A D2C supplements brand at $220K/mo Shopify revenue was over-spending — running blind on platform-attributed dashboards, ramping budget every time CAC looked stable, and burning $16K/mo of waste. We pulled spend back, rebuilt structure, and let efficiency compound. Real ROAS started lifting at month 4. Stabilized by month 12.
This brand had crossed $200K/mo Shopify revenue and was scaling Meta and Google together. The agency before me had one playbook: see stable platform CAC, increase budget 30%, repeat. Topline grew. Profit didn't.
The founder reached out after a quarterly P&L review. CFO wanted to know why blended margin was down 6 percentage points despite a 40% revenue lift over 6 months.
We were scaling Meta CBO at +30% per week. The platform said we were profitable. The bank account said we weren't. I needed someone to tell me which number was right. — Founder, week 1
Reconciling against Shopify Reports surfaced the gap: Meta retargeting ROAS was reporting 6.2× — real Shopify-attributed contribution was 1.8×. Prospecting was paying to "acquire" customers who'd already bought (no purchaser exclusions). PMax was bidding on the brand. Search was disorganized.
The fix wasn't to scale faster. It was to pull spend back, rebuild structure, and let real efficiency compound. That's exactly what we did over 12 months.
Four separate leaks, all over-attributing platform performance, all under-counting how much the brand was actually paying per real new customer.
Retargeting CBO was getting 31% of total Meta budget — reporting 6.2× ROAS. But after applying purchaser exclusions and reconciling against Shopify Reports, real incremental contribution was 1.8×. Most of those "conversions" were repeat buyers and warm cart abandoners.
Recovery: $9K/mo redirected to prospectingMeta prospecting CBO was running broad with no 90-day purchaser exclusion. The algorithm was finding past buyers and counting them as new conversions. Typical for accounts that hit $200K/mo without ever resetting fundamentals.
Recovery: 22% of prospecting waste eliminatedBrand exclusions off. PMax was bidding on the supplement brand name across Google and Shopping, taking credit for branded conversions that would have happened organically.
PMax true incrementality: 50% of reportedGoogle Ads tROAS was set to 4.0 — based on platform-reported revenue. Recalibrated to the 26% CM2 target, the real tROAS needed was 3.2 (against Shopify revenue). Account had been over-bidding on the wrong number for months.
Bid efficiency: +18% on non-brandFirst action wasn't optimization — it was removing waste. Spend dropped before structure was rebuilt. Then 11 months of patient compounding.
Retargeting reduced from 31% to 12% of Meta spend. Purchaser exclusions added across all retargeting and prospecting. Immediate $9K/mo recovery — reallocated to Meta prospecting and Google non-brand Search.
Brand exclusions configured on every PMax campaign. Asset groups split by margin tier (high-margin formulas vs commodity protein). Custom labels in Merchant Center for margin signal.
Bid strategy targets recalibrated to 3.2 tROAS (against real Shopify revenue, not platform-attributed). Max CPC caps added to prevent runaway costs during scaling phases.
3–5 new concepts every 2 weeks. Founder-story long-form was the surprise winner for prospecting. UGC review format graduated to scale at month 3. Lifestyle-only creative killed.
Real Shopify ROAS started climbing at month 4 (1.4× → 1.6×). CAC dropped to $58. CM2 climbed to 22%. We held spend flat to let the structure prove itself before scaling.
By month 9, structure was solid. We added 15% to non-brand Search (the highest-ROI layer) and 10% to PMax NCA. Spend trended down further as efficiency compounded — total spend at month 12 was 38% below month 0.
Total ad spend dropped 38% over 12 months. CAC dropped 32%. CM2 climbed 7 percentage points. MER moved from 1.9 to 2.4 — Shopify-attributed.
Takeaway: Topline grew 17% on 38% less spend. The growth came from real new customers, not from over-attributed retargeting. CM2 went from 19% to 26% — that's the entire game.
The default agency response to "stable CAC" is "scale spend." But platform-reported CAC is often wrong by 30–80% in D2C accounts that haven't reconciled tracking against Shopify. Scaling on the wrong number compounds the wrong way.
Months 1–3: Cleanup phase. Retargeting cap, purchaser exclusions, PMax brand exclusions, tROAS recalibration. Numbers move sideways while the algorithm re-learns. This is the phase where most agencies panic and revert.
Month 4: First credible Shopify-side ROAS lift. The algorithm has stable signal, the architecture is doing its job, and the brand sees the first real efficiency improvement.
Months 5–12: Compound phase. Each month of clean structure makes next month a little better. Creative library grows. CM2 climbs. Spend gradually drops while topline holds and grows.
Anyone promising material 30-day improvements in this kind of work is reading you a sales script. The founder's quote sums it up better than I could:
Took 4 months to see ROAS turn. Took 12 to compound. Worth the wait. Madhukar was the only person who told me to spend less, not more. Everyone else wanted me to scale. He was right.
— Founder, D2C Supplements · 12 months inThe supplements brand above started with the $499 Profit Audit. live walkthrough across Google + Meta, written 17-point plan, real number on the leak. Credited to month-1 if you sign within 14 days.