Multi-location HVAC operator, Southeast US

1. The case study

The company

A privately held, second-generation family HVAC business operating across four metros in North Carolina and South Carolina. Roughly $15M in annual revenue when we engaged, about 65 field technicians and 15 office staff. Predominantly residential (repair, replacement, install, maintenance plans) with a small commercial book that mostly came from referrals from residential customers who happened to own small businesses. Founded in the early 2000s and profitable every year since 2011.

The situation they came to us with

The business was profitable but stalled. Paid ads were burning approximately $65K per month on Google with a return the owners described as "we think it works, but we cannot actually prove it." HubSpot was in place as the CRM but nobody was using it beyond storing contacts. Phones rang steadily, roughly 900 to 1,100 inbound calls a month across four locations, but there was no channel attribution on any of them. Google Business Profile presence was inconsistent: one of four cities was in the map top three, two were on page two of local results, and the fourth was invisible in its own metro. The book skewed toward repair rather than install, which meant the average job margin was compressing at the same time labor and parts costs were rising. Ownership wanted growth without adding office staff.

The specific brief when we started: "Prove which channels actually produce installs, and make the paid spend defensible to our accountant. If we can also fix the map in Raleigh and Charleston, that would be excellent."

What we did

1. Attribution and phone tracking, first

Nothing else mattered until we could see the data honestly. We deployed CallRail across every inbound channel with dynamic number insertion on the website, unique numbers for Google Ads, Google Business Profile, Bing Ads, organic listings, print, and yard signs. We wired call outcomes into HubSpot through the native integration, so every call became a contact record with a source, a duration, and a recorded outcome. Within 30 days the reporting could answer, for every closed job in the CRM, which channel produced the first touch and which channel produced the booking. About 34% of the paid-media spend was traced to calls that never converted to jobs, and another 12% was traced to calls from existing customers who would have called anyway. That was the reallocation runway that funded the rest of the engagement.

2. Google Ads restructure

The account had accumulated 40+ campaigns over five years of "someone kept adding, nobody removed." We collapsed the account to eight campaigns organized by intent and service line: emergency AC repair, heat pump installation, furnace installation, ductwork, indoor air quality, commercial (separate account structure), and two PMax layers (one for install queries, one for repair). We rebuilt the negative keyword lists at both the account and campaign level (roughly 2,400 negatives added over the first quarter), tightened match types, deployed call bid adjustments for business hours plus the after-hours emergency window, and configured proper conversion goals for booked jobs rather than form submissions or calls under 60 seconds. PMax got explicit audience signals and a fenced Merchant-Center-style feed for install-specific creative.

3. Google Business Profile overhaul across four locations

We rebuilt each profile from the categories up. Primary and secondary categories were wrong in three of four locations. Service areas were either too broad (dilutes ranking) or too narrow (misses ZIP codes with real demand). Hours, holiday hours, and the service list were incomplete. We pre-seeded the Q&A section with the actual questions callers asked most often, then answered them in the brand voice. We deployed weekly Google Posts (a mix of financing offers, seasonal reminders, and case-style project photos) and stood up a review generation flow through Podium that hit customers 45 minutes after a completed job while the technician was still fresh in memory.

4. Site rebuild for CWV, local SEO, and structured data

The existing site was on a legacy WordPress theme with a Largest Contentful Paint above 4 seconds on mobile and a Cumulative Layout Shift that failed on every landing page. We migrated to a lean custom build on the same WordPress core (kept the CMS the office staff already knew), then rebuilt the template layer for performance. We shipped per-location landing pages with genuinely different content per city (real testimonials from that market, actual project photos, the technicians who cover it, the ZIP codes served), plus a service-city grid so every service-metro combination had a dedicated URL. We layered on the schema stack: LocalBusiness for each location, Service for each service line, FAQPage on the answers, Speakable for voice search, and ImageObject on the project photo library.

5. HubSpot activation for lifecycle

HubSpot had been a contact bucket. We turned on the two workflows that mattered most for a residential HVAC book. First, a repair-to-install conversion sequence: every customer whose system was over ten years old at the time of a repair got a 6-week email drip on the economics of replacement (energy costs, financing math, warranty comparison, tax credit availability), with the technician's actual next-visit appointment as the CTA. Second, a maintenance plan enrollment sequence for anyone whose first call was a repair, with a discount tied to enrolling before the next seasonal changeover. Both sequences were plain-text, from the local branch manager's actual email address, and none of them mentioned marketing automation.

6. Content engine for AEO and GEO

We wrote long-form guides on the questions homeowners actually ask, structured for the answer engines: "AC not cooling, what to check first," "heat pump vs. gas furnace in the Carolinas," "tonnage sizing for a 2,400 square foot home," "how often to change your filter," "signs your ductwork is leaking." Each piece opened with a direct-answer TL;DR (60 to 90 words), used FAQPage schema on the subheads, and included Speakable markup on the summary. Within six months the guides were cited in Google's AI Overviews and in Perplexity answers for high-intent local queries, which fed a compounding organic layer that the paid channels could not produce on their own.

The numbers

Timeline, team, budget

• Timeline: 18 months, structured as six 3-month sprints with a stakeholder review at each boundary.
• Team: One strategist (fractional CMO role), one paid-media specialist, one web designer, one long-form content writer, working part-time and coordinated through a shared HubSpot workspace.
• Retainer band: $18K to $22K per month, plus paid-media spend passed through at cost.
• Tools deployed: HubSpot Marketing Hub, CallRail, Google Ads, Google Business Profile, Podium, Google Search Console, GA4, Google Tag Manager, Ahrefs, Screaming Frog, Lighthouse, PageSpeed Insights, Merchant Center, WordPress (kept), custom theme (built).

2. How discovery works in HVAC in 2026

Residential HVAC is one of the most competitive local-service categories in the United States, and the discovery landscape has shifted more in the last three years than in the fifteen before that. A homeowner in Charlotte or Charleston or Greenville looking for an HVAC contractor in July 2026 encounters a discovery flow that looks nothing like the phone-book-and-yard-sign world the industry grew up in.

The intent split

HVAC search intent divides sharply between emergency and considered. Emergency intent ("AC not cooling," "furnace stopped working," "heat pump making noise") is short, mobile, phone-call-heavy, and often happens outside business hours. It rewards operators who show up in the top three of the map pack, have a live phone answered by a human, and offer same-day service. Considered intent ("heat pump vs gas furnace," "how much does a new HVAC system cost," "SEER rating explained") is longer, desktop-friendly, research-heavy, and happens weeks or months before a purchase. It rewards operators who publish real answers to real questions, with structured data that lets AI assistants extract those answers cleanly.

An HVAC operator who optimizes only for one of those two intents leaves half the market. Emergency-only operators become dependent on paid ads and lose the compounding organic layer that considered content produces. Considered-only operators write blog posts nobody clicks on because the operator has no visibility in the map pack when the emergency happens. The right posture treats emergency as the demand-capture floor and considered as the demand-generation compounding layer, and staffs both.

The seasonality tax

HVAC demand is bimodal and violent. In the Carolinas the summer AC repair season runs mid-May through late September, with peak spend and peak margin in July. The winter heat season runs late November through early March, with peak in January. Between those two peaks are two shoulder periods where demand collapses by roughly 55 to 65 percent versus peak. The operator who does not plan for this pays for it two ways: paid-media spend gets wasted in shoulder months buying clicks that cannot fill a schedule, and the sales team gets over-hired in peak months and under-worked (or laid off) in shoulders. Sophisticated operators use the shoulder months for maintenance plan enrollment (converting one-time customers into recurring revenue) and for indoor-air-quality upsells (whole-home dehumidifiers, air purifiers, UV lights) that carry decent margin and are less weather-dependent. The marketing calendar should mirror this: heavy paid media in peak, heavy content and lifecycle in shoulder.

The discovery surfaces homeowners actually use

Ranked by frequency of first-touch for the operator we worked with (measured through call attribution over 18 months):

1. Google Business Profile (~38% of attributed first touches). The map pack is where emergency and near-emergency queries resolve. The homeowner types "AC repair near me" and clicks one of the three profiles that appears. Ranking outside the top three costs roughly 90% of that click volume.
2. Google Search organic (~22%). Long-tail queries with informational and commercial intent, including comparison searches ("best HVAC company Raleigh"), spec searches ("Trane vs Carrier"), and the long-form guide traffic.
3. Google Ads (~19%). Emergency queries with commercial modifiers where the operator wants to buy the top of the SERP above the map pack. High intent, high cost, high return when structured correctly.
4. Referral / word-of-mouth (~11%). Underrated. The single strongest driver of install-tier revenue, but hard to influence at the top of funnel.
5. AI Overviews / Perplexity / ChatGPT / Gemini (~5%, growing). Small in absolute volume in 2026 but doubling every six months. Cited operators appear in AI answers to comparison and how-to questions and skip the SERP entirely.
6. Facebook / Nextdoor / Angi / Thumbtack (~4%). Directory-style discovery with variable quality. Nextdoor is the strongest of the four for residential HVAC.
7. Everything else (~1%). Yard signs, truck wraps, print, radio. Still nonzero, but rounding error against the digital surfaces.

The margin gap that makes marketing look wrong when it is right

The economics of HVAC create a specific optimization problem. A repair call generates $200 to $700 in revenue on a 90-minute truck roll. An install job generates $6,000 to $14,000 on a full-day two-technician crew. The install has 3x to 4x the margin per hour of technician labor. Every HVAC marketing engagement worth doing has to shift the mix toward installs, not just increase call volume. Adding 20% more repair calls to a fleet that is already at capacity does not help the business. Converting 20% of repair calls into installs, or moving a repair call scheduled for next week into an install decision this quarter, does.

This is why the lifecycle work in the case above (repair-to-install email sequence) mattered more than the paid-media work in year two. Paid media brought calls in the door. Lifecycle converted the calls that were already there into higher-margin work. An HVAC operator whose marketing dashboard reports "leads" as the top-line metric is optimizing for the wrong thing. The right top-line metrics are installed system revenue, install ticket average, install close rate on qualified opportunities, and maintenance plan attach rate.

The financing dynamic

Most residential HVAC installs are financed. The homeowner facing a $9,000 to $14,000 cash outlay is a very different buyer from the same homeowner presented with a $120 to $180 per month financed plan. Operators partnered with Synchrony, GreenSky, Wells Fargo Retail Services, or Service Finance close installs at meaningfully higher rates than operators who only quote cash. The marketing implication: financing math belongs on the site, in the ad copy, in the estimator, and in the lifecycle emails. The single biggest CRO lever we found on the HVAC site was a financing calculator on install landing pages that let the homeowner see the monthly payment before ever calling.

The licensing and trust layer

HVAC is a regulated trade in every state, with license requirements that vary meaningfully. In North Carolina the H-3-I (heating group III, class I) license is required for full residential HVAC work. In South Carolina it is the mechanical contractor license under the SC LLR. Both states require the license number to appear on advertising. A site that surfaces the license number, the technician certifications (NATE, EPA 608), the manufacturer authorizations (Trane Comfort Specialist, Lennox Premier Dealer, Carrier Factory Authorized), and the local BBB accreditation reads as legitimate. A site that hides them or omits them reads as fly-by-night. In an industry with a reputation problem, this trust layer is the difference between showing up in results and being clicked once you show up.

The technician labor market

The single biggest constraint on HVAC growth in 2026 is not demand, it is technician supply. The industry is short an estimated 55,000 to 80,000 technicians nationally, and the average tenure of a new hire in the first 12 months is roughly 40 to 50 percent. Every marketing engagement worth doing has to account for this. Adding call volume that cannot be serviced does not grow the business, it degrades the customer experience and the online reviews that follow. Sophisticated operators run recruiting marketing alongside customer marketing (careers pages that rank, LinkedIn recruiting, apprenticeship pipelines with local trade schools), because the growth ceiling is set by how many good technicians you can attract, hire, train, and retain, not by how many leads you can generate.

The private-equity roll-up dynamic

Since roughly 2019, home-services private equity has consolidated the HVAC market at an accelerating pace. Regional and national roll-ups are buying independent operators at 5x to 8x EBITDA multiples, then unifying branding, technology, and back-office operations under a holding company. This changes the competitive landscape in two ways. First, the operator competing in your metro next year might be a rebranded PE portfolio company with a much larger marketing budget than yours. Second, the independent operator's positioning as "family owned, been here for 30 years, know the neighborhoods" becomes more valuable, not less, as the market consolidates and homeowners react against the private-equity-owned brands. The marketing implication: independent operators should lean into local identity, real owner presence, and community depth as differentiation, not just service quality (which is table stakes).

The manufacturer dealer program dynamic

Every major HVAC manufacturer runs a tiered dealer program. Trane Comfort Specialists, Lennox Premier Dealers, Carrier Factory Authorized Dealers, Bryant Factory Authorized, American Standard Customer Care, York Certified Comfort Experts, Rheem Pro Partners, Goodman Comfort Alliance. These programs come with three benefits that matter for marketing: extended warranty offers (typically 10 years parts and 10 years labor versus the base 5/1), co-op marketing dollars (typically 2 to 5 percent of eligible equipment purchases returned as marketing credit for pre-approved spend), and trust signals for the SERP (a "Trane Comfort Specialist" badge on the site converts install visits at a measurably higher rate than an unbadged Trane dealer).

The marketing implication is direct. The dealer badge belongs on the homepage, in the footer, on install landing pages, and in the Google Business Profile description. The co-op dollars belong inside the paid media budget as a distinct line item, not folded into general spend (using co-op for Google Ads requires pre-approval by the manufacturer and specific creative rules, but the ROI on co-op-funded spend is meaningfully better because the effective cost is roughly half). The extended warranty belongs in the ad copy, on the estimator, and in the lifecycle emails, because it is a real differentiator against a competitor who does not have the dealer status. Independent HVAC operators who ignore their dealer program are leaving between 4 and 8 percent of their annual marketing budget on the table in unused co-op alone, and losing conversion on install pages that fail to display the trust signal a competitor with the same badge is using.

The review economy in HVAC

An HVAC operator's Google review count is the single most-referenced trust signal in the industry. Homeowners scroll the map pack, see three profiles, and click the one with more reviews and a higher aggregate rating. The rough rule: below 100 reviews the operator is not seriously in the game, 100 to 300 is competitive in most metros, 300 to 800 is dominant in secondary metros, and 800+ is the bar in top-25 metros. Getting from 40 to 220 reviews in twelve months (which the case above accomplished) requires a systematic flow, not a monthly "please leave us a review" email.

The mechanics that work: an SMS review request from the technician's own phone within 45 minutes of job completion, a direct link to the Google review URL (not a landing page that redirects), an offer to help draft the review if the customer is unsure how to write one, and a rapid response cadence on every review (positive or negative) from a named person at the company. Response rate matters. Profiles that respond to 95 percent or more of reviews within 48 hours rank higher than profiles that ignore reviews, because Google reads response cadence as an activity signal.

Negative reviews are the trickiest lever. The instinct is to fight or to ignore. The right move is to respond professionally, offer to resolve the issue offline (never in the review thread), and never argue on the profile. A well-handled negative review with a professional response converts better than no negative reviews at all, because it demonstrates the operator handles problems in the open. Homeowners are suspicious of profiles with only 5-star reviews and no responses. The healthy ratio for a real, working HVAC operator is roughly 92 to 96 percent 5-star, 3 to 6 percent 4-star, and 1 to 3 percent 3-star-and-below with visible professional responses. Anything above 98 percent 5-star reads as manipulated.

The recruiting-marketing parallel

The single biggest constraint on HVAC growth in 2026 is not customer demand, it is technician supply. The industry is short an estimated 55,000 to 80,000 technicians nationally, and the average tenure of a new hire in the first 12 months is roughly 40 to 50 percent. Every marketing engagement worth doing has to account for this. Adding call volume that cannot be serviced does not grow the business, it degrades the customer experience and the online reviews that follow. Sophisticated operators run recruiting marketing alongside customer marketing (careers pages that rank for "HVAC technician jobs [metro]," LinkedIn recruiting, apprenticeship pipelines with local trade schools, retention marketing to technicians already on staff), because the growth ceiling is set by how many good technicians you can attract, hire, train, and retain, not by how many leads you can generate. In the engagement above we allocated roughly 15 percent of the total marketing budget to recruiting marketing in year two, and the operator was able to grow tech headcount by 22 percent (from 65 to 79) while holding the industry-typical churn rate. Without that, the customer marketing would have generated demand the fleet could not serve.

3. The Ranking Surfaces Playbook, applied to HVAC

The Ranking Surfaces Playbook covers 13 distinct discovery surfaces. Not all 13 matter equally for HVAC. Below is the priority order and why, based on what actually moved the needle over the 18-month engagement above and across other HVAC work I have touched. Each surface gets an assessment of how much revenue it can move for an HVAC operator, what to actually build, and what to skip.

Tier one: the surfaces that produce revenue this quarter

LSO (local search) — the highest-leverage surface for HVAC

Nothing else in the Playbook produces revenue as fast for an HVAC operator as fixing the Google Business Profile. Categories, service areas, review velocity, and post cadence are the four levers. Getting the primary category right (Heating Contractor or Air Conditioning Contractor, not the generic HVAC Contractor) and adding the correct secondary categories can move a profile from page two to page one in 60 days for a metro with average competition. From page one to the top three takes review velocity and proximity. There is no substitute for asking every completed job for a review within 45 minutes of the technician leaving. Beyond the profile itself, LSO for HVAC also covers local citations (Yelp, Angi, HomeAdvisor, BBB, chamber of commerce, industry directories like Nextdoor and NextInsurance), consistent NAP (name, address, phone) across every citation, and a local schema layer on the website with LocalBusiness on the location pages, HomeAndConstructionBusiness as a secondary type, and the areaServed property populated with real ZIP codes rather than generic city names.

SEO — per-location and service-city grid

The site architecture that ranks for HVAC in 2026 is a grid: one page per service line, one page per metro, and a unique URL for every service-metro intersection. "AC repair Raleigh" is a different page from "AC repair Charlotte" and both are different from "heat pump installation Raleigh." Each page needs genuinely different content, not just a swapped city name. Real testimonials from that metro, real project photos, real technician profiles, real ZIP codes served, real Google Maps embed. This grid is what lets an HVAC company rank organically for the long tail of "[service] [neighborhood]" and "[service] near [landmark]" queries. Technical SEO for HVAC also means clean crawlability (no orphan pages, sitemap segmented by page type, robots.txt that does not block CSS or JS), fast internal linking (every service page links to every location page and vice versa), and a URL structure that mirrors the information hierarchy rather than the CMS default.

CWV — because HVAC traffic is mobile-first and emergency-intent

The homeowner in the crawl space with a leaking condensate line is on a phone. If the page takes 4 seconds to render they call the next result. Core Web Vitals were a nice-to-have three years ago and are a ranking factor now. LCP under 2 seconds on 4G, CLS under 0.1, INP under 200ms. The site rebuild in the case above moved every metric into the green and the ranking gains followed within one quarter. Practical HVAC CWV work: inline critical CSS, defer non-essential JavaScript, preload the hero image and the branded logo, host web fonts locally with font-display: swap, compress every image with modern formats (WebP or AVIF), and remove the auto-playing hero video that the marketing agency added because "it looks impressive." It does not look impressive when the page takes 6 seconds to render on the homeowner's phone.

Tier two: the surfaces that compound

AEO — direct-answer content on the questions homeowners actually ask

The considered-intent side of HVAC (research-heavy queries weeks before purchase) is where AEO wins. TL;DR paragraphs that answer the question in 60 to 90 words, FAQPage schema on subheads, spec tables when the answer involves numbers. Google's AI Overviews cite content structured this way disproportionately, and the citation carries a click-through rate roughly 2x higher than a classical top-three organic result. The mechanics for HVAC: pick the 30 to 50 highest-intent informational queries in the vertical (comparison queries, cost queries, how-to queries, spec queries), write a long-form guide for each with a strict answer-first structure, include a spec table with real numbers, mark the entire piece up with FAQPage schema on the subheads and Speakable on the summary paragraph. Update quarterly. The compounding effect is real: within six months of shipping a good AEO layer, the organic traffic mix starts to shift from long-tail money keywords toward informational queries that convert on longer sales cycles but at higher install ticket averages.

GEO — being cited inside ChatGPT, Perplexity, Claude, and Gemini

LLMs increasingly answer HVAC comparison and how-to questions directly. The operators cited in those answers get a growing share of the informational-intent traffic that used to land on a blog post. The mechanics are the same as AEO but with an additional layer: brand entity clarity (sameAs across web, LinkedIn, Wikidata if the company is notable enough), attributable numbered facts, and a llms.txt file that tells the crawlers what is here. For an HVAC operator that means an Organization schema block with sameAs pointing to the Google Business Profile URLs, the LinkedIn company page, the Facebook business page, the BBB profile, and the manufacturer authorization pages (Trane, Lennox, Carrier list authorized dealers publicly). It also means writing content with attributable numbers ("SEER 16 typically saves 12 to 18 percent versus SEER 14 for a typical 2,400-square-foot home in the Carolinas") rather than generic claims. LLMs reward attribution.

E-E-A-T — the trust layer that lifts every other surface

NATE certification, license numbers displayed on every relevant page, response-time promises with real median data, technician profiles with real photos and real bios, project photos with location and date metadata, review response cadence with the owner or manager's actual name. E-E-A-T is not a single ranking factor. It is the trust layer that lifts every other surface and it is disproportionately important for HVAC because of the industry's reputation problem. The specific E-E-A-T deliverables for HVAC: an About page that names the owner and the year founded, individual technician profile pages with photo and certifications, a licenses page listing every state license number, a service area page listing every ZIP served, a project gallery with dated location-tagged photos, and a reviews page that surfaces the aggregate rating with source attribution to Google and Facebook rather than just cherry-picked testimonials.

Tier three: the surfaces worth doing but with lower ROI in HVAC

VxSO — Google Lens for HVAC is small but present

ImageObject schema on project photos, alt text that describes the actual system installed (not "HVAC unit"), photos indexed in Google Images and Pinterest Lens. Small volume but the intent is high: someone reverse-searching a photo of a compressor already has a real problem. The build cost is low (adding ImageObject schema and better alt text to an existing photo library takes a few days), so the ROI is fine even at low absolute volume.

VSO — voice search for HVAC is real, small, and growing

Speakable markup on FAQ answers, natural-question subheads. Homeowners increasingly ask their phones or their smart speakers HVAC questions, and Speakable is the mechanism that gets your answer read aloud. Small in 2026, larger in 2028. The build cost is negligible if you are already doing AEO (Speakable and FAQPage often mark up the same content), so there is no reason not to do it.

ASO — only if the operator has an app

Not applicable for most HVAC operators. A minority of operators (typically the roll-up-owned brands) have consumer-facing apps for scheduling, payment, and maintenance plan management. If that describes your operation, ASO applies (keyword optimization on the App Store and Play Store, screenshot copy, review generation inside the app). Otherwise skip.

Tier four: not a fit for HVAC

KGO, GLOBO, Web3

Knowledge graph entry matters for national brands. Most regional HVAC operators are not notable enough for Wikidata inclusion, and forcing it before organic notability produces a rejected entry. International locale ranking does not apply to a US-only service business. Web3 identity (ENS, Farcaster) is a first-mover opportunity for national and consumer brands, not for a residential HVAC operator with a 45-minute service radius.

AAO — the honest answer is "not yet, but soon"

Agentic search (AI agents booking appointments on your site) is emerging in adjacent categories (travel, hospitality, dining) and will reach home services eventually. In 2026 it is not yet producing meaningful volume for HVAC. Worth deploying llms.txt v2 and PotentialAction schemas as a first-mover play (cheap to do) but do not expect measurable revenue from it this year. The operators who deploy the AAO stack in 2026 will be positioned for the volume shift when it arrives, likely in 2027 to 2028 based on the pace in adjacent verticals.

How Playbook priority shifts by operator size

The 13-surface Playbook does not apply identically to every HVAC operator. A solo owner-operator running one truck out of a residential garage has different priorities than a $15M multi-location business. Rough guide by revenue band:

Solo / 1-2 trucks (under $1M annual): LSO is the entire game. Fix the Google Business Profile, get to 100 reviews, be findable in the map pack. SEO matters but a well-structured 6- to 10-page site with the right schemas is enough. Skip AEO, GEO, and everything below it until revenue justifies the investment. Total marketing spend should be 6 to 10 percent of revenue at this scale, weighted heavily toward paid search on emergency queries and reputation management.

Small ($1M-$5M, 3-8 trucks): Add technical SEO with per-service pages and per-service-city pages if the operator serves more than one metro. Layer a light content engine (6 to 12 long-form guides per year on the highest-intent informational queries). The attribution stack becomes essential because paid media is now a real budget line, not petty cash. Consider one call-tracking number per major channel. CWV becomes worth the site rebuild investment. E-E-A-T deliverables (licenses page, technician profiles, certifications displayed) should be built out.

Mid ($5M-$25M, multi-location): The full Playbook makes sense at this scale. This is where LSO + SEO + CWV + AEO + GEO + E-E-A-T all pay back within 12 to 18 months. Content becomes a compounding asset that outpaces paid on a 24-month view. HubSpot or ServiceTitan lifecycle becomes worth the investment. The measurement stack (see below) becomes non-negotiable. Recruiting marketing enters the mix at 10 to 20 percent of total marketing budget.

Regional ($25M+, multi-metro or state-crossing): Add KGO (Wikidata entity, Knowledge Panel work), ASO if the operator has a consumer-facing app, GLOBO if crossing state lines meaningfully (each state gets its own hreflang-style locale variant of the site, with state-specific license and regulatory content). AAO becomes a real first-mover play worth 5 to 10 percent of the content budget. Custom software (proprietary online estimator, virtual site survey tool, homeowner portal) may be worth building. At this scale the marketing team is 8 to 15 people internal plus agency support, and the marketing budget is 3 to 5 percent of revenue (lower percentage than smaller operators because the operational leverage is higher).

The measurement stack for a multi-surface HVAC engagement

For a multi-surface engagement the measurement stack has to reflect the surface complexity. Skimping here is why most HVAC marketing engagements fail to prove ROI even when the underlying work is good. The stack I deployed on the case above:

GA4 as the base analytics layer, configured with proper events (call_click, form_submit, estimator_start, estimator_complete, financing_click, booking_complete, chat_open). Enhanced measurement enabled. Cross-domain tracking across the main site and any subdomain landing pages. Custom conversions mapped to install revenue rather than lead count.

Google Search Console segmented by property (each subdomain gets its own property, each language variant gets its own property if applicable). Not merged into GA4 for reporting because the discrepancy between GSC clicks and GA4 sessions matters and gets lost in a merged view.

CallRail as the phone attribution layer, with unique numbers per channel and dynamic number insertion (DNI) on the site. Call transcription enabled on booked calls only (privacy) to feed keyword insights back into paid.

HubSpot as the CRM and lifecycle engine, with proper contact source mapping so every contact record knows which channel produced the first touch. Deal stages mapped to the actual sales process (Quote Requested → Quote Sent → Quote Accepted → Job Scheduled → Job Completed → Review Requested).

Looker Studio as the executive dashboard layer, pulling from GA4 + CallRail + Google Ads + HubSpot into one weekly view the owner can read in three minutes. Not built to impress marketing, built to answer the owner's three questions: which channels made money this week, what is the ROAS by channel, and what is the install pipeline for the next 30 days.

Cost caps by surface: paid media budgeted at 3 to 4 percent of trailing 12-month revenue as a hard ceiling, allocated roughly 60/25/15 across Google / Bing / Meta. SEO and content budgeted separately at 1 to 2 percent. Local citation management as a fixed annual line ($1,200 to $2,400 depending on scale). Software stack (HubSpot, CallRail, Ahrefs, Screaming Frog, Podium) as a fixed monthly line typically running $1,800 to $3,500 for a mid-size HVAC operator.

How the surfaces combine in practice

The Playbook works because the surfaces compound. LSO drives the map pack traffic that produces most emergency-intent calls. SEO drives the organic long-tail that produces most considered-intent traffic. CWV lifts both. AEO and GEO capture the AI-answer layer that increasingly sits above both. E-E-A-T is the trust signal that lifts every one of those. VSO and VxSO capture the marginal-but-real voice and visual layers. Skipping any one of them leaves margin on the table, but skipping the top three (LSO, SEO, CWV) means the rest of the stack does not have a foundation to compound on.

A concrete example of the compounding: in the HVAC case above, month three saw the LSO overhaul move the Charlotte GBP from position 6 to position 2 in the map pack. That alone would have been a win. But because the site rebuild (CWV work) had shipped in month two, the increased GBP click-through fed a site that could actually convert (LCP under 2 seconds versus the old 4-plus). And because the per-location landing page for Charlotte had already been indexed with clean LocalBusiness schema (SEO work), the organic ranking for "AC repair Charlotte" moved from page two to position four in parallel. Three months later the AEO layer began being cited in AI Overviews for "how much does AC repair cost Charlotte," pulling considered-intent traffic that the map pack alone would never have reached. Each surface amplified the others. Running them in sequence rather than in parallel would have delayed the compounding effect by roughly six months.

The 18-month HVAC engagement above pulled 8 of the 13 surfaces meaningfully. Two more (AAO and KGO) would have made sense for a larger operator with the notability signal to support them. Three (ASO, GLOBO, Web3) were not applicable. That distribution (8 of 13 pulled, 2 aspirational, 3 not applicable) is typical for a regional home-services operator. A national brand or a SaaS company would pull a different subset. The framework's value is that it makes the trade-offs explicit rather than defaulting to whichever surface the vendor happens to specialize in.

4. What most HVAC operators get wrong

Across the HVAC engagements I have worked on and the audits I have run for operators before deciding whether to take them on as clients, the same seven mistakes appear over and over. If you are running an HVAC business and reading this, some of these will land uncomfortably. That is the point.

1. Buying paid clicks without attribution

The single most common HVAC marketing failure is spending on Google Ads with no phone tracking on the resulting calls and no CRM record of which calls became jobs. The account looks like it is working because clicks and calls come in. Without attribution the operator cannot see that 30 to 40% of the spend is producing calls that never book, calls from existing customers, or calls that get scheduled for two weeks out and then cancel. Fixing attribution before touching the ads unlocks a reallocation that pays for the rest of the work.

2. Treating Google Business Profile as "set and forget"

Most HVAC operators claim their GBP, fill in the basics, and never touch it again. In 2026 the profile is a living surface. Weekly Posts, monthly Q&A updates, service list edits when service lines change, and constant review generation are the operating rhythm. A profile that has not been updated in six months looks abandoned to Google's ranking algorithm and to the homeowner scrolling the map pack.

3. One website landing page for every city

The operator serves five cities. The website has one "service areas" page with a bulleted list. That page ranks for zero of the five cities. The right answer is five separate landing pages with genuinely different content, real testimonials from each metro, real technician profiles for the crew that covers that market, and real project photos from work done there. Duplicated content across five URLs is worse than one URL.

4. Chasing more calls when the fleet is already at capacity

An HVAC operator whose techs are booked out three weeks does not need more calls. They need higher-margin calls. Optimizing the mix (repair-to-install conversion, maintenance plan enrollment, higher-tier system upsells at replacement) produces more revenue with the same fleet. Adding call volume to a booked-out fleet just extends the schedule and increases cancellations.

5. Ignoring reviews because "we get some naturally"

An HVAC operator doing 900 jobs a month who gets 8 organic reviews a month is capturing under 1% of possible reviews. A review generation flow that asks every completed job at the 45-minute mark captures 15 to 25% of jobs as new reviews. That is a 20x difference in review velocity, which compounds in the map pack ranking, in the click-through rate on the profile, and in the trust signal on the SERP.

6. Writing blog posts nobody clicks

The HVAC blog is usually 300-word posts on "5 tips for a healthy AC" written by a marketing intern. Nobody clicks them, they do not rank, and they do not get cited by AI. Long-form guides (2,500 to 4,000 words) that actually answer real homeowner questions, with direct-answer TL;DRs, spec tables, and FAQ schema, will do more for organic visibility in six months than three years of the tips posts.

7. Treating marketing as separate from operations

The best HVAC marketing lever is a technician who leaves the job site having done exceptional work, and the second-best lever is that technician asking for a review before they pull out of the driveway. If the marketing team never talks to dispatch, the ops team, or the technicians, the marketing is running blind. The engagement above worked because we sat in on the weekly ops meeting for the first six months.

5. Frequently asked questions