Paid Media · 5-Day Performance & Scaling Review

Czarina — where to spend the next dollar

17 – 21 April 2026 · AEDT · AU market
Prepared for Michael (Meta) & Steven (Google Ads)
$3,771
Paid Media Spend
$8,055
Attributed Revenue
2.14×
Blended ROAS
+$20K
Identified 14-Day Lift
Meta ROAS figures use Meta’s official purchase_roas metric (1d-click / 1d-view attribution) to match the in-platform Ads Manager view.
Google figures from Google Ads API v21. Full methodology on final page.
Executive Summary

Revised honest read

After re-baselining Meta against the in-platform purchase_roas metric, the numbers are more sober but the opportunities are sharper. Google is the clearest scale story. Meta has pockets of strong performance buried under mis-allocated budget.

Methodology correction: This report’s Meta figures use purchase_roas (Meta’s default Ads Manager ROAS field) on 1-day-click / 1-day-view attribution. An earlier version of this brief used summed omni_purchase action values which inflated results by counting cross-day attribution twice. Michael flagged the discrepancy; this is the corrected view.
Google

Clean feed, clean signal

After the GMC cleanup (10,529 → 27 product issues, 99.7% reduction), the new PMAX campaign hit 8.71× ROAS on Apr 21 at only 15% impression share. 85% of eligible demand was lost to rank — pure budget headroom.

Total Google: $451 spend → $1,827 revenue at 4.05× ROAS.

Meta

Misallocated, not broken

One adset returned 11.45× on $131. Five adsets burned $935 at 0.47× ROAS. The flagship Purchase adset at 2.22× ROAS accounts for 64% of Meta spend and is approaching creative fatigue (freq 3.2).

Total Meta AU: $3,320 spend → $6,228 revenue at 1.88× ROAS.

Scaling Opportunities

Four evidence-based plays

Revised after re-pulling Meta data against the Ads Manager purchase_roas baseline (the numbers Michael sees in-platform). Conservative lift projections assume ROAS decay at scale.

Google Priority 1 · High confidence

Scale the new PMAX — 85% of demand is still unclaimed

The PMAX campaign relaunched Apr 20 on the cleaned GMC feed hit 8.71× ROAS on Apr 21 at 90% budget utilisation but only 15% impression share. 85% of eligible shopping demand was lost to rank. This is a clean scaling signal — limited only by budget.

Apr 21 ROAS
8.71×
Imp Share
15%
Rank-Lost IS
85%
Daily Cap
$150
Actual Spend
$135
Recommended action
  • Week 1: Raise daily budget to $300/day (2×).
  • Week 2: If ROAS holds above 5×, push to $500/day.
  • Ceiling signal: watch for Imp Share > 40% or ROAS decay below 4× — that's the saturation point.
  • Configure listing group filter: custom_label_0 = in_stock_or_oversell (the 504-product advertisable set).
+$10,000 – $15,000 incremental revenue over 14 days
at 5× ROAS (conservative) on ~$2,500 incremental spend
Meta Priority 1 · High confidence

Fund the 11.45× adset — it's currently under-budgeted 4×

The ADV+ | Max Conv Value | AU adset inside SI · ADV+ · Catalogue + Sales · AUS Apr26 returned 11.45× ROAS on just $131.12 over 5 days (~$26/day). Frequency is 1.6 — plenty of headroom before fatigue. Reach of 972 is tiny against the 17,946 reach of the main Purchase adset.

ROAS
11.45×
Spend (5d)
$131
Revenue
$1,502
Frequency
1.6
Reach
972
Recommended action
  • Immediate: Raise daily budget from ~$26 to $100/day (4×).
  • Week 2: If ROAS holds above 6×, push to $200/day.
  • Monitor frequency weekly — refresh creative once it passes 2.5.
  • Funding: redirect from Opp 3 below — zero incremental spend required.
+$6,000 – $10,000 incremental revenue over 14 days
at 5–7× ROAS (assuming decay from 11.45× as we scale) on ~$1,200 incremental spend
Meta Redeploy

Stop the bleed — 5 adsets returning below 1× ROAS

Across 5 adsets, $946 was spent over 5 days at sub-1× ROAS — a net loss against store margin. Redeploying this budget to Opp 2 at even a conservative 5× would generate ~$4,730 incremental revenue.

AdsetCampaignSpendPurchValueROASAction
ADV+ Prospecting | 23–64 | Women ADV+ · Sales · AUS Apr26 $502 1 $318 0.63× Cut 50%
LAL AU 3% ATC & Purch 90D · 55–65 DABA · AUS Apr26 $205 1 $119 0.58× Cut 100%
Social Engagers · 55–65+ Dynamic RM · AUS Apr26 $107 0 $0 0.00× Cut 100%
Open_AU_25+_Sale DD_Meta_Traffic_AU_TOF $99 0 $0 0.00× Cut 100%
Website Visits 180D · 18–50 MOFU · Sales · AUS Apr26 $22 0 $0 0.00× Cut 100%
Redeployable $935 2 $437 0.47× → Opp 2
Net effect: zero incremental spend, revenue swings from $437 (current) to ~$4,700 (at 5× when redeployed) — a +$4,300 swing over the next 5 days.
Meta Priority 2 · Protect

Refresh the $2,135 flagship adset — frequency is at 3.2

Purchase | AU in SI · DD · Adv+ Catalogue Conv PUR is the workhorse: $2,135 spend, 19 purchases, $4,743 revenue at 2.22× ROAS. Frequency at 3.2 — approaching the 3.5 fatigue threshold. This is 64% of total Meta AU spend; ROAS decay here is costly.

Spend
$2,135
Purch
19
ROAS
2.22×
Frequency
3.2 ⚠
Reach
17,946
Recommended action
  • This week: Upload 3 new creative variants ahead of Mother's Day window (peak May 11).
  • Champion to beat: current best-performing DPA carousel.
  • Test UGC, new product hero imagery, and refreshed offer copy in parallel.
  • Goal: pull frequency back under 2.5 while lifting ROAS above 3×.
Protect: preventing ROAS decay from 2.22× → 1.8× over 2 weeks saves ~$1,800 at current spend levels.
Google Ads · Data

5-day Google Ads performance

Apr 21 highlighted — first full day on the cleaned GMC feed. Google spend was intentionally throttled during the mid-report feed rebuild.

Daily trend

DateSpendImpClicksCTRConvValueROAS
2026-04-17$72.142,986712.4%0.0$16.520.23×
2026-04-18$64.554,957821.7%0.0$0.000.00×
2026-04-19$52.591,855593.2%0.8$254.244.83×
2026-04-20$88.132,527491.9%1.0$379.004.30×
2026-04-21$173.206,401991.5%2.0$1,177.216.80×
TOTAL$450.61———3.8$1,826.974.05×

Campaign-level (5-day aggregate)

CampaignStatusSpendConvValueROAS
121 | Search | BrandENABLED$174.601.8$633.243.63×
121 | PMAX | All Products (New Feed)ENABLED$144.812.0$1,177.218.13×
121 | PMAX | All ProductsPAUSED$123.760.0$16.520.13×
121 New GMCPAUSED$7.450.0$0.000.00×

Brand Search daily (headroom analysis)

DateBudgetSpentUtilImp ShareValue
2026-04-17$75$43.8358%97%$0
2026-04-18$75$11.9716%100%$0
2026-04-19$75$17.9824%92%$254
2026-04-20$75$62.7384%98%$379
2026-04-21$75$38.1051%94%$0
Brand Search read: budget is $75/day, peak utilisation 84% (Apr 20), impression share 92-100%. Demand is largely already being captured. Not a scaling opportunity. The Apr 21 anomaly (30 clicks, $38 spend, 0 conversions) is worth investigating as a potential attribution lag.
Context — why the spend looks low: Google spend was intentionally throttled while rebuilding GMC feed integrity. Before cleanup (Apr 19 morning): 10,529 issues across 4,261 products, 0 approved. After cleanup (Apr 22 morning): 27 issues, 1,605 approved. The Apr 21 8.71× PMAX ROAS is the first clean signal after the cleanup landed in Google’s feed evaluation.
Meta Ads (AU) · Data

5-day Meta Ads performance

All figures based on Meta’s purchase_roas metric — the ROAS column shown in Ads Manager. Apr 19 highlighted as the best-ROAS day.

Daily trend

DateSpendImpClicksCTRATCPurchValueROAS
2026-04-17$628.8119,6678864.5%262$648.001.03×
2026-04-18$602.7619,0549314.9%264$897.401.49×
2026-04-19$820.5025,6071,4295.6%589$2,651.153.23×
2026-04-20$649.5119,9031,1055.6%325$1,165.401.79×
2026-04-21$618.6717,1229895.8%446$866.001.40×
TOTAL$3,320.25————26$6,227.951.88×

Campaign-level (5-day aggregate)

CampaignSpendPurchValueROAS
SI - DD | Adv+_Catalogue_Conversions_PUR$2,13519$4,7432.22×
SI - ADV+ | Sales | AUS - April 2026$5021$3180.63×
SI - DABA | AUS - April 2026$2051$1190.58×
SI - ADV+ | Catalogue + Sales | AUS | April 202$1314$1,50211.45×
SI - MOFU | Sales | AUS - April 2026$1291$1991.54×
SI - Dynamic Remarketing | AUS - April 2026$1190$00.00×
DD_Meta_Traffic_AU_TOF$990$00.00×

Adset scaling map

Sorted by ROAS. Action tags reflect frequency + spend-level + performance combined signal.

AdsetCampaignSpendFreqPurchROASAction
ADV+ | Max Conv Value | AUSI - ADV+ | Catalogue + Sales$1311.6411.45×🚀 Scale 4×
Purchase | AUSI - DD | Adv+_Catalogue_Conve$21353.2192.22×🔄 Refresh creative
VV 3 sec | 55 - 65+ | WomenSI - MOFU | Sales | AUS - Apri$932.112.13×📈 Hold + monitor
ADV+ Prospecting | 23 - 64 | WomenSI - ADV+ | Sales | AUS - Apri$5022.310.63×⚠ Cut
LAL AU 3% ATC & Purchase 90D | 55 - 65SI - DABA | AUS - April 2026$2051.610.58×⚠ Cut
Open_AU_25+_SaleDD_Meta_Traffic_AU_TOF$991.500.00×⚠ Cut
Website Visits 180D | 18 - 50 | WomenSI - MOFU | Sales | AUS - Apri$221.200.00×—
Social Engagers + Followers | 55 - 65+SI - MOFU | Sales | AUS - Apri$131.100.00×—
Website Visitors 180D | 18 - 50 | WomeSI - Dynamic Remarketing | AUS$121.700.00×—
Social Engagers & Followers | 55 - 65+SI - Dynamic Remarketing | AUS$1071.800.00×⚠ Cut
The Plan

14-day scaling roadmap

Side-by-side per channel. Revenue projections are conservative, assuming ROAS decay as we scale.

Google · Steven

Week 1

  • New PMAX daily budget: $150 → $300
  • Configure listing group filter by custom_label_0
  • Investigate Apr 21 Brand Search 0-conv anomaly
  • Baseline PMAX asset group performance for week 2 optimisation

Week 2

  • If PMAX ROAS ≥ 5×: scale to $500/day
  • Test splitting PMAX by margin-tier using Shopify product tags
  • Activate dormant NZ PMAX feed if AU ROAS holds
+$10–15K projected revenue
at 5× ROAS on ~$2,500 incremental spend
Meta · Michael

Week 1

  • Scale ADV+ Max Conv Value: $26 → $100/day (4×)
  • Cut LAL 55-65 ATC/Purch ($205, 0.58× ROAS)
  • Cut Social Engagers 55-65+ ($107, 0× ROAS)
  • Cut Open AU 25+ Sale Traffic ($99, 0× ROAS)
  • Reduce ADV+ Prospecting 23-64 50% ($502 → $250)
  • Upload 3 new creatives to the Purchase | AU adset

Week 2

  • If 11.45× adset holds ≥ 6×: scale to $200/day
  • Mirror GMC cleanup methodology on Meta Commerce catalogue
  • Prep Mother’s Day creative pack (peak May 11)
+$6–10K projected revenue
net spend neutral — reallocation-funded (+$1,200 from cuts)
Methodology

How these numbers were pulled

Full transparency so Michael and Steven can reproduce every figure independently.

Meta Ads (AU)

  • API: Graph API v20 · act_664878987453372 (AU account)
  • ROAS source: purchase_roas field (default Ads Manager metric)
  • Attribution: 1-day-click + 1-day-view (Meta default, matches Michael’s UI)
  • Revenue calculation: spend × purchase_roas (avoids double-counting from summing omni_purchase action values across days)
  • Date range: 2026-04-17 to 2026-04-21 inclusive · AEDT

Google Ads

  • API: Google Ads API v21 · Customer 959-176-7044 (Czarina) via MCC 841-066-8004
  • Query: GAQL on campaign and asset_group views
  • Metrics: cost_micros, conversions_value, search_impression_share, search_rank_lost_impression_share
  • GMC cleanup context: Content API v2.1 + Merchant API v1beta on Merchant ID 5769512617
Prior version correction: The initial scaling brief cited 22.9× ROAS for the ADV+ Max Conv Value | AU adset. That figure came from summing daily omni_purchase action values across the 5-day window, which double-counts $1,502 of revenue attributed to days where the adset had $0 spend (delayed Meta attribution). The correct in-platform figure is 11.45×. Michael caught this; every Meta number in this document has been re-pulled using purchase_roas to match Ads Manager.