Every NC ROAS target from Liam stays as the floor. On top of it, this plan sets Spend, NC Revenue, NC Orders and subscription targets per brand and campaign type through December 31, plus the actions that get us there, with Black Friday and the New Year peak planned in advance.
Grey bars are what Google Ads delivered in 2026. From September 15 each bar stacks the base scaling of what already runs, plus what each action adds. The dashed line is today's run-rate (last 30 days), so the gap above it is the growth.
Left: monthly spend. Right: portfolio NC ROAS against the spend-weighted floor from Liam's targets. The plan keeps NC ROAS above the floor every month; spend is added where there is measured room.
NC revenue per day, weekly average. Base scaling alone, and the full plan with every action and both seasonal pushes.
Ledisa is the only brand that ran Google in Q4 2025 (Cleantra's first new-customer revenue is January 2026), so it is the only evidence we have. It changes where the push goes.
NC revenue each action adds on top of base scaling, per month. Each one has a gate: it only scales if the gate holds. Task IDs are in ClickUp.
The growth plan (base scaling plus actions), by month. NC ROAS is the planned figure. Tests (Demand Gen, new PMax in its learning weeks) start below their floor on purpose and scale only past their gate; each line's floor is in the campaign-type view.
NC revenue per period. Conservative assumes a soft November and December (NC ROAS x0.80) and no actions. Base scaling assumes x0.90. The growth plan is base scaling plus the actions above.
It spent $0 on Sep 14. It is Ledisa's most efficient line: $76k spend and $156k NC revenue in the last 30 days at 2.05x. Base scaling assumes it is back by October; every week it stays off costs about $36k of NC revenue. Task 1243w4q4ugv.
Cleantra non-brand (1.10 vs 1.34) and Ledisa non-brand (1.15 vs 1.40) are 57% of spend. Their spend stays flat until NC ROAS reaches the floor, and the advertorial and exact-match work is aimed at them first. More budget alone would push them further down.
That is the model in Aaron's scaling report. Last Click 7d reads about 30% lower (2x lower on Ledisa Search). Which window the floor uses is open with Gianfranco (86cbgx1c0); if it changes, the targets move with it.
Ledisa in Q4 2025. Treat the BFCM and New Year numbers as planning assumptions, check them weekly, and cut back to the floor if November softens the way it did last year.
Baseline: Aug 15 – Sep 13, 2026. $1,026,774 spend, $1,429,864 NC revenue, 16,770 NC orders, 12,917 of them subscriptions, NC ROAS 1.39 (Demand Gen excluded). Spend comes from Google Ads, joined in Triple Whale on campaign_id; August matched within 1.5%.
Floor: Liam's NC ROAS targets (Aug 21): Search 1.4, PMax 1.7 and DGEN 1.0 for video-ad products (Lymphatic, Ledisa GLP-1, CELMOR Body Oil, Nervalley Magnesium Cream, BlueForce Follifusion), Search 1.3 and PMax 1.6 for the rest, weighted by spend per brand and campaign type.
Base scaling: types above target grow 15% a month where budget is the wall, 10% where demand or rank is, and 20% for Search Brand through bids and impression share. Diminishing returns are modeled from Google's own bid simulation. Types below target hold spend until they reach the floor.
Actions: sized conservatively from measured baselines: Ledisa's brand line, 12% of non-brand spend for new PMax (Cleantra runs 17%), +10% conversion rate for the winning advertorial, and a Display-to-Search shift on generic PMax. Each is a planning assumption with a gate, not a forecast.
Subscriptions: Triple Whale pixel_orders_table, new-customer orders flagged as subscription (72% to 92% by brand). Orders are revenue divided by each line's NC AOV.
Excluded: GL-ME (closed Aug 21) and non-brand accounts. Known data issue: Celmor's Last Click NC revenue is 3% to 5% of Triple Attribution on every campaign. Full model and CSVs: reports/acentecom/targets/.