Refund Rate by Acquisition Channel
Refund rates on paid social regularly run 2-3x higher than paid search or email. Here's why discount-led Meta cohorts return more — and how to bake channel quality into your scaling decisions.
Quick answer
Refund rate typically runs 2-3x higher on paid social (Meta, TikTok) than on paid search or email. The driver is intent: discount-led, impulse-triggered social cohorts return more than buyers who searched for your product by name. Track refund rate alongside RPV per channel before you decide where to scale spend.
Refund Rate by Acquisition Channel
The percentage of orders refunded, segmented by the marketing channel that acquired the customer — a channel-quality signal that sits beside RPV.
Refund rate by acquisition channel measures returns and refunds split by the source that drove the order: paid social, paid search, email, organic, affiliate, and so on. The same store can see a 12% refund rate on Meta-acquired orders and a 4% rate on branded search in the same week — not because the products differ, but because the intent of the buyer does.
Performance Managers use the metric to discount the gross ROAS of impulse-heavy channels and to redirect budget toward sources that bring buyers who keep what they ordered.
Most stores look at refund rate as a single store-wide number. That hides where the returns are actually coming from.
Once you split refunds by acquisition channel, the picture flips: the channel you thought was your best ROAS performer may be your worst margin performer after returns clear.
Why Meta ads return more than search
Paid social interrupts. A shopper scrolling Reels sees an apparel ad, taps, and checks out in 90 seconds — often pushed by a 20% off code or a free-shipping threshold. They did not wake up wanting that hoodie.
Paid search is the opposite. Someone typed your brand or product category into Google, compared options, and clicked. They arrived with intent already formed, which is why their refund rate is consistently lower.
The intent gradient
Roughly: branded search < non-branded search < email to existing buyers < retargeting < prospecting paid social < influencer/affiliate discount codes. Refund rate climbs as you move down that list. Sizing-sensitive verticals (apparel, footwear) amplify the spread; consumables (beauty refills, supplements) compress it.
How to detect channel-driven refund skew
You need three things stitched together: the order, the refund event, and the first-touch (or last-touch) acquisition channel. On Shopify, this means joining order data with refunds and your attribution source — UTMs, Shopify's attribution, or a server-side feed.
Use a 30-60 day refund window. Most returns land in the first 30 days post-delivery, but apparel and footwear can stretch to 60. Anything shorter understates the channels with longer return tails.
Pair the result with RPV by traffic source. A channel with high RPV but high refund rate has inflated upfront numbers; a channel with modest RPV but low refunds may be the quiet winner once you net it out.
Typical refund rate ranges by channel
Refund rate ranges by acquisition channel, apparel & accessories DTC (€1M-€15M stores)
| Channel | Typical refund rate | Discount-heavy cohort | Drivers |
|---|---|---|---|
| Branded search | 3-5% | 4-6% | High intent, knows the brand |
| Non-branded search | 5-8% | 7-10% | Category intent, less brand fit |
| Email (existing buyers) | 4-7% | 6-9% | Familiar with sizing & quality |
| Meta retargeting | 7-11% | 10-14% | Warm but discount-pushed |
| Meta prospecting | 10-15% | 14-20% | Impulse, first-time buyers |
| TikTok prospecting | 12-18% | 16-22% | Strongest impulse skew |
| Influencer / affiliate codes | 11-17% | 15-22% | Code-led, low brand affinity |
Beauty and electronics compress these ranges by 30-50% because sizing isn't a factor. Footwear inflates them. If your store is mixed-category, segment the table by SKU type before you act on it.
How to fix it (without killing growth)
First, recalculate channel ROAS net of refunds. If Meta prospecting shows 2.4x gross ROAS at 14% refund rate, the net is closer to 2.1x. Compare that to search at 3.1x gross and 4% refunds — the gap widens, not narrows.
Second, attack the cohort drivers, not the channel itself. Add a sizing guide above the add-to-cart on social-traffic landing pages, gate the 20% code to email signup so buyers slow down, or A/B test a fit-quiz pre-checkout. These reduce refund rate without cutting volume.
Experiments worth running
Test a 48-hour cooling-off email to Meta prospecting buyers — a soft "here's how to style it / how others sized it" message before shipment. Stores running this see refund rate drop 1-3 points without affecting reorder rate.
Test removing the largest discount from prospecting creative for two weeks. Volume usually dips 10-20%, but if refund rate falls more than that, net contribution goes up. This is the kind of trade you can only see once you've split refunds by channel.
Frequently asked questions
Meta interrupts a scroll with an ad and a discount; search captures someone who already wanted the product. The intent gap is the single biggest driver of refund-rate variance between channels — typically 2-3x in apparel, 1.5-2x in beauty.
No — recalculate ROAS net of refunds and compare. Paid social often still pays back at scale, just at a lower effective return than the gross number suggests. The decision should be net contribution, not refund rate alone.
Use the same attribution window as your acquisition reporting (typically 7-day click for Meta, 30-day for search) and a 30-60 day refund window post-delivery. Shorter windows understate apparel and footwear channels.
In our benchmark range, yes — TikTok prospecting tends to run 2-4 points higher than Meta prospecting because the impulse skew is stronger and the audience skews younger and more price-sensitive.
RPV by traffic source tells you how much revenue each visitor generates upfront. Refund rate by channel tells you how much of that revenue stays. You need both: a high-RPV channel with high refunds can be worse than a low-RPV channel with low refunds.
Yes. Code-led traffic behaves more like prospecting social than like email — refund rates of 11-17% are typical. Group it with paid social for net-margin analysis, not with organic.
Use first-touch for diagnosing acquisition quality (which channel brought this customer) and last-touch for tactical optimization (which campaign closed the order). For refund rate, first-touch usually tells the cleaner story.
Segment refund rate by channel AND by SKU category. If one channel skews toward a single high-return product, the channel itself may be fine — your creative is just pushing the wrong SKU.
For mixed-category apparel stores in the €1M-€15M band, a blended 6-9% post-optimization is realistic. Beauty stores should target 2-4%; electronics 4-6%. Anything above 12% blended suggests a product or sizing problem, not just a channel mix issue.
Yes — run the calculator per channel using each channel's orders and refund count, then weight by spend to get a net-of-refunds ROAS view. That's the number to make budget decisions on.
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.