A strong week under the surface: brand's share of demand reached an all-time high, YouTube pre-roll was upgraded to Scale, and Belgium keeps compounding quietly. Two things need attention — [BRAND_NL_Q3] is underperforming on reach and CPC, putting the NL awareness goal at risk, and TikTok's uncertainty widened enough to blur its verdict. The Meta holdout reads out Thursday; hold any budget shifts until then.
Composite of branded search, share of search, direct traffic, mentions & recall proxies across Powr and Wondr — revenue-weighted, benchmarked against each brand's category in BeNe.
TOFU channels — modeled contribution
| Channel | Spend / mo | Modeled revenue contribution | Contribution index | Effect lag | Verdict |
|---|---|---|---|---|---|
Meta — awareness VIDEO VIEWS · REACH |
€14,200 | €39,800 ±7.1k | 5–7 wks | Scale | |
YouTube pre-roll SKIPPABLE · 15S |
€8,600 | €19,400 ±5.9k | 6–8 wks | Scale | |
TikTok — top view CREATOR · SPARK ADS |
€6,100 | €9,200 ±4.8k | 3–5 wks | Hold | |
Podcast sponsorships HOST-READ · NL |
€4,300 | €5,100 ±3.6k | 8–10 wks | Watch |
BOFU — what brand does for performance
| Channel | Spend / mo | CPA | Brand tailwind | Strength |
|---|---|---|---|---|
Branded search GOOGLE · EXACT + PHRASE |
€3,800 | €14 | −41% ±8%CAT −38% | Strong |
Meta retargeting DPA · ENGAGED AUDIENCES |
€9,200 | €22 | −12% ±5%CAT −15% | Moderate |
Google Search — non-brand GENERIC · CATEGORY TERMS |
€11,400 | €31 | −3% ±4%CAT −8% | Weak — within noise |
Performance Max SHOPPING + DISPLAY |
€7,600 | €27 | −2% ±5%CAT −6% | Weak — within noise |
Brand's tailwind on non-brand BOFU is currently weak — within the model's noise floor. That is not a verdict on brand; it's a reach measurement. Your awareness converts existing demand more cheaply (branded search −41%, retargeting −12%) but isn't yet wide enough to pre-warm cold traffic. This is what "TOFU needs to improve" looks like in the data — the model expects the weak rows to move 6–8 weeks after awareness reach expands, and Powr NL is the first place to look.
The echo — what today's awareness buys later
Your April awareness push of €22,900 is still generating revenue today. Performance stops when you stop paying. This doesn't.
Brand ROI — decay per cohort, compounding in total
Each month's spend decays — read any row left to right and the effect fades. But cohorts stack: consistent investment compounds, which is why the total keeps climbing while every individual row falls. Stop investing, and the stack unwinds.
| Cohort | Jan | Feb | Mar | Apr | May | Jun |
|---|
12% of GA4 sessions untagged — fix raises attribution coverage; TikTok history is only 9 months, so its confidence intervals stay wide until Q4.Fed into the Strategist's first plans
META BRAND CONTRIBUTION
COEFFICIENT CI: ±34% → ±14%
REALLOCATED TO TOFU
Data sources
Context layer
Stream BrandStock out
get_brand_score — score, components & historyget_demand_decomposition — baseline / brand / performanceget_cohort_roi — per-cohort decay curvessimulate_budget_shift — what-if scenarioswrite_media_plan — create plans externallyYour modeled data, wherever your team works — pull the brand score into a Looker board, run budget simulations from Slack, or let your own agents query the MMM directly. Read-only by default; write scopes are opt-in per tool.