Home services runs its marketing on Cost Per Lead. The business runs on booked, completed, margin-positive jobs. The space between those two numbers is where multi-location HVAC and plumbing brands pay for calls nobody can dispatch, window and home improvement companies celebrate a cheap lead month and discover at the estimate stage that half the leads were out of territory, and private-equity roll-ups spend the integration budget buying the same phone call three times under three brand names.
Open the dashboard at any home services marketer in 2026 (a single-market plumber, a 40-branch HVAC platform, a national window and remodeling brand) and the spine of the reporting is the same. Impressions, clicks, calls, form fills, chats, leads. A cost-per-lead number, probably split by Local Services Ads and "everything else." Maybe a booking rate somewhere in a call-tracking export that the marketing team does not own and the call center does not read.
That structure made sense when a lead was something the marketer produced and the office turned into a job. The homeowner searched "furnace repair near me," clicked, called, and a CSR booked the appointment. The marketer's job ended at the ring. Booking, dispatch, close rate and ticket size belonged to operations.
That separation no longer holds, and Google is the reason. In August 2026 Google began moving Local Services Ads into Google Ads as Performance Max campaigns with pay-per-lead goals, starting with plumbing, HVAC, electrical, roofing, pest control and the rest of the home services categories. The pay-per-lead model stays. The manual maximum cost per lead does not: it is deprecated in favor of a campaign-level Target CPA. Google defines what a lead is, decides which ones are charged, and now sets the price inside the same bidding system that runs everything else in the account. Performance Max for lead generation already bid on signals the marketer never sees, toward whichever conversion the account defines. Feed either system a lead target and it will produce leads: cheap ones, out-of-area ones, ones that call at 11pm when nobody is booking, ones from the neighboring brand you also own. The platform does what it is told. It is the metric that is wrong.
What Cost Per Lead actually optimizes for
Cost Per Lead treats every contact as identical at the moment it happens. A homeowner with a failed compressor in July, inside the service area, calling during booking hours, who schedules a same-day visit that closes at a $9,000 replacement ticket, is worth the same in CPL as a renter asking for a quote the company does not offer, in a ZIP code the trucks do not reach, who hangs up after 20 seconds. The dashboard cannot tell the difference. The platform optimizes toward whichever audience and whichever hour produces the ring cheapest.
The downstream effect is predictable in home services accounts of any size. The cheapest leads cluster where booking rates are worst: after hours, in the outer ring of the service area, on the broadest service terms, and on the brand with the thinnest CSR coverage. Marketing hits the CPL number. Operations watches the booking rate slide. Finance sees cost per completed job drift upward and cuts the marketing budget in the one place it was actually working.
Run the arithmetic on a 40-branch HVAC platform, with illustrative numbers. Say it buys 10,000 leads a month at a blended $60, so $600,000. Daytime leads inside the core service area book at 55 percent; after-hours and outer-ring leads book at 25 percent, and the outer-ring jobs that do book carry an extra hour of drive time. If 40 percent of the volume is in that second group because it is where the lead is cheapest, the platform is producing 4,300 booked jobs, and the 4,000 cheap leads contributed 1,000 of them at an effective $240 per booked job against $109 for the rest. Move a third of that spend into daytime core-area inventory at a 30 percent higher CPL and booked jobs rise while total spend does not. CPL goes up. Cost per booked job goes down. Only one of those numbers is on the dashboard.
Local Services Ads made this concrete in a way search never did, and the migration makes it sharper. LSA already priced in a unit closer to the business than a click. But the unit was always Google's: a lead is what Google says is a lead, a charge is what Google decides to charge, and with manual per-lead bids gone, the price is what Google's bidding system decides it is. A company running pay-per-lead campaigns on cost per lead is optimizing inside a unit the seller defined and now prices, which is the same mistake The Seller's Unit describes, with a truck attached.
Last-click attribution had this same shape. The metric was cheap, defensible and quietly destructive. Meridian MMM exists because enterprise marketers finally got tired of the Search-attribution illusion. Cost Per Lead is the home services version of the same illusion, and the AI-mediated local funnel of 2026 makes it more expensive every month.
The metric that actually runs the business
Cost Per Booked Job is the marketing investment required to produce one scheduled, dispatchable job, weighted by the expected completed-ticket margin of that job. It is not a new number. Every home services CFO and every private-equity operating partner already runs it in the board deck, usually as cost per completed job or marketing cost as a percentage of revenue. The novelty is putting it on the marketing dashboard, in the bidding signal, and in the agency scorecard, instead of letting it live in a ServiceTitan or Housecall Pro report the marketing team never opens.
Three components matter. The booking rate: what share of leads from this source, this hour and this ZIP become a scheduled job. The dispatch rate: what share of scheduled jobs a technician actually reaches, net of cancellations and capacity. The expected ticket margin: what a completed job from this source and service line is worth after labor, parts and the cost of the truck being somewhere else.
The math is simple. The data is scattered. Booking lives in the call-tracking platform or the CSR software, dispatch and ticket value live in the field-service management system, and marketing data lives in Google Ads, Lead Manager and GA4. Joining the three at the lead level, by source, campaign, hour and service area, is the actual work. It is why Cost Per Booked Job is not on most dashboards: not because nobody believes in it, but because nobody has owned the plumbing between the phone system, the FSM and the ad platform.
What changes when you optimize on Cost Per Booked Job
Run the same media plan on both metrics against the same lead data and the budget shifts in three predictable ways.
Audience and geography move toward the trucks
CPL is indifferent to where the lead is. CPBJ is not. The ZIP codes and service lines with the highest booking and dispatch rates are usually not the cheapest to buy a lead in, because competitors are already there. CPBJ moves spend toward the service area the company can actually serve at margin, pulls it out of the outer ring where the lead is cheap and the drive time is an hour, and starts treating capacity as a marketing input. The first thing most teams discover is that they have been paying to make the phone ring in markets where the calendar was already full.
Dispatch capacity moves from operations to marketing
When the optimization signal is "lead," a 42 percent booking rate is a call-center problem. When the signal is "booked job," it is a marketing efficiency loss that shows up the same day, by source and by hour. The two teams start looking at the same number. Dayparting follows CSR staffing instead of impression share. After-hours leads get routed to an answering service with booking authority or get bid down. In the roll-ups, the three brands bidding against each other in the same metro get a shared capacity view, and the duplicate-lead problem becomes visible for the first time.
Creative shifts from urgency to fit
CPL rewards creative that maximizes the ring: "24/7 emergency service," "free estimates," "call now." The caller most responsive to that creative is the caller least likely to book, because the ad promised something the company does not sell at that hour or in that ZIP. CPBJ rewards creative that pre-qualifies: the service area in the ad, the service lines offered, the booking window, the financing. The lead volume drops. The booking rate rises. The CFO notices inside one season.
"The lead most responsive to '24/7 emergency service' is the lead most likely to call at 11pm in a ZIP code your trucks do not reach. That is not a call-center problem. It is a metric problem."
Cost Per Lead vs. Cost Per Booked Job, side by side
How to feed Cost Per Booked Job into Google Ads
The implementation is shorter than most teams expect, because Google has quietly built most of the primitives. Enhanced conversions for leads ties an offline event back to the click or call. Conversion value rules let the value vary by geography and audience. Value-based bidding inside Search and Performance Max acts on it. What is missing is the instrumentation and the value.
The shape of the build on the Search and lead-gen Performance Max side: define the booked-job conversion as an offline event, fired from the field-service management system or the call-tracking platform when a job is scheduled, and keyed back to the click or call with enhanced conversions for leads. Calculate booking and dispatch rates by source, hour and service-area polygon from the FSM. Multiply by expected ticket margin by service line. Pass the result as the conversion value, and use value rules by location to encode the service area directly into the bid. Move the campaigns to a value-based bid strategy once the offline events are flowing. Google's AI will reweight toward higher expected booked-job value within two to four weeks, and the lead mix changes before the lead volume does.
On the pay-per-lead side the levers are narrower, which is the point. Lead Manager in Google Ads, which replaces the old LSA inbox, carries the charged status, the lead type, a lead feedback form and a download for the CRM. Use all four as a discipline: feed every lead back as booked or not, export the lead file into the same booking-rate model the Search campaigns run on, and set the campaign-level Target CPA by service type and market as a proxy for expected booked-job value rather than as a budget-pacing knob. Google's own guidance already says that failing to respond to leads hurts ranking; the booking signal is the version of that rule the business actually earns in. The goal is one unit across both surfaces, so a pay-per-lead campaign and a Search campaign in the same metro are judged on the same number.
The harder work is organizational. CPBJ forces the call center, dispatch and marketing into the same data. In a roll-up it forces the brands into the same capacity view. That requires a sponsor, usually the operating partner or the COO, and it is also the moment marketing starts speaking the language in Marketing on the P&L: how the decision reads in completed-job margin, not in leads.
Where compliance actually sits
Home services is lightly regulated compared with the other verticals in this series, but two things need saying. Call recording and lead tracking fall under state consent rules and, for outbound follow-up, the TCPA, so the booking-rate model should run on consented, first-party call data the company already captures for operations. And the Google Verified badge and Google Guarantee carry license, insurance and background-check requirements that the marketing team does not control. CPBJ makes the cost of a lapsed license visible as a lost booked job, which is a feature.
What this looks like on the P&L
The financial argument for Cost Per Booked Job is the same shape as every argument on this site: the marketing decision and the business decision become the same decision.
A home services company running on CPL allocates budget to whatever produces the ring cheapest. The leads come in. Some fraction books. Some fraction of those is dispatched. Some fraction of those completes at a margin that covered the truck. The blended numbers look fine for a month. By the end of the season, cost per completed job has drifted upward, the branches with full calendars were buying leads they could not serve, and the budget conversation with the operating partner is about cutting marketing, not fixing the unit. Both reports are technically correct.
A company running on CPBJ allocates budget against expected booked-job value at the point of the bid, by service area and by hour. The leads come in already weighted toward the jobs the trucks can reach and the tickets that carry margin. The seasonal reconciliation still happens, but the variance is information rather than damage: it improves next season's booking-rate model instead of indicting this season's spend. The same marketing dollars produce more completed jobs, more margin per truck-day, and a lower Cost Per Decision, because the dispatch decision and the media decision are finally the same decision. FinTech has the same shape of metric problem with KYC in place of the dispatch board. The lesson generalizes.
This is one vertical instance of a broader argument: every marketing KPI in common use measures activity, not decisions, and agentic AI is the first technology that makes the correction operable rather than aspirational. Home services is where the gap is most literal, because the decision has a truck and a technician attached to it.
The marketing team is not optimizing wrong. The metric is. Cost Per Lead was the best a home services company could do when the marketer's job ended at the ring and the platform did not price the lead itself. Neither condition holds anymore. The companies that rebuild their KPI around the booked job first will spend the next two seasons compounding a margin-per-truck advantage the rest of the market is still arguing about in leads.
Sources & further reading
- Local Services Ads transition to Performance Max campaigns with pay-per-lead goals (Google Ads Help). The August 2026 migration: pay-per-lead stays, manual maximum cost per lead is deprecated, Lead Manager replaces the LSA inbox.
- Improve your Local Services Ads performance (Google Ads Help). Responsiveness, reviews and booking as ranking inputs.
- About enhanced conversions for leads (Google Ads Help). The mechanism for tying an offline booked job back to the click or call.
- About offline conversion imports (Google Ads Help). Including phone-call conversion imports.
- About conversion value rules (Google Ads Help). How to encode service-area and hour-of-day booking rates into the bidding signal.
- Value-based bidding best practices (Google Ads Help).
- About Performance Max campaigns (Google Ads Help).
- About Smart Bidding (Google Ads Help).