- Blended — why ROAS doubled:
ROAS 1.36x → 2.75x is an accounting artifact, not a performance gain. Spend fell $1,458 WoW and $1,448 of that (99%) is Demand Gen ($1,358 → $69) plus PMax ($159 → $0) switching off — both carry 0 last-click orders, so removing them mechanically lifted the blend. The honest read is below the ROAS line: revenue −20% ($3,273 → $2,619) and orders −23% (52 → 40), with spend at 29% of the $3,289 plan and revenue at 31% of the $8,486 plan. The same mix shift explains CTR −54% / CPC +54% (cheap Discovery clicks left the blend) and the +198% jumps in ATC and CVR (DG traffic that never added to cart is gone) — the site did not get better.
- Search Brand:
Softened on more spend: 30 orders (was 36, −17%) and $1,895 revenue (−23%) on $439 (+16.5%), ROAS 6.54x → 4.31x. CPC_Core did most of the damage — 25 orders / $324 at 5.18x vs 31 / $304 at 7.12x — while Brand IS actually rose to 91.9% (was 90.3%), so the extra spend bought coverage, not conversions: this is demand softness, not a rank problem. Products is the clear leak, spend +61% to $108 for one fewer order at 1.58x (was 4.52x). Modifiers is immaterial at $7 but returned its first order of the month at 6.51x.
- Search Non Brand (Reviews):
The week's only genuine win. 5 orders (was 3, +67%) on 15% less spend ($185 vs $218), revenue $261 vs $120, ROAS 0.55x → 1.41x and CPA $37 against a $90 NB target. IS climbed to 76.5% (was 74.3%), leaving roughly a quarter of the auction unclaimed — this is the one unit with proven headroom to fund right now.
- Shopping Brand:
The real drag. Orders collapsed to 5 from 13 (−62%) on nearly flat spend ($258 vs $298), revenue $463 vs $686 and ROAS 2.30x → 1.79x, with CPA at $51.59 against the $12 Shopping Brand target. It is a single campaign, SHP_TSD_US_PPT_ACQ_VOLUME_CBO_WK45, so the cause is contained: check whether the WK45 promo / CBO window closed and whether the top SKUs are still serving in the feed. At $92.56 AOV these are the account's most valuable orders — each one lost is worth about 1.4 Search Brand orders.
- Pmax Non Brand (Serum):
Dark this week: $0 spend and 0 orders, after $159 and 0 last-click orders the week before. There is no performance to read here — the open question is whether it was deliberately paused or ran out of budget/eligibility. Confirm status before next week: with PMax at zero, Reviews is the account's only remaining non-brand prospecting.
- Demand Gen (Discovery):
Cut 95% to $69 from $1,358, with 0 last-click orders in both weeks. 1dc is blind to Discovery's upper-funnel contribution, so this is not proof it failed — judge it on GA4 assisted conversions, and note that Brand Search softened (30 orders vs 36) in the very week the DG budget disappeared. The decision needs to be explicit: fund it against an assisted read, or reallocate the ~$1,300/week it freed. Today that money is simply not being spent, which is why the account sits at 29% of plan.