Everything your PPC specialist needs to take over the Kertújítók Google Ads account:
the diagnosis (why leads are low), the three action plans (what to do), the account facts & access,
the do-not-break rules, and a sequenced first-week checklist. All figures were pulled live from
Google Ads, GA4 and GoHighLevel and cross-checked across ≥2 sources.
Read in this order
Three documents. This brief is the map; the other two are the detail. Start here, then read the diagnosis, then choose a strategy.
① You are here
This master brief
The 60-second situation, account access, the do-not-break rules, the two priority fixes, and the first-week checklist.
② Why (the evidence)
Diagnosis report
Full triangulated diagnosis — the funnel, the "140 vs 3" reconciliation, ranked root causes, and what is not broken.
Three distinct Ads+website action plans (Lean · Conversion-led · Full-funnel), every setting/layer, a side-by-side comparison, and a 1-page SOP appendix each.
Note: the 3-strategies document (③) is the deep deliverable — open it for the
full settings-level plans + comparison + SOP appendices. If the link is not live yet it is finishing
generation; it publishes to the same URL.
The situation in 60 seconds
New site + new ad campaigns went live 2026-06-12, but the old 2025 campaigns and old tracking were left running alongside. Result in the ~11 days since: 3 real leads.
Lead capture is NOT the problem. GHL has 5 contacts / 4 opportunities and 0 dropped leads — every form submit became a lead. The 4th opportunity is a manual-entry duplicate ("Kis József" typed twice). Real = 2 website leads (both /gepi-foldmunkak) + 1 phone lead.
Conversion tracking is ~96% fiction. The "~140 conversions" Google reports = a different website's (kisgepi-foldmunka.hu) contact-page views (96) + quote-button clicks (32) + Google-Business "website visits"/"directions". Only ~2 were real form submits. There are ~32 conversion actions in the account. → Smart Bidding is optimising toward junk.
Old campaigns interfere. Old 2025 campaigns = 39% of spend, incl. ~11,600 Ft going to a second website (kisgepi-foldmunka.hu) inside the same account. An old Performance Max campaign = 44% of all traffic @ 0.40% CTR, dumped on the home page. New search campaigns are starved (losing up to 77% impression share to budget).
The money pages get almost no traffic. Kertépítés (the 70% priority) got ~30 sessions in 11 days and 0 leads. Forms leak: 15 starts → ~3 completes. 20 phone-clicks of real intent never reach the CRM.
Good news: billing is fine (ads serving, ~143k Ft/mo), the pages work, and the new search ads have 5–38% CTR — they're just starved. The fixes are structure + measurement, not "spend more".
Realistic target after the fix: ~20–40 leads/month at this budget (vs ~3 now).
Account, access & key IDs
What the PPC specialist needs to operate. Credentials themselves live in the agency vault / with Matt — never in this doc.
System
Identifier
Notes
Google Ads
9678294672 — "Balogh Péter ev." (HUF, Europe/Budapest)
Reachable directly; also under MCC 9494412337 (Rocketfuel Digital). Budget ~150–250k Ft/mo.
Lives in the GHL site head code (not in the repo). All conversion tags fire from here.
GoHighLevel (CRM)
location PuOENnvq1aSnCFcgWxdH
Pipeline "Kertújítók leadek" PmiVcXS4EAeBbHzhKfJx: Új lead → Beszéltünk → Ajánlat kint → ✅ Ügyfél lett / ❌ Nem.
Website
kertujitok.hu (7 pages, GHL CMS)
Thank-you: /koszonjuk. Service pages: /kertepites, /automata-ontozorendszerek, /robotfunyiro, /gyepszonyegezes, /gepi-foldmunkak, /kertfenntartas.
Second site (separate!)
kisgepi-foldmunka.hu
A related earthworks brand currently tangled in the SAME Ads account — must be separated out.
⚠️ Recurring risk: the Google Ads API OAuth app is in "Testing" publishing
status (refresh tokens expire after 7 days) → the reporting dashboard's Ads feed will keep breaking
(~every 7 days). Publish the OAuth app to Production to stop this.
Do-NOT-break rules (invariants)
Never touch the lead-capture wiring: the GHL external-tracking.js tag
(id tk_9a445b55c75a4d3cb9a294baad3399f2), the form ids (hero-form-main,
contact-form-main, widget-form-chat), or the field names
(full_name, phone, email, city, message, main_service, other_services, area_size, page_url, form_type).
The submit handler must preventDefault + delay before redirecting to
/koszonjuk — don't "fix" it to redirect instantly (kills the tracking request).
The live GTM (GTM-P2BRFFK) is in the GHL site head — don't replace it with any repo file.
kertujitok.huDNS + MX live on the OLD host (nethely/webspacecontrol) — don't touch hosting/DNS.
In the GHL page builder the CSS panel must stay empty; replace the HTML block → Publish.
Owner (Balogh Péter) is non-technical — anything he must do has to be dead-simple; the agency/PPC does the technical work.
Keep kisgepi-foldmunka.hu separable — ideally its own Ads account & conversion actions.
Fix #1 (do this FIRST) — conversion-tracking cleanup
Tracking must be clean before any Smart Bidding — otherwise Google optimises toward junk. This is the single highest-leverage fix.
One primary conversion = the lead form submit (hero/contact/widget_form_submit, category "Submit lead form", count "One", 90-day window). Combine into one "Lead form submission" goal assigned to all Search campaigns.
Demote everything else to Secondary (observe): phone-link clicks, quote/CTA button clicks (currently mis-categorised as PAGE_VIEW), GBP "website visits"/"directions", page views.
Remove the second website (kisgepi) actions from this account (the ~96-"conversion" page-view action). Don't delete history blindly — secondary first, archive after ~60 days.
Use native Google-tag/GTM conversions as primary; GA4-import = observe only (GA4 import is delayed and can't do Enhanced Conversions).
Turn on Enhanced Conversions for Leads (hash the email in the GTM form-submit tag).
Set up phone-call tracking properly: "Calls from a website" via a Google forwarding number (supported in HU), min 60s duration; keep the tel:-click as secondary only.
Highest-leverage:GHL offline / CRM conversion import (OCI). Enable Ads auto-tagging → add a hidden gclid field to the forms → create an OCI "Qualified/Converted lead" action → GHL workflow on Opportunity = Won uploads it. This feeds Google real revenue outcomes, not just form fills.
Fix #2 — account structure & budget
Consolidate to 2–3 Search campaigns (no PMax, no Demand Gen at this budget; LSA isn't available in Hungary): Kertépítés-core 60–65% · "Kiemelkedő" (öntöző + robotfűnyíró + gyepszőnyeg) 25–30% · Gépi földmunka 10%.
Themed ad groups (3–8 exact/phrase keywords, 2–3 RSAs each), not SKAGs. Uncheck Search Display-expansion.
One website per account:hard-pause all kisgepi-foldmunka.hu campaigns on day 1 (cross-site self-competition inflates your own CPCs and leaks ~11,600 Ft).
Retire the rest of the old 2025 campaigns safely: pause (don't remove), stagger 2–3/week, only after the new campaigns exit learning.
Match types: exact + phrase only; no broad until 30+ conv/mo + Smart Bidding.
Bidding roadmap: Phase 0 launch = Maximize Clicks + max-CPC cap (~800–1,200 Ft) → Phase 1 (<15 conv/30d) Manual/Max Clicks → Phase 2 (15–29) Maximize Conversions (no target) → Phase 3 (30+) tCPA at +20–25% of observed CPL. Never change budget/bids >20%/week.
Geo = "Presence only", Budapest + Pest megye + explicit exclusions (or ~40 km radius). Campaign-level budgets (not shared).
Starter negative-keyword list (account level)
Cluster
Negatives
Jobs / careers
állás · munka · fizetés · tanfolyam · képzés · vizsga · iskola
DIY / informational
hogyan · diy · ingyen · ingyenes · saját magam · youtube · fórum · reddit
gép vásárlás · ásó · kistraktor árak · bérlés (if not renting)
Then run a weekly search-term audit: any term with 3+ clicks and 0 conversions → add as negative. (In the broken account, ~35% of spend went to 0-conversion terms incl. "földmunka balaton".)
First-week action checklist (sequenced)
Golden rule: fix measurement before touching bidding.
Day 1: hard-pause every campaign pointing to kisgepi-foldmunka.hu; publish the OAuth app to Production; add the account-level negative list above.
Day 1–2: audit all ~32 conversion actions; set the one form-submit goal as primary, demote/secondary the rest, remove the kisgepi/UA junk. Verify in real time with a test form submit.
Day 2–3: turn on Enhanced Conversions for Leads; set up phone-call tracking (forwarding number, 60s); wire GHL GCLID capture + the OCI "Won" workflow.
Day 3–5: rebuild to 2–3 consolidated Search campaigns (Presence-only geo, exact+phrase, Maximize Clicks + CPC cap); point each ad group to its service page (NOT the home page).
Week 1–2: stagger-pause the remaining old 2025 campaigns; on the website, cut form fields (low-friction step 1, drop required email), add a sticky mobile call+CTA bar, fix ad→page→form message match.
Then: let it run, audit search terms weekly, and only move to Smart Bidding once a campaign clears the conversion threshold (see the chosen strategy's SOP appendix).
Reference material (best practices behind the plans)
The recommendations above are grounded in current (2025–2026) best practice from trustworthy sources (Google Ads Help, Search Engine Land/Journal, WordStream/LocaliQ, Optmyzr, CXL, Baymard, GHL docs). Full sourced briefings:
Account structure / budget / bidding / PMax-vs-Search / negatives / benchmarks — see the strategies doc appendix + research1_ads_structure.md.
Conversion tracking done right / Enhanced Conversions / offline CRM import / phone tracking / landing-page & lead-form CRO — see the strategies doc appendix + research2_tracking_cro.md.