Returns-Adjusted ROAS: Why Your Holiday ROAS Will Shrink in January
Returns adjusted ROAS is ad-attributed revenue minus the refunds on those same orders, divided by ad spend. It is the number your December ROAS turns into once January returns come in. Most holiday reports never get revised, so teams plan Q1 off a figure that was never real. If you sell apparel, the gap can be big enough to change which channel you think won the season.
Why does holiday ROAS shrink after December?
Holiday ROAS shrinks because the revenue is booked at purchase and the returns arrive weeks later. Extended holiday return windows make that lag longer. Amazon, for example, let most items bought between November 1 and December 31, 2025 be returned until January 31, 2026, and plenty of DTC brands copy that policy to stay competitive.
The volume is large. The National Retail Federation estimated that 17% of holiday sales would be returned in 2025, and 19.3% of online sales across the year. Your December dashboard counts every one of those orders at full value.
Ad platforms do not go back and fix it either. Meta's reported purchase value is based on the order at the time it was placed. As one 2026 analysis put it, when an order is later returned, Meta does not reduce its reported conversion value in your historical campaign data. Neither does a December screenshot in your Q4 recap deck.
Why apparel feels it hardest
Apparel has the highest return rates in ecommerce, mostly because of fit. A June 2026 survey of online shoppers found clothing made up 38.3% of all reported online returns, driven by sizing and shoppers ordering multiple sizes of the same item. Holiday adds gift returns on top: a sweater that fits fine still goes back if the recipient does not like the color.
How to calculate returns adjusted ROAS
The formula is simple. The work is in matching refunds to the orders that ads actually drove.
Net revenue ROAS goes one step further and uses net revenue: gross revenue minus discounts and refunds. Use whichever matches how your finance team defines revenue, and stay consistent.
The key detail is order-level matching. Taking your store-wide return rate and applying it to every channel is better than nothing, but it hides the thing you most need to see: return rates differ by channel, campaign, and product. If you calculate your true ROAS without that, you will undercount returns on some channels and overcount them on others.
An apparel example: three channels in December
Here is an illustrative apparel brand with $200,000 in December ad spend. The return rates are hypothetical, but the pattern is common: discovery traffic and discount-driven orders tend to come back more often than high-intent search orders.
- Meta: $100,000 spend, $380,000 attributed revenue, 3.8x ROAS. 24% of that revenue refunded by January 31. Returns adjusted ROAS: 2.89x.
- Google: $60,000 spend, $300,000 attributed revenue, 5.0x ROAS. 14% refunded. Returns adjusted ROAS: 4.30x.
- TikTok: $40,000 spend, $140,000 attributed revenue, 3.5x ROAS. 31% refunded. Returns adjusted ROAS: 2.42x.
Blended, the account goes from 4.1x in the December report to about 3.2x by February. That alone changes the Q1 conversation. More important, the gaps between channels move. Meta and TikTok looked close in December, 3.8x versus 3.5x, but TikTok's net revenue fell by nearly a third, and Google's lead over both widened. If you were about to shift budget from Meta to TikTok because of a 3.5x December number, the returns data says to wait.
Your numbers will be different. That is the point: you will not know your ecommerce return rate by channel until you measure it.
What causes return rates to differ by channel?
A few patterns show up often enough to check in your own data:
- Intent. Search shoppers usually know what they want. Social shoppers are more often buying on impulse.
- Offer depth. Heavier discounts pull in buyers who are more willing to order two sizes and return one.
- Product mix. If one channel sells more fitted tops and another sells more accessories, their return rates will differ for reasons that have nothing to do with the channel.
- Gift share. Campaigns built around gifting carry gift returns into January.
Product mix is the one people miss. Before you decide a channel brings in bad customers, check which products it sold. Sometimes the fix is the product set in the campaign, not the budget.
AdBeacon ties every attributed order to its channel, campaign, ad, and products, so refunds can be matched to the ads that drove them.
What to do before and after the holidays
Before BFCM 2026 (Black Friday is November 27, Cyber Monday November 30):
- Pull last year's refunds by order and match them to the channel and campaign that drove each order.
- Set holiday ROAS targets on a returns-adjusted basis, so a 3.5x target on a 25% return channel is understood as roughly 2.6x net.
- Agree with finance on the date you will restate holiday performance. February 15 works for most brands with a January 31 return deadline.
After the return window closes:
- Restate December ROAS by channel on net revenue and send the updated numbers to whoever approved the budget.
- Compare returns adjusted ROAS to platform-reported ROAS. The gap often widens, because platforms already include view-through conversions on top of not netting out refunds.
- Carry the adjusted numbers into Q1 budget decisions, not the December screenshot.
This is where first-party, order-level attribution earns its keep. AdBeacon attributes orders using click-only, first-party data from your Shopify, BigCommerce, or WooCommerce store, so every attributed order can be tied back to the channel, campaign, and ad that drove it, and the products in it. That makes it practical to match refunds to campaigns instead of guessing with a store-wide average. AdBeacon also shows its own ROAS next to platform-reported ROAS, plus POAS, CAC, and new vs returning customers, so the restated view sits beside what each platform claimed. Look at more than one attribution model while you are at it. Last click and first click often tell different stories about which channel's customers keep what they buy.
How to report the restated numbers without losing trust
Nobody enjoys sending a February email that says December was smaller than everyone thought. The way to make it routine is to set the expectation in November. Report holiday ROAS twice: a preliminary number in early December, labeled as gross, and a final returns adjusted number after the return window closes. When leadership or a client knows the second number is coming, the drop reads as process, not as a miss.
Agencies have an extra reason to do this. If your client's finance team restates revenue in February and you are still quoting December ROAS, the gap lands on you. Getting ahead of it with a restated channel view is a simple way to look like the partner who knows the numbers better than the client does.
Two formatting tips that make the restatement easier to read:
- Show gross ROAS, returns adjusted ROAS, and the refund rate side by side for each channel, so the reader sees why each one moved.
- Call out the biggest change explicitly. "TikTok lost the most to returns" is the line people will remember.
The takeaway
December ROAS is a draft. The final version arrives in February, after returns, and for apparel brands it can be 20% or more lower, with very different gaps between channels. If you want to see attributed orders, products, and ROAS by channel from your own first-party data before you lock Q1 budgets, book an AdBeacon demo.
Frequently Asked Questions
What is returns adjusted ROAS?
Returns adjusted ROAS is attributed revenue minus refunds on those same orders, divided by ad spend. It shows what a campaign earned after customers sent products back.
Do Meta and Google subtract returns from reported ROAS?
Not by default. Reported conversion value reflects the order at purchase, and later refunds do not reduce historical reported revenue.
How much do holiday returns reduce ROAS?
It depends on your return rate. NRF expected 17% of 2025 holiday sales to be returned, and apparel brands often run higher, so a 4.0x December ROAS can land closer to 3.0x after returns.
When should I recalculate holiday ROAS?
Recalculate after your holiday return window closes, typically early to mid February for a January 31 deadline. That is when most holiday returns have landed.
What is net revenue ROAS?
Net revenue ROAS divides net revenue (gross revenue minus discounts and refunds) by ad spend. It is a stricter version of returns adjusted ROAS that also accounts for discounts.