A premium D2C skincare brand spending $2,000/day across Google + Meta thought everything was working. Their dashboard agreed. Their CFO didn't. Margins kept shrinking month-over-month while revenue grew. We rebuilt the account around Shopify-attributed reporting and surfaced a $60K/year leak hiding in plain sight.
This brand came to us spending $2,000/day across Google Ads (Search + PMax) and Meta Ads. On paper, every metric was green:
Reported ROAS: 4.2×. Monthly attributed revenue: $260K. The marketing team was celebrating. The CFO was asking why margins kept compressing despite topline growth.
"Our ROAS is up 60%. Our profit is down 12%. What's going on?" — Founder's email to me, week 1
The first thing I do in every audit is reconcile the platform-reported revenue against Shopify Reports → Sales by traffic referrer. That single comparison usually surfaces the gap. In this account, the gap wasn't subtle — it was a chasm.
Real Shopify-attributed ROAS, recalculated against actual Shopify revenue: 0.99×. The platform was over-reporting by 4×. The brand wasn't growing profitably — it was burning cash in slow motion.
Each leak is a different mechanism by which platform-reported ROAS over-states real Shopify revenue. Together they compounded a 4× misreport.
Of the "Google Ads revenue" reported, 38% came from people searching the brand name directly. Demand the brand had already earned through Meta, organic, and email — but was now paying Google to capture again.
Real non-brand ROAS: 2.1× (not 4.2×)Performance Max had no brand exclusions configured. PMax was bidding on the brand's name, winning at low CPCs, and reporting those wins as PMax conversions — making itself look like the hero campaign while stealing credit from organic and Search.
PMax true incrementality: 45% lower than reported28% of total ad spend was going to Meta retargeting campaigns reporting an 8× ROAS. Most of that revenue came from cart abandoners and repeat buyers who would have converted anyway — paying a premium to reach the warmest possible audience.
Retargeting incrementality: 30% of reportedThe brand could quote ROAS to two decimals but couldn't tell you the margin per campaign after COGS, shipping, returns, and discounts. With CM2 unknown at the campaign level, "scaling profitably" was a guess. We rebuilt the math first.
CM2 visibility installed Day 14No spend increase, no creative changes — just architecture. The first 90 days were structural; the next 6 months compounded.
Branded Search isolated into its own campaign. Non-brand Search rebuilt on phrase + exact-match. PMax restructured with brand exclusions ON and new-customer-acquisition goal turned on. Standard Shopping added at SKU-level with margin-tier asset groups.
Pixel + GA4 + Enhanced Conversions wired correctly. Server-side CAPI events for ViewContent, AddToCart, InitiateCheckout, Purchase. Monthly reconciliation pulled directly from Shopify Reports established.
Prospecting CBO with 90-day purchaser exclusions. ASC layer added. Testing CBO budget at 15% of total. Retargeting capped at 12% of total spend (down from 28%) with purchaser exclusions on every campaign.
tROAS targets recalibrated to the 19% CM2 target instead of platform default. Higher bids on high-margin SKUs (serums, retinol). Lower bids on low-margin (cleansers, samples). Custom labels for margin tiering pushed to Merchant Center.
3–5 new creative concepts tested every 2 weeks. UGC reviews graduated as the winning format (texture-focused product demos a close second). Founder-story long-form for retargeting only.
First Shopify-side ROAS lift visible at Month 4 (1.0× → 1.2×). By Month 6, 1.4×. By Month 9, 1.6× stabilized. Spend held at $2,000/day throughout — the lift was pure efficiency.
No spend increase, no creative cost increase. The full lift came from architecture, attribution, and bid calibration. CM2 doubled. Real ROAS 1.6×'d.
Note: Platform ROAS went down (4.2× → 3.8×) — most agencies would panic. But it went down because we removed the inflation, not because performance got worse. The only number that pays the bills is real Shopify ROAS — and that nearly doubled.
Blended ROAS actually went down — from 4.2× to 3.8×. Most brands would panic at that. But actual profit went up 30% in year-1 because every dollar was now working harder on real new-customer acquisition instead of paying for conversions that would have happened anyway.
1. Brand exclusions on PMax. The single most-impactful change. PMax stopped claiming credit for branded search wins, and the real incremental contribution became visible.
2. Retargeting cap at 12%. Stopped paying premium CPMs to reach an audience that was going to convert anyway. Reallocated to prospecting and PMax NCA — where the next dollar actually buys new customers.
3. Shopify-tied reporting. Once the founder + CFO could see the gap between platform-reported revenue and Shopify-attributed revenue, every spend decision improved. The right number, in the right column, every month.
Madhukar showed us a number our previous agency had been hiding from us for 18 months. We were obsessed with ROAS. He showed us that we were basically paying Google to take credit for our brand equity. Now we know which campaigns are actually driving growth.
— Head of Growth, D2C Skincare · 9 months inThe skincare brand above started with the $499 Profit Audit. live walkthrough, written 17-point plan, and a real number on what the leak is. Credited to month-1 if you sign within 14 days.