of the order — devastating. On a €200 AOV outerwear order, €6 of return shipping is 3% — much more manageable. Low-AOV stores with high refund rates have the worst-bent break-even ROAS curves.
Take your current gross break-even ROAS and divide by (1 − refund_rate). On a 2.5x floor with 20% refund rate, that's 2.5 ÷ 0.80 = 3.125x as a fast approximation. It ignores return-shipping cost but gets you 80% of the answer in 10 seconds — useful for sanity-checking a campaign that looks too good.
Track CAC, channels, and funnel conversion in one place
Metricuno connects ad spend, funnel events, and revenue so you can see CAC by channel, cohort, and campaign — without stitching together five tools.