How Returns and Refunds Erase ROAS Gains in Apparel DTC
Platform-reported ROAS ignores returns entirely. For apparel stores running 25-40% return rates, that gap quietly erases most of the campaign profit you think you're booking.
Quick answer
Meta and Google report ROAS on gross attributed revenue at checkout. They never subtract the 25-40% of apparel orders that get returned, the reverse-logistics cost, or the restocking write-down. A reported 4.0x ROAS on a dress collection with a 35% return rate and €6 of reverse logistics per unit usually nets out to roughly 2.4x — and that's the number your P&L actually sees.
Net ROAS after returns (apparel)
Attributed ad revenue minus refunds, exchange shrinkage, and reverse-logistics cost, divided by ad spend.
Net ROAS is the version of return on ad spend that survives contact with your warehouse. Where platform ROAS counts every checkout as revenue, net ROAS subtracts the orders that come back: full refunds, the gap between what an exchange ships out and what comes back as resaleable inventory, and the per-unit cost of inspecting, refurbishing, and restocking returned apparel.
For categories with structurally high return rates — denim, dresses, footwear, outerwear — net ROAS is typically 35-45% lower than reported ROAS. Treating the two as interchangeable is the single most common reason apparel stores hit their ROAS target and still lose money.
This page is the operational answer to one question: how much of your reported apparel ROAS is fictional, and how do you measure the real number weekly without waiting 60 days for return windows to close?
Why apparel ROAS overstates so badly
Ad platforms fire the purchase event when checkout completes. From their point of view a sale is a sale. Your warehouse disagrees three weeks later when 35% of the dresses come back in a polybag.
Apparel return rates sit in a different universe to other DTC verticals. Beauty SKUs return at 4-8%, electronics around 8-12%, and homeware around 10-15%. Apparel averages 25-30%, with dresses and footwear regularly hitting 35-45% on new-customer cohorts.
The driver is bracketing — shoppers buy three sizes intending to keep one. That behaviour is rational for the customer and catastrophic for your reported ROAS, because the platform counts all three units as attributed revenue while two are already destined for the returns bin.
The return-lag trap
Most apparel returns land 14-35 days after purchase. A Monday-morning weekly ROAS report on last week's campaigns therefore shows a number with effectively zero returns netted out — the worst possible version of the metric. By the time the returns catch up, you've already scaled the campaign two more times.
How to detect the gap on your own data
Pull a 90-day cohort, not a weekly one. Take all orders attributed to paid social or paid search in a single completed month, then wait until that month's 60-day return window has closed before you score it.
Compute three numbers per channel: gross attributed revenue (what Meta/Google reported), net retained revenue (gross minus refund value minus exchange shrinkage), and net retained margin (net revenue minus reverse-logistics cost per returned unit). Divide each by spend. The gap between line one and line three is the lie you've been scaling against.
Reported vs net ROAS by apparel category (illustrative, €60 AOV, €6 reverse-logistics cost per returned unit)
| Category | Return rate | Reported ROAS | Net ROAS (revenue) | Net ROAS (margin) |
|---|---|---|---|---|
| Denim | 28% | 4.0x | 2.88x | 2.60x |
| Dresses | 38% | 4.0x | 2.48x | 2.10x |
| Footwear | 35% | 4.0x | 2.60x | 2.25x |
| Outerwear | 22% | 4.0x | 3.12x | 2.88x |
| T-shirts & basics | 12% | 4.0x | 3.52x | 3.35x |
Segment the cohort by acquisition source while you're at it. TikTok-acquired apparel buyers consistently return at higher rates than Google Shopping buyers — often 8-12 percentage points worse — because TikTok drives more impulse and trend-led purchases that fail the at-home try-on.
How to fix it without throttling growth
Fix one: stop bidding to reported ROAS. Set a return-adjusted ROAS floor in your Meta and Google bid strategies. If your true breakeven is 2.4x net, and your category runs a 35% return rate, your reported-ROAS target needs to be roughly 3.7x — not 2.4x.
Fix two: weight channels by net contribution, not gross. A channel delivering 3.5x reported ROAS at a 15% return rate beats one delivering 4.5x at 40%. Most paid-media dashboards still rank by the wrong column.
Fix three: separate cash refunds from exchange credit in your reporting. An exchange that ships a replacement size keeps the revenue in the building; a cash refund moves money back out. Treating them the same overstates the damage on well-run exchange flows and understates it on cash-heavy ones.
Fix four: attack the upstream cause. Size guides with fit-finder quizzes, model height/weight stats on every PDP, and post-purchase fit-confirmation emails all measurably reduce size-driven returns. A two-point return-rate drop on a dress collection moves net ROAS more than any bid optimisation.
What "good" looks like
Best-in-class apparel teams reconcile net ROAS by channel and category on a 30-day rolling cohort, set bid floors against the net number, and report both figures side-by-side in the weekly trading meeting. The gap between reported and net becomes a managed KPI, not an accounting surprise at month-end.
Experiment ideas to test this month
Run a PDP fit-finder quiz on your three highest-return SKUs and measure return rate on quiz-completers vs non-completers across a four-week purchase window. A 3-5 point return-rate reduction is a realistic win and translates directly into 10-15% net ROAS uplift on those SKUs.
Test exchange-first return flows: default the returns portal to "swap for another size" rather than "refund". The conversion from refund-intent to exchange-completion is typically 20-35%, and every percentage point retained is revenue your campaigns get to keep crediting.
Frequently asked questions
Use 25-30% as a blended starting point, then adjust by category: 12-15% for basics and t-shirts, 22-28% for denim and outerwear, 30-40% for dresses and footwear. Override with your own once you have 90 days of post-return-window cohort data.
Net ROAS = (gross attributed revenue − refund value − exchange shrinkage − reverse-logistics cost) ÷ ad spend. The most rigorous version also subtracts payment-processing fees on refunded orders, which Stripe and Adyen do not refund on the return leg.
It's the only number available in near-real-time for bid optimisation. Use reported ROAS for in-flight campaign decisions, but anchor your targets to a return-adjusted floor so the optimiser is chasing a goal that maps to actual profit.
Yes, materially. Discovery channels (TikTok, Meta Reels, Pinterest) drive higher-impulse purchases with weaker fit intent; demand-capture channels (Google Shopping, branded search) attract buyers who already know the brand's fit. The gap is commonly 8-12 points on apparel.
Weekly for directional reporting, monthly cohort for bid-strategy decisions. Weekly reports on last week's spend will undercount returns by 70-80% because the lag is structural — most apparel returns arrive 14-35 days after purchase.
An exchange keeps revenue in the business and only costs you the reverse-logistics leg plus inventory friction. A cash refund removes the revenue entirely. Same return event, very different ROAS impact — track them as separate line items.
In the EU, expect €4-€8 per unit for return shipping, inspection, and repackaging, plus 10-20% inventory write-down on items that come back damaged or out-of-season. Premium brands with white-glove returns can hit €12-€15 per unit.
Yes — paid returns reduce return rates by 4-8 points but also reduce conversion rate by 3-6 points. Most apparel stores find net contribution roughly flat, so the decision is about brand positioning, not maths. Test it on a single category before rolling out.
Returns are one of the three main mechanisms — alongside COGS and shipping subsidies — that turn a healthy reported ROAS into a negative ROI. On apparel specifically, returns are usually the largest of the three. Reconcile all three monthly to keep the gap visible.
Set a return-adjusted target ROAS floor in your bid strategies for the three highest-return categories. It takes an afternoon, doesn't touch the site, and immediately stops the optimiser from scaling spend against revenue you'll refund in three weeks.
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.