by Trivas.aiAsk a skincare founder what "good ROAS" looks like and you'll get a different answer than you'd get from someone selling protein powder or phone cases. Beauty just behaves differently. Repeat purchase rates run higher than almost any other DTC vertical, which means blended ROAS climbs well past what paid media alone is actually generating.
That gap is where most brands get confused. They see a 4x blended ROAS in their monthly P&L, then check Meta Ads Manager and see 1.9x, and panic. Neither number is wrong. They're just measuring different things. Blended ROAS covers all revenue against all ad spend. MER (marketing efficiency ratio) is often used interchangeably but calculated slightly differently depending on who you ask. Platform-reported ROAS only counts what that platform thinks it caused, using its own attribution window and its own rules.
So before you benchmark anything, know which number you're holding. For a ROAS benchmark for beauty DTC brands 2025 to mean anything, you need to compare like against like.
Here's the range worth memorizing: most healthy beauty DTC brands land between 2.5x and 4.5x blended ROAS. Paid-social-only ROAS, measured properly across channels rather than pulled from a single dashboard, typically runs 1.5x to 3x. If you're outside those bands in either direction, that's worth digging into, not ignoring.
Channel performance varies more in beauty than category-level numbers suggest. Breaking it down:
One thing worth saying plainly: platform dashboards inflate their own ROAS. Each platform claims credit generously under last-click or platform-attributed models, and if you add up what Meta, Google, and TikTok each claim, the total will exceed what actually happened in your bank account. You need cross-channel reporting to see the real number, not four separate "my ROAS is best" claims from four different ad managers. This is exactly the blind spot brands running Meta, TikTok, and Google Ads simultaneously run into: each platform looks great in isolation, and the combined math never adds up the way any single dashboard promised.
Category matters almost as much as channel. A skincare brand and a fragrance brand can run identical campaigns and land on wildly different ROAS, purely because of how the product itself gets bought and reordered.
Skincare typically runs 3x to 4.5x. Subscription and replenishment models drive this. Once someone's serum runs out in six weeks, they're back, and that repeat revenue keeps blended ROAS strong long after the initial ad spend.
Color cosmetics and makeup sit at 2x to 3.5x, and it's a more volatile band. Trend-driven demand means a viral shade or format can spike sales fast, then drop just as fast. Higher return rates (wrong shade, doesn't match skin tone) also eat into net numbers after the fact.
Haircare lands at 2.5x to 4x, helped by bundling. A shampoo-conditioner-mask set has a much higher AOV than any single item, and that multi-SKU cart math is often the difference between a 2.5x and a 4x on the same ad spend.
Fragrance is the choppiest, 1.5x to 3x. It's gift-driven and heavily seasonal (hello, Q4), so a brand's blended ROAS in November can look nothing like its ROAS in March. Don't benchmark fragrance brands against an annual average. Look month over month against the same month last year instead.
A few structural shifts are pushing these ranges around this year, and none of them are temporary blips.
CAC keeps climbing on Meta and TikTok. Auction costs are up, and brands are responding by leaning harder on retention, email, and SMS to keep blended ROAS respectable even as paid acquisition gets more expensive per new customer.
UGC and creator content are lowering CPMs, but there's a catch. Cheaper impressions only help if you're feeding the algorithm enough creative variety to keep performance from decaying. Brands running 3-4 ad concepts a month are seeing worse ROAS than brands running 15-20, even at similar spend levels.
iOS and privacy changes are still distorting attribution. Brands relying solely on platform-reported data are commonly underestimating true ROAS by 15-30%, because conversions that happened days after a click or view simply never make it back into the ad platform's counting.
Subscriptions and bundles have become the main lever. Beauty brands pushing subscribe-and-save skincare or bundled haircare sets are the ones consistently landing above category average, because they're converting a single transaction into recurring revenue against a one-time acquisition cost.
Blended ROAS is simple in theory: total revenue divided by total ad spend, across every channel, not per platform. The math breaks down the moment you calculate it separately for each channel and add the results, because that double-counts customers who saw a TikTok ad, then converted from a retargeting email, then got attributed by Meta anyway.
Get this right by pulling first-party revenue data first. Shopify orders, GA4 sessions, and each ad platform's spend numbers all need to sit in one unified source before you compare anything to a benchmark. If you're pulling Meta's number, Google's number, and your Shopify total into three separate tabs and eyeballing them, you're not calculating blended ROAS. You're guessing.
If you want to see exactly where your numbers land against the ranges above, plug them into the ROAS calculator rather than estimating by feel.
One more thing worth flagging: don't compare MER to ROAS and expect them to match. They use different denominators.
Metric
ROAS
MER
What it measures
Revenue generated per dollar of ad spend
Total revenue against total marketing spend
Formula
Ad Revenue / Ad Spend
Total Revenue / Total Marketing Spend
Typical use
Channel or campaign level
Whole-business efficiency
They answer different questions. Treating them as interchangeable is how brands convince themselves they're underperforming (or overperforming) when they're really just misreading the math.
Most of the ROAS "problems" brands bring to us aren't spend problems. They're measurement problems hiding as spend problems.
Attribution gaps between what Meta or TikTok reports and what actually lands in Shopify are the single biggest cause of a ROAS mismatch. The platform says one thing, the bank account says another, and nobody's reconciled the two.
Overspending on prospecting is common when a brand's real efficiency is coming from retargeting and email flows, not new-customer acquisition. If retargeting is running 4x and prospecting is running 1.6x, and budget keeps flowing 70/30 toward prospecting, blended ROAS suffers for no good reason.
Ignoring returns quietly deflates net ROAS after the sale closes. Color cosmetics especially: a campaign can look like a 3x win on the day of purchase and drop to 2.4x once shade-mismatch returns come back three weeks later.
Manual spreadsheet reporting delays all of this. If it takes three days to pull last week's numbers together, budget keeps flowing to the underperforming campaign for three more days than it should.
The fix for most of the above isn't a new ad strategy. It's a single source of truth that pulls Shopify, Meta, Google, and TikTok data into one place, so you're not reconciling four dashboards by hand every Monday.
Trivas built its reporting layer on Amazon Redshift specifically to handle this kind of cross-channel reconciliation: platform-reported ROAS gets matched against actual store revenue, so you're looking at one number instead of guessing which of four is closest to real. It's the kind of infrastructure that matters more once you're spending across three or four channels and can't afford to trust any single platform's self-reported math. For teams evaluating this kind of setup, marketing leaders are usually the ones who feel this pain first, since they're the ones presenting the "real" number upward every month.
Quick recap: healthy beauty DTC brands are landing at 2.5x to 4.5x blended ROAS in 2025, with paid-social-only performance running 1.5x to 3x. Skincare and haircare tend to sit at the higher end thanks to repeat purchases and bundling. Color cosmetics and fragrance run more volatile, driven by trends and seasonality rather than steady repurchase behavior.
Before you decide you're under- or over-performing against any of this, run your own numbers through the ROAS calculator and see where you actually land, not where a single ad platform's dashboard says you land.
And if you're tired of stitching together Shopify, Meta, Google, and TikTok numbers by hand every week just to answer "is our ROAS actually good," it might be worth seeing how a unified BI reporting setup handles that reconciliation for you. Either way, subscribe to keep an eye on where these benchmarks move as the year plays out. They shift more than most category reports let on.
explore more insights
3 min read
3 min read
3 min read
.d53b12e5.png&w=3840&q=75 "Trivas.ai")