US Beauty Brand: 5.2x Meta Ads ROAS from Full Account Rebuild
A US-based DTC skincare brand spending $18K/month on Meta Ads with a blended ROAS of 1.9x, well below break-even, due to stale creative, overlapping audiences, no retargeting, and broken iOS 14 attribution.
Published: January 2026 · Timeframe: 5 months
Company Overview
- Industry
- D2C Beauty & Skincare
- Campaign Duration
- 5 months
- Primary Service
- Meta Ads Management
- Type
- Direct-to-Consumer Skincare Brand
- Status
- Active Client
The Challenge
The client came to us after 9 months of self-managed Meta Ads that had peaked at 2.4x ROAS and been declining ever since. The account had accumulated all the structural problems that compound over time in self-managed DTC accounts: stale creative, audience overlap, broken attribution, and no retargeting architecture.
- Blended ROAS of 1.9x against a 3x break-even threshold, losing money on every Meta sale at scale
- Same three static product images running for 4 months with no creative refresh or testing framework
- iOS 14 pixel changes had broken 40% of attribution, Smart Bidding was optimising on phantom data
- No CAPI (Conversions API) implementation, server-side events not being sent to Meta
- Add-to-cart abandoners never re-engaged, no Dynamic Product Ads retargeting in place
- Audiences overlapping across 6 ad sets, causing internal auction competition and inflated CPMs
Our Solution
Phase 1: Attribution Fix & Account Architecture (Weeks 1–2)
Before optimising any campaigns, we fixed the measurement foundation. Making budget decisions on broken attribution data was compounding the ROAS problem.
Conversions API Implementation
Implemented Meta CAPI alongside the browser-side pixel with server-side event deduplication using the event_id parameter. Immediately recovered attribution for 38% of mobile conversions that iOS 14 privacy changes had made invisible to the pixel. Event Match Quality score improved from 4.2 to 8.8 within 48 hours of CAPI going live.
Audience Architecture Rebuild
Audited all existing ad sets and found 6 audiences with significant overlap, each competing against the others in auction. Rebuilt the account into a clean three-stage funnel: TOF Prospecting (cold audiences), MOF Engagement (video viewers, page engagers), BOF Retargeting (website visitors, ATC, checkout abandoners). No audience overlap between stages.
Campaign Structure Rebuild
Separated Prospecting and Retargeting into distinct campaigns with different ROAS targets. Prospecting targeted 3x ROAS; Retargeting targeted 6x. This prevented the high-ROAS retargeting audience from masking poor prospecting performance in blended reporting.
Phase 2: Creative Testing Framework (Months 1–3)
Creative is the primary performance variable in Meta Ads. We built a systematic creative testing framework that consistently produced new winning ad variations and prevented the creative fatigue that had stalled performance.
Creative Testing System
Launched 4–6 new ad variations weekly across 4 creative formats: UGC testimonial videos, product demonstration videos, before-and-after static ads, and lifestyle photography carousels. Each new creative was tested against the control in a structured A/B framework with statistical significance thresholds before declaring winners.
UGC Creative Programme
Briefed 12 micro-influencers to produce authentic UGC video reviews focused on 3 key benefit claims. UGC creative consistently outperformed branded studio content, achieving 2.3x higher CTR and 40% lower cost-per-initiate-checkout than equivalent branded video.
Dynamic Product Ads
Set up Dynamic Product Ads for cart abandoners (7-day window) and product page viewers (14-day window) using the full product catalogue. DPA retargeting achieved a 7.8x ROAS, becoming the highest-performing campaign in the account.
Phase 3: Scale & Advantage+ Testing (Months 3–5)
With a proven account structure and creative system in place, we scaled spend while testing Meta's latest campaign formats for incremental gains.
Lookalike Audience Optimisation
Built 1% and 2% lookalike audiences from the top 500 purchasers by lifetime value. LTV lookalikes outperformed interest-based audiences by 60% on ROAS. Expanded to 3% and 5% lookalikes as performance proved stable.
Advantage+ Shopping Campaign Test
Ran a 4-week incrementality test, Advantage+ Shopping vs. standard Manual Sales campaign, with equal budget. ASC delivered 18% higher ROAS on equivalent spend and became the primary prospecting format from month four.
Results & Impact
Blended account ROAS up from 1.9x at engagement start, maintained above 5x for the final 8 weeks of the engagement period.
From $48 average CPP at start to $18, driven by improved attribution accuracy, creative performance, and audience quality.
Monthly revenue attributable to Meta Ads grew 4.1x over the 5-month period, with spend increasing only 40%, the remainder driven by efficiency gains.
DPA retargeting became the highest-performing campaign in the account, achieving a 7.8x ROAS on cart abandoner and product page viewer audiences.
"DigiBlazon rebuilt our Meta account from scratch and the results were transformational. Our ROAS went from 1.9x to 5.2x in five months. The creative testing system they built is something we never had before, and it's completely changed how we think about paid social."
Sophia Harrington
Head of Growth, US-Based DTC Beauty Brand
Tags
Related Case Studies
UK B2B SaaS Platform: 5.4x Google Ads ROAS with Full Account Rebuild
US Home & Lifestyle Brand: 5.2x ROAS Scaling to $120K Monthly Ad Spend
Ready to Achieve Similar Results?
Book a free audit and we'll walk through your current marketing programme — what's working, what isn't, and where the biggest opportunities are.
Request Your Free Audit
Usually $1,500 value, No commitment, delivered in 48 hours.
By submitting, you agree to receive insights and consultation. We never share your information.