Account 4321874619. Grounded in the bridge audit pack collected 2026-06-09, USD, API v20. Windows: last_7 06-02→06-08, prior_7 05-26→06-01, last_30 05-10→06-08, last_90 03-11→06-08. Nothing was executed; this is the team-ready read with actions tiered by priority.
TL;DR
Two tiers, two jobs. Cold acquisition (Domain Broker Leads PMax) spends 93% of budget at $287/lead. Three small retargeting/display campaigns spend under 7% but deliver ~24% of leads at $41 to $74 each. They are the efficient layer, and the cold campaign feeds them.
Workhorse trend is up. The 90d-to-30d CPA "rise" ($124 to $232) is a tracking-swap artifact, not decay: conversions moved off a legacy goal onto the thank-you tag in late April. Same-tracker read is improving. WoW CPA $224 to $162, CPC halved $11.54 to $7.36. Cold lead cost is about $287. Open question is what share of those leads close.
Biggest lever: fix the workhorse headlines. 8 of 15 enabled headlines are Google-rated LOW, only 1 is BEST, and 4 BEST broker-intent headlines were removed (Hire A Domain Name Broker, Buy Premium Domain Names, Your Premium Domain Broker, Expert Domain Broker Services). Swap them back in.
Two compliance problems in retargeting copy: a sell-side line ("Serious Buyers, Sellers") and an unverified "$775K saved" claim. Both should come out.
One dead campaign holding a slot: Display "4-20-26" is ENABLED but did $0 / 0 impressions in 90 days.
No search campaign live. Zero high-intent "domain broker" capture and no search-term visibility while PMax runs blind.
Account picture
Window
Spend
Conv
CPA
CVR
CPC
last_7
$4,040.51
25.0
$161.62
4.13%
$6.67
prior_7
$3,931.08
17.53
$224.20
4.25%
$9.52
last_30
$16,141.99
69.53
$232.17
4.26%
$9.88
last_60
$21,251.32
137.53
$154.52
6.15%
$9.50
last_90
$23,667.34
191.28
$123.73
7.10%
$8.79
Spend is steady (~$4k/wk). The 90d-to-30d CPA climb is the tracker swap (legacy zaraz goal to thank-you tag, late April), not decay. Same-tracker trend is improving: WoW CPA $224.20 to $161.62, conversions 17.53 to 25.0, CPC $11.46 to $7.36.
By campaign (last_30), ranked by efficiency not spend:
Campaign
Role
Spend
Conv
CPA
Display Whois
Warm retargeting (whois visitors)
$82.94
2.0
$41.47
PMax Retargeting
Warm retargeting
$461.72
8.0
$57.72
Display Retargeting
Warm retargeting
$520.78
6.99
$74.47
Domain Broker Leads PMax
Cold acquisition
$15,076.54
52.53
$286.99
Display 4-20-26
dead (0 delivery)
$0.00
0
n/a
The retargeting/display layer is $1,065 spend to ~17 leads (about 7% of spend, ~24% of leads) at a $41 to $74 CPA. That is expected of warm audiences and is capped by audience size, but it is the efficient end of the account and depends on the cold PMax to feed it. The cold campaign carries the volume at the higher $287 CPA.
Working: WoW is genuinely better. More conversions, lower CPA, CPC nearly halved.
Not working: (1) headline pool is quality-starved: enabled set is 1 BEST / 6 GOOD / 8 LOW, the LOW cluster is almost all Enterprise / Elite / High-End / Stealth framing, while 4 BEST broker-intent headlines sit REMOVED. (2) The $1M VPN.com proof story is absent from this campaign entirely. (3) Hard-tier "Enterprise/Elite" language sits in the main pool; the profile says it belongs in a dedicated high-roller group with the main groups kept accessible.
Clean: no banned 97% / $75M claims here. 3,201 campaign negatives, heavy sell-side/registrant coverage.
Levers: swap 4 LOW elite headlines for the 4 removed BEST ones (Tier 2). Add 1 to 2 assets built on the verifiable $1M story (Tier 2). Hold budget: the $287 CPA is per-form-fill, so confirm the form-fill to closed-deal rate before scaling (Tier 3, hold).
PMax Retargeting · $50/day · warm retargeting
Verdict: most efficient campaign in the account, but the copy breaks directives.
last_30 $461.72 / 8 conv / $57.72 CPA / 9.76% CVR. Best CPA and CVR of any campaign, and improving (last_7 $46.04 CPA).
Directive violations (must fix): long-headline "Premium Domain Brokerage For Serious Buyers, Sellers and High-Value Deals" breaches BUY-SIDE ONLY. Long-headline "VPN.com Was Not For Sale. We Bought It Anyway and Saved $775K At The Table" is an unverified specific claim. The verified proof point is the $1M acquisition, not a $775K saving. Same class of claim purged before (97% / $75M).
Levers: rebuild both long-headlines buy-side and verified-claims-only (Tier 2). Confirm "$775K" true or false (needs Ben).
Display Retargeting · $50/day · warm retargeting
Verdict: efficient ($74 CPA), but watch the ramp. last_30 $520.78 / 6.99 conv / $74.47 CPA, well under the $287 cold CPA. last_7 spend ramped ~3.5x WoW and CPA rose with it ($119.98), so efficiency thins as it scales (expected for a finite pool). Lever: keep running; if last_7's ~$120 CPA holds next week, pull spend back toward the cheaper $50/day level (Tier 2 later).
Verdict: cheapest CPA in the account ($41.47), but starved. last_30 $82.94 / 2 conv / $41.47 CPA on a $10/day cap. Only 2 conversions so the read is thin, but it is the lowest cost-per-lead here. The 0.19% CTR is normal for display. Lever: hold $10/day one more week for confidence, then consider a modest bump if CPA holds (Tier 3).
Display 4-20-26 · $30/day
Verdict: pause and rethink. ENABLED but $0 / 0 impressions over 90 days (reconciliation flagged no_delivery). Not wasting money, but a dead ENABLED shell, likely a disapproved ad or empty audience. Lever: pause it (Tier 2) or diagnose why it never served.
Cross-campaign
Budget shape: 93% sits in the $287-CPA cold campaign; the $41 to $74 warm layer is capped at $10 to $50/day. You cannot just shift budget down: retargeting is bounded by audience size and the cold campaign generates that audience. But the efficient campaigns have headroom. Give them more where the audience supports it (Whois at $41 is the clearest candidate).
Conversion-action clutter (hygiene): 5 lead-form actions plus several GA4 engagement actions are enabled and primary account-wide. None double-count domain broker today, but www.vpn.com and go.vpn.com thank-you are both live and primary, and landers deploy to go.vpn.com, so it is a latent dedup risk worth tidying.
No search layer: with the search campaign paused, there is no high-intent exact-match capture and no search-term mine. PMax does all the work and you cannot see its queries. A tight search campaign would add intent capture and a readable search-term report (Tier 3, propose).
The $1M story is underused: it appears (mangled as "$775K") in the small retargeting PMax and is absent from the workhorse, the one place it would matter most.
Judge CPA (~$287/form-fill) against close rate before any budget move
3
Workhorse PMax
CPA is real; the unknown is form-fill to closed-deal rate
9
Re-enable a tight buy-side search campaign
3
new/paused
No intent capture, no search-term visibility today
10
Give Whois (+ efficient retargeting) more budget if CPA holds
3
Whois / Retgt
$41 to $74 CPA capped at $10 to $50/day
Items 1 to 5 are Tier 2 and ready to apply on your go, but they touch live campaigns, so per the hard rule the bridge runs ads down first, then applies. Items 7 to 10 are proposals.
Data caveats
Conversion tracking: workhorse conversions run about 1:1 on one thank-you action (50.53 of 52.53 last_30), so CPA is real at the form-fill level. The form-fill to closed-deal rate is the real unknown and Google cannot see it. www and go.vpn.com thank-you both fire as primary (latent double-count if traffic shifts to go). Ties to the conversion-tracking work in google-ads-offline-conversion.md.
PMax opacity: no search terms, no keyword/QS, no ad-group breakdown by design.
Display creatives were not pulled in this collector version (metrics + negatives only).
161 paused campaigns exist (mostly legacy VPN-software product), not audited.
Reconciliation: all ad-group sums tie to campaign totals; one no_delivery flag (Display 4-20-26).