Performance Max lead quality for US home services — stop buying $12 forms dispatch ignores

Direct answer — how do US home services fix Performance Max lead quality?

Demote soft form submits as primary, instrument calls with duration floors, feed Accepted / Booked / Won via offline conversion imports, harden metro geos and landers against aggregator-style junk, and pause or starve PMax until Enhanced Conversions, GCLID→CRM, and dispatch definitions tell the truth. US plumbers, HVAC, electrical, roofing-adjacent, and multi-metro locals see PMax “working” as cheap CPL while CSRs ignore half the queue — Soft Primary theatre at automation speed.

This insight owns Performance Max lead quality for US service / home services businesses — call thresholds, CRM stages, Angi/HomeAdvisor-culture junk, form spam, and OCI fixes. UK twin: PMax lead quality UK. Setup OS: PMax SMB playbook. Attribution theatre: PMax attribution inflation. Sequencing: PMax vs Search. Structure: home services Google Ads. Measurement: US conversion tracking, OCI, call tracking, primary vs secondary.

Working rule: CPL is not a KPI. Sales-accepted rate and cost per booked job are. If PMax lowers CPL while booked jobs stay flat, you lost.

What “good” looks like after 30–60 days: money-adjacent primary (Accepted or Booked); raw Lead secondary; call conversions with duration floors; OCI live with GCLID; metro geos match vans; brand exclusions on; cost per Booked inside allowable (break-even CPA); MER holds (metrics).

What this URL owns vs siblings

Own here: US home-services PMax quality operating system — call vs form primacy, CRM stage ladder for Smart Bidding, aggregator-junk patterns, metro/multi-location hygiene, weekly quality scoreboard.

Delegate: full PMax launch → playbook + 38-point checklist; view-through / brand ROAS lies → attribution inflation; emergency vs scheduled Search structure → home services; UK GDPR/LSA twin → PMax lead quality UK; Meta Instant Form twin → US Meta lead quality; map pack → US Local SEO.

Out of scope: ecommerce Shopping PMax; full HIPAA programmes (counsel for regulated health); legal advice on state privacy notices.

Entities: Performance Max, Maximize Conversions, Target CPA, call conversions, Enhanced Conversions, offline conversion imports (OCI), GCLID, Local Services Ads (US), Angi/HomeAdvisor-style leads, brand exclusions, Diagnostics, MER, CRM stages, metro geos.

Why US PMax lead quality collapses

US home-services PMax accounts often celebrate $8–$20 CPL while dispatch ignores half the leads. Typical failure modes:

1. Soft primary — any form submit or “get a quote” thank-you as primary; PMax finds tire-kickers and job seekers.
2. Phone closes the job; forms are primary — PMax floods forms while real bookings happen on calls that are untracked or secondary.
3. Call conversions without duration floors — 8-second pocket dials train Maximize Conversions.
4. Aggregator culture — Angi/HomeAdvisor-style junk expectations; owners accept low close rates as “normal,” then PMax scales them.
5. Statewide or national geos for a three-county shop — cheap fills outside drive time.
6. No CRM stage imports — Google never sees Accepted/Booked, only browser Lead.
7. Multi-metro blended CPL — one national PMax hides which city books.
8. Attribution inflation mistaken for quality — high ROAS from brand/view-through while CRM quality falls (attribution).
9. Speed-to-lead broken — after-hours forms sit until morning; competitor already booked; Google still learned from the submit.
10. Franchise freestyle — locations invent their own Lead definitions; HQ cannot reconcile QBR quality.

PMax did not invent bad home-services leads. It scales them faster than manual Search when the conversion definition is soft (bad conversion data).

Define quality before you scale budget

Work backwards from ServiceTitan / Jobber / Housecall Pro / CRM:

1. Lead — form, call, or chat received
2. Qualified / Accepted — in service area, job type fits, price band plausible, homeowner/landlord rules if you use them
3. Booked — appointment or job on the diary
4. Completed / Won — invoice or job complete

Primary for PMax should be stage 2 or 3 once volume allows — not stage 1 forever. Import via OCI. Demote raw Lead to secondary (primary vs secondary).

Dispatch and marketing must sign the stage dictionary. If “qualified” means different things to the owner and the CSR, Smart Bidding cannot help.

Call-duration floors (non-negotiable for US trades)

Phone is often 40–70% of premium bookings. Instrument:

• Dynamic numbers where volume allows
• Duration threshold that filters spam and pocket dials (call tracking guide, call assets migration)
• Path priority so one customer who clicks, forms, then calls is not three primaries (multi-path)

A 15-second “conversion” is how PMax looks busy while the diary stays empty. Score coverage with Lead Path Coverage.

Primary conversion hygiene (US checklist)

• One primary money-adjacent action per lead-gen objective.
• Enhanced Conversions on website forms.
• Call conversions with duration floors when phone books jobs.
• No primary on page view, scroll, or “button click.”
• GCLID into CRM at lead create for OCI.
• State privacy / CMP posture respected where you run opt-outs — do not invent PII (US tracking).
• Score stack: Tracking Trust, Google Ads Health Score, PMax Readiness.

If prerequisites fail, do not “fix quality with creative.” Fix measurement (conversion tracking guide).

Form spam and aggregator-style junk

PMax will send traffic to whatever final URL and conversion you reward.

• Require ZIP or city, mobile phone, service type, and timing (today / this week / researching).
• reCAPTCHA / bot filters appropriate to your stack.
• Message-match landers (service + metro) — not a homepage dump (landing message match).
• Speed-to-lead SLA: after-hours Instant Form culture without answering burns fee and still trains bids.
• Exclude never-fit / out-of-area from offline uploads so they never become “conversions.”

Thin “free quote” pages with three soft fields are how Maximize Conversions finds $5 junk forever — the same junk aggregator markets sold for years, now with Google AI speed.

Offline imports — the quality fix that sticks

When the diary is truth:

1. Store GCLID on the lead.
2. Upload Accepted / Booked / Won with correct conversion time and value when possible.
3. Keep raw Lead as secondary.
4. Flip PMax toward the offline money action when weekly volume supports learning.

Expect temporary CPL rise after the flip. Judge cost per Booked. Full playbook: OCI.

LSA, Search, and PMax — do not pay three times for one job

US Local Services Ads (where eligible) plus Search plus PMax can all claim the same booking.

• One booked-job definition in CRM with source tags.
• Do not treat every LSA lead and every PMax form as additive without matching phone.
• Sequence: measurement → Search foundations → LSA where eligible → PMax only if ready (home services, PMax vs Search).
• Capacity gates: if techs are full, pause urgency and starve PMax before buying more spam.

Metro and multi-location hygiene

• Pin counties, cities, or radii you staff — Phoenix is not “Arizona.”
• Split metros when economics differ.
• Franchise: HQ locks event names; locations own geo and answer rates.
• Map pack still owns much of near-me — PMax does not replace US Local SEO.

Geo, brand, and Search before quality theatre

• Brand exclusions on PMax; fund branded Search (playbook).
• High-intent Search for emergency / core services before PMax “coverage.”
• Quality problems that are really brand cannibalization will not be fixed by a new form field alone (attribution).

Optimization ladder for US PMax lead gen

| Event Google sees | Dispatch experience | Role | |-------------------|---------------------|------| | Soft form Lead | Tire-kickers, spam | Secondary only | | Call under duration floor | Pocket dials | Not a conversion | | Accepted / Qualified | CSR will chase | Good mid primary | | Booked | Job scheduled | Strong primary | | Won + value | Revenue | Best when volume allows |

Flip when weekly Accepted/Booked volume supports learning. Target CPA only after quality floors (Max Conv vs tCPA).

Board KPIs that survive the owner meeting

• Cost per Accepted and cost per Booked
• Accepted / Booked rates from PMax vs Search
• Answer rate / speed-to-lead
• Call conversion quality (duration distribution)
• MER / contribution — not PMax ROAS alone
• Allowable booked CPA from job margin math

Kill vanity: platform CPL, All conversions ROAS, and “leads” that never enter the CRM.

Weekly quality ops (30 minutes)

1. Accepted and Booked rates by metro and campaign (PMax vs Search).
2. CPL vs cost-per-Booked gap — if widening, stop celebrating CPL.
3. Out-of-area / spam share; call duration histogram.
4. OCI upload health + Enhanced Conversions.
5. Brand exclusion and negative candidates.
6. Capacity vs spend — pause urgency if full.
7. Aggregator vs paid quality comparison (close rate, not CPL).

Speed-to-lead and capacity gates

Write an SLA:

• Who calls within X minutes during business hours
• After-hours and weekend rules if ads promise same-day
• Pause or soften urgency creative when the diary is full

PMax will keep buying leads you cannot answer. Retarget and cold both need capacity honesty — same lesson as US Meta lead quality.

Unanswered evening forms still trained yesterday’s Maximize Conversions model. Media efficiency is an ops problem as much as a bid problem.

30-day quality rebuild

Week 1 — Truth. Sign stage dictionary with dispatch. Screenshot primaries. Score PMax Readiness and Tracking Trust. List call vs form share.

Week 2 — Hygiene. Demote soft Lead; add ZIP/phone/job type; set call duration floors; pin metro geos.

Week 3 — OCI. GCLID into CRM; upload Accepted/Booked; verify in Google Ads. Keep Search foundations funded.

Week 4 — Flip and judge. Point PMax at money-adjacent primary; accept uglier CPL; report cost per Booked. Kill or hold PMax if quality still collapsed after clean signal.

Proof pattern (what “good” looked like)

Published US home-services Google Ads work (four-metro HVAC/plumbing) followed the same philosophy: fix call + CRM truth before trusting automation dashboards. PMax quality is not a separate religion — it is that measurement stack applied when Performance Max is allowed to spend.

If your account still celebrates platform CPL while ServiceTitan / Jobber booked jobs are flat, you are looking at Soft Primary theatre. Flip the event; do not raise budget.

Vertical snapshots

HVAC / plumbing / electrical: Emergency vs scheduled as a form field; after-hours creative needs after-hours answering or you buy abandoned urgency that still trained bids.
Roofing / exterior: Storm-chaser and out-of-area ZIP spam — ZIP required; photo upload optional on website path.
Multi-location franchise: HQ event taxonomy; per-metro PMax or clear reporting — no national soup day one.
Clinics / specialty local: Counsel on regulated data; still define Booked in CRM even if some payloads are constrained.

US vs UK twin — what differs

| Topic | US emphasis | UK twin | |-------|-------------|---------| | Aggregators | Angi / HomeAdvisor-style lead markets | Checkatrade / directories / LSA screening | | Close path | Call-first + duration floors | Call + Consent Mode + LSA overlap | | Privacy | State notices / CMP; browser thinning | GDPR + Consent Mode v2 defaults | | Currency vanity | $ CPL theatre | £ CPL theatre |

Same ladder. Different SERP and call ops. Read both if you run multi-market accounts: UK.

FAQ

Why is my Performance Max CPL low but dispatch hates the leads?

Soft primary and/or no call floors. PMax found what you rewarded. Fix the event — not only the creative.

Should US home services use PMax for lead gen?

Yes after tracking, Search foundations, and a quality loop. No as a day-one national dump (PMax vs Search, home services).

What call duration should count as a conversion?

Whatever dispatch says is a real conversation — often 60–90+ seconds for estimate calls; tune with spam rates. Document it; do not use 15 seconds.

Do I need offline conversions?

If phone or delayed CRM stages book the job — yes. Form-only primaries are how US service PMax accounts rot.

How is this different from Meta lead quality US?

Same philosophy (stages over soft Lead). Different products. Meta Instant Forms: US Meta lead quality.

Should we pause PMax while we fix tracking?

Often yes — or cut to learning-minimum while Search captures high-intent. Teaching Maximize Conversions on spam for another month costs more than a quiet rebuild week.

How do we know quality improved?

Accepted rate up, cost per Booked down or stable at higher volume, out-of-area share down, call-duration distribution healthier, dispatch complaint volume down. CPL alone never answers this.

Related frameworks

• PMax lead quality UK — twin
• Performance Max SMB playbook · PMax launch checklist · PMax Readiness
• PMax attribution inflation · PMax vs Search
• Home services Google Ads · US conversion tracking
• OCI · Call tracking · Primary vs secondary
• US Local SEO · US Meta lead quality
• Google Ads agency US · Free Google Ads audit (US)

Next steps

1. Screenshot current PMax primary; score PMax Readiness.
2. Sign Accepted/Booked with dispatch; set call duration floors.
3. Harden forms + metro geos; wire OCI.
4. Flip primary; judge cost per Booked for two weeks.
5. Request a US Google Ads audit if CPL looks great and the diary does not.

US Performance Max lead quality is not a creative trick. It is a call-floor + CRM stage + OCI Booked + metro honesty system. When Google sees who books, PMax can help. When it only sees $12 form spam, every budget step buys prettier junk faster.