Last updated: April 2026
Paid social for luxury is a different animal than paid social for DTC or SaaS. The platforms are the same, the ad formats are the same, and the bidding mechanics are the same, but the economics, the creative standards, and the measurement requirements are fundamentally different. Most luxury brands don't account for those differences, and the result is a paid social program that burns budget without building anything durable.
We've audited paid social accounts for over 40 luxury brands at Deus since 2022. The same five mistakes show up in nearly every one. Not occasionally. Nearly every one.
This is the most expensive mistake on the list and the hardest to see from inside an Ads Manager dashboard. Platform-reported ROAS includes revenue from people who would have bought anyway: brand loyalists who clicked a retargeting ad on their way to the checkout they were already heading toward.
Across 12 luxury accounts where we ran incrementality lift studies in 2024 and 2025, platform-reported Meta ROAS overstated the true incremental number by an average of 42%. That means a reported 4.5x ROAS was actually closer to 2.6x in real incremental value. For a brand spending €500K per quarter on Meta, the gap between those two numbers represents roughly €190K in misattributed revenue.
The fix isn't complicated, but it does require discipline. Run geo-holdout or conversion-lift tests quarterly. Measure incrementality by campaign type, not just blended. And stop treating platform-reported ROAS as gospel when you're making budget allocation decisions.
Luxury brands produce beautiful brand films, editorial-style imagery, and cinematic product content. Then they run it on a purchase-optimised campaign with a CPA target. The algorithm dutifully serves that creative to the tiny slice of the audience most likely to convert today, which is almost never the audience the creative was designed to move.
Brand content and conversion content have different jobs. Brand content builds salience, desire, and consideration over weeks or months. Conversion content closes a sale in the moment. When you run a 60-second brand film against a purchase objective, Meta's algorithm will throttle delivery because the creative doesn't generate enough immediate conversions to justify the spend. You end up with low reach, low frequency, and a cost-per-view that makes the CFO flinch.
The correct structure is a two-layer media plan. Awareness and consideration campaigns run brand creative on reach, video view, or ThruPlay objectives. Conversion campaigns run product-specific creative on purchase or add-to-cart objectives. The two layers talk to each other through audience sequencing, not through a single campaign trying to do both jobs at once.
Luxury brands have smaller site traffic than mass-market DTC brands. A fashion house doing €15M in annual DTC revenue might get 80,000 to 120,000 unique visitors per month. When you carve that pool into retargeting segments (viewed product, added to cart, visited three-plus times), the audiences get tiny fast.
The problem is that Meta and Google will happily spend your retargeting budget against a 3,000-person audience all month long. The frequency climbs past 15, 20, sometimes 30 impressions per user per month. At that point you're not retargeting. You're stalking. And your brand, which is supposed to feel exclusive and considered, starts feeling desperate.
Deus client data suggests the frequency ceiling for luxury retargeting sits between 8 and 12 impressions per user per 30-day window. Beyond that, incremental conversion rates drop to near zero and brand perception scores start declining. The fix is to cap frequency, broaden your prospecting pool to feed the retargeting funnel, and accept that a small retargeting audience means a small retargeting budget. Not every channel needs to scale.
Luxury creative takes longer to produce than DTC performance creative. A brand shoot can take weeks of planning, a day or two of production, and another week of post. That means most luxury brands refresh their ad creative two to four times per year. Performance brands refresh monthly or more.
The mismatch creates a predictable cycle. A new campaign launches, performance is strong for four to six weeks, then CTR declines, CPA rises, and the team wonders what changed. Nothing changed in the market. The creative just fatigued.
For luxury, creative fatigue tends to set in after 20,000 to 30,000 impressions per unique user per ad variation on Meta, and slightly higher on Google Display. The solution isn't to churn out low-quality variations. It's to shoot more assets during each production cycle (cover shots, detail shots, styling alternatives, behind-the-scenes footage) and build a library that supports 8 to 12 distinct ad variations per quarter rather than 2 to 3.
Google captures existing demand. Someone searches "buy Bottega Veneta Jodie bag" and you bid on the keyword. The intent is already there. Meta creates demand. Someone is scrolling their feed, not shopping, and your creative needs to interrupt them with something beautiful or compelling enough to plant a seed.
The strategic difference is enormous, but many luxury brands plan and evaluate both channels using the same framework. They set the same ROAS target for Meta and Google, allocate budget based on last-click performance, and then wonder why Meta "doesn't work" while Google "crushes it."
Google will almost always show higher last-click ROAS than Meta for luxury. That's structural, not a reflection of channel quality. Google is closing sales that Meta (and organic, and PR, and brand marketing) opened. A brand that cuts Meta spend to shift budget to Google will typically see Google ROAS decline within two to three months, because the demand Meta was generating dries up.
The correct framework is to evaluate Google on efficiency (ROAS, CPA) and Meta on incremental contribution (lift in total revenue, new customer acquisition rate, brand search volume). They're different tools. Measure them differently.
All five mistakes share a common root: applying mass-market DTC playbooks to a luxury context. The budget sizes are different, the audience sizes are different, the creative production cycles are different, and the purchase consideration windows are different. A framework built for a €40 AOV skincare brand selling 50,000 units per month has no business governing a €2,400 AOV jewellery brand selling 200 pieces per month.
The structural fix is to build a paid social operating model that reflects luxury economics from the start. That means incrementality testing baked into the quarterly calendar, campaign structures split by objective, frequency caps enforced at the ad set level, creative libraries sized to support 90-day refresh cycles, and channel-specific KPIs that don't pretend Meta and Google do the same thing.
Most of this is setup work. Once the operating model is in place, the ongoing management is actually simpler than the chaotic approach most luxury brands default to.
For the benchmark ranges these recommendations sit within, see Luxury Marketing Benchmarks 2026. For the broader AI visibility context, see The Luxury Brand Entity Stack.
