by Trivas.aiROAS is revenue divided by ad spend. Spend $1,000, make $3,000 back, that's a 3x. Simple math, except beauty brands keep asking us for "the number" as if one blended figure could tell them anything useful.
It can't. A single blended ROAS mashes together your branded search, your cold Meta prospecting, and your retargeting into one average that hides more than it reveals.
Here's the general range worth anchoring to for the ROAS benchmark for beauty DTC brands 2025: 1.8x to 2.5x blended counts as average, 3x or higher is strong, and anything under 1.5x is a flag to go dig into what's broken. Beauty runs lower than most verticals because paid social is saturated and influencer/UGC-style ad creative isn't cheap to produce or test at volume. Apparel and home goods brands often post higher blended numbers just because their acquisition costs haven't been bid up the same way.
The rest of this piece breaks these numbers down by channel, AOV tier, and business model. Blended ROAS is a starting point, not a verdict.
Channel-level numbers tell a very different story than the blended one.
Meta
Google Shopping/Search
TikTok
The mistake we see constantly: brands compare a 4x on Meta retargeting to a 2x on TikTok prospecting and conclude TikTok is underperforming. It's not underperforming, it's doing a different job. Retargeting ROAS should always look better than prospecting ROAS, full stop. If your Meta prospecting campaigns are pulling 4x, either your funnel stages are mislabeled or something unusual is happening with attribution windows.
The same logic applies across TikTok and Google Ads: pull the numbers apart by funnel stage before you compare anything across platforms.
Not all beauty brands should be chasing the same number, and treating them as if they should is where a lot of benchmark confusion starts.
Subscription and replenishment brands, think skincare regimens or haircare systems people reorder every 60 to 90 days, can tolerate a lower first-order ROAS. If a customer's lifetime value over 3 to 6 months makes the math work, a 1.5x on the initial purchase might be perfectly healthy.
One-and-done products, gift sets, trend-driven items, seasonal color launches, don't get that luxury. There's no repeat purchase to lean on, so the first sale has to carry its own weight. These brands need a higher first-purchase ROAS or they're just buying revenue at a loss.
AOV tier changes the math again. Sub-$30 AOV brands often need 2.5x or higher just to clear COGS and fulfillment costs after the ad spend is accounted for. A $60+ AOV brand can run profitably at 1.8x because there's more margin dollar per order to absorb the acquisition cost.
One more thing worth naming directly: if you're reading ROAS straight off Meta Ads Manager or Google Ads, you're probably looking at an inflated number. Platform-reported ROAS uses attribution windows and click models that favor the platform. Blended MER (marketing efficiency ratio, total spend against total revenue) is almost always lower, and it's the number that actually reflects reality.
The brands hitting the top of the ROAS benchmark for beauty DTC brands 2025 range aren't doing anything mysterious. They're just measuring more precisely.
First, they break ROAS down by SKU, not just by campaign. Hero products routinely carry 2 to 3x the ROAS of the rest of the catalog. If you're only looking at campaign-level numbers, a strong hero SKU can mask a weak long tail, and you'll never know which products are actually earning their ad budget.
Second, they separate ROAS for first-time customers from returning customers. Blending the two flatters the number, because returning customers convert on far less ad spend. A brand that looks like it's running a healthy 2.8x might actually be running a 1.6x on new customer acquisition once returning buyers are pulled out.
Third, they reconcile ad-platform ROAS against actual Shopify or GA4 revenue weekly, not monthly. Attribution windows shift, discount codes get double-counted, and a month-end reconciliation means you're finding problems 30 days late instead of 7.
Fourth, and this one's specific to beauty: they account for returns and refunds. Color-match issues and allergic reactions drive return rates that other verticals just don't deal with at the same scale. A reported 3x ROAS that turns into a 2.4x after refunds isn't a rounding error, it's a real gap between what the platform tells you and what actually landed in the bank.
Most miscalculations aren't math errors, they're definition errors.
The biggest one: treating platform-attributed ROAS and blended MER as interchangeable in board reporting. They measure different things and mixing them up in a deck is how founders end up defending a number that was never real to begin with.
Second: ignoring the effect of discounts, bundles, and gift-with-purchase offers on true revenue per order. If half your orders shipped with a free deluxe sample and a 15% code, your revenue per order dropped, but if your ROAS math still uses list price, you're overstating performance.
Third: not isolating brand search spend from non-brand search spend on Google. Branded search ROAS is almost always inflated, because that demand was coming to you anyway. Blending it with non-brand prospecting numbers makes your Google performance look better than your actual acquisition engine.
Fourth: benchmarking against a generic ecommerce average instead of a beauty-specific, AOV-adjusted number. A $25 AOV lip gloss brand and a $90 AOV skincare set brand shouldn't be judged against the same target, and neither should be judged against an all-industry average that includes electronics and furniture.
Start with blended ROAS: total ad spend across every channel against total store revenue for the same period. That's your sanity check before you go any deeper.
From there, segment two ways. First, new versus returning customer, so you can see your true acquisition efficiency instead of a number propped up by repeat buyers. Second, your top 10 SKUs versus the rest of the catalog, to see where you're over- or under-indexing.
If you want to run the actual math instead of eyeballing it, the ROAS calculator takes real spend and revenue inputs and gives you the number without a spreadsheet.
Do this monthly. Daily platform swings are noise, mostly driven by attribution windows still settling, and reacting to them is how budgets get yanked around for no good reason.
Here's the core problem: Meta, Google, and TikTok each report their own version of ROAS, and each one is optimized to look good in its own dashboard. Reconciling those against actual store revenue by hand, in a spreadsheet, eats hours every week, and it's usually the first task that gets skipped when things get busy.
A unified dashboard that pulls Meta, Google, TikTok, and Shopify or GA4 into one place removes that guesswork. Instead of three tabs telling you three different stories, you get one source of truth for what actually happened. Trivas's BI reporting is built around exactly this: one place where platform-reported numbers get checked against real revenue automatically.
For beauty brands running heavy paid social spend, channel-specific views matter too. You want to catch it fast when one platform's reported ROAS starts drifting from what actually landed in revenue, not find out at month-end close.
If you're trying to figure out where you actually sit against the ROAS benchmark for beauty DTC brands 2025, and want that comparison built into your everyday reporting instead of a one-off spreadsheet exercise, it's worth starting a trial or talking to a founder directly about what your numbers should look like.
explore more insights
3 min read
3 min read
3 min read
.d53b12e5.png&w=3840&q=75 "Trivas.ai")