One-line takeaway

More traffic cannot repair an unclear booking path, an unowned lead, or an unknown completed-job outcome.

Source-backed signal

Noah Igler's video, “I Made a Home Service Business $1.27M in 90 Days (Hormozi's Playbook),” presents a self-reported HVAC marketing case study. Igler shows a ServiceTitan dashboard with $1.27 million in total revenue for the preceding 90 days and says almost all of that revenue was tied to search and paid traffic. He says the company's Google Ads had previously exceeded $600 per lead, later fell to about $150 per lead depending on the service, and reached 6.9 times return on ad spend during the 90-day period.

His operating sequence is conversion first, then local and AI-search visibility, then paid advertising. He treats the website experience as the offer and recommends clearer outcome language, visible booking options, reduced form effort, prompt response, and a measurement shift from raw lead cost toward customer economics. He also recommends separating campaigns by service category rather than valuing every inquiry alike.

Source boundary: the video is a marketing practitioner's account and sales presentation, not an independent case study. It does not disclose the client, ad spend, lead and customer counts, booked and completed rates, attribution model, organic-versus-paid allocation, gross profit, prior-period revenue, membership or repeat revenue, or the records behind the claimed return. The displayed dashboard and presenter's statements support attribution of the claims to Igler; they do not independently establish that the marketing work caused the reported total revenue or profit.

Google Ads' official offline-conversion guidance distinguishes an online inquiry from later qualified-lead and converted-lead events. It recommends separate conversion actions for different funnel stages rather than treating every lead event as the same outcome. The documentation also describes passing later offline outcomes back into advertising measurement.

Google Business Profile's official review guidance says businesses may ask customers for reviews with a direct link or QR code. It also says reviews must reflect genuine experiences and prohibits offering incentives in exchange for posting, changing, or removing reviews.

Those official sources establish two useful capabilities and boundaries: marketing measurement can continue after a form submission, and a compliant review request can be designed as a real post-service handoff. They do not prove that a website is the main constraint at every HVAC company, prescribe one booking workflow, or establish a universal conversion rate, response-time target, channel order, lead cost, or customer value.

The video's website-conversion percentages, lead-volume multiplier, review-volume target, LSA close-rate range, local-ranking prescriptions, and AI-search recommendations are presenter claims or recommendations. They are not treated here as industry benchmarks or platform rules.

Interpretation

Interpretation, not a sourced fact:

An HVAC website is part of the booking operation.

The page can make a clear promise, show an appropriate next step, and collect a request. But the conversion path does not end when a homeowner taps a button. Someone or something still has to recognize the request, determine whether the work fits the service area and availability, respond through the permitted channel, book the right job type, and preserve enough source information to connect the completed work with the marketing decision that created it.

That makes “we need more leads” an incomplete diagnosis.

More traffic may be useful when the current path is understood and the operation has capacity. It can also feed more homeowners into an unclear form, unavailable appointment window, missed call, delayed response, unqualified job type, or attribution gap. A larger lead count does not reveal which of those outcomes occurred.

Cost per lead has the same limitation. It prices the arrival of an inquiry. It does not, by itself, show whether the inquiry was qualified, booked, completed, profitable, converted into an ongoing relationship, or lost because the company could not respond. Lifetime value has the opposite risk: it can make an acquisition channel look attractive by assigning future value that the company has not actually observed.

That distinction matters in this source. Igler explicitly labels his LTV-to-CAC example as an illustration rather than the featured company's data. The example compares lead cost with average job-ticket revenue, then describes the quotient as LTV to CAC. An operator cannot reuse that ratio without a known lead-to-customer rate, acquired-customer cost, direct job cost, gross profit, repeat-work horizon, and retention evidence. Replacing cost per lead with an unsupported lifetime-value estimate does not improve the decision.

The useful first move is not to declare traffic too expensive or the website broken. It is to trace a fixed set of recent requests through the full handoff and keep unknown outcomes visible.

Operator lesson

Treat the website, office response, dispatch rules, field outcome, and marketing record as one connected system.

For each meaningful website action, define four things:

  • **State:** What happened—requested, contacted, qualified, booked, completed, canceled, declined, duplicate, or unknown?
  • **Evidence:** Which timestamp, call record, form record, booking record, job record, or customer decision supports that state?
  • **Owner:** Which role is accountable for the next action or exception?
  • **Next action:** What should happen, by when, and through which permitted channel?

Then separate the handoffs:

  1. **Search or referral to website:** Which service, location, or need brought the homeowner to this page? Do not assume every direct visit or branded search belongs to the last paid click.
  2. **Website to request:** Is the next step clear on mobile? Can the homeowner call, request service, or view an honest booking option without decoding generic language?
  3. **Request to human response:** Did the office receive the request? Is the response state visible across phone, form, chat, and SMS rather than split into unowned inboxes?
  4. **Response to booked job:** Was the request in area, in scope, and matched to a real appointment or documented disposition?
  5. **Booked job to completed work:** Did the job run, cancel, reschedule, or turn into a different job type? Keep a booking separate from completed revenue.
  6. **Completed work to known economics:** What revenue and direct job-cost information is available? If gross profit, membership enrollment, repeat work, or replacement opportunity is unknown, report it as unknown.
  7. **Completed experience to review request:** Was a compliant request made after a genuine customer experience, without rewarding positive sentiment or suppressing negative feedback?

The website can reduce friction at the front of that system. It cannot promise appointment capacity the dispatch board does not have, repair an unanswered phone, choose a qualification policy, or create trustworthy profit and attribution records after the job. Those are operating decisions with marketing consequences.

Practical playbook

  1. Choose one service category and one meaningful market. Do not mix emergency repair, maintenance, replacement, and every service area in the first audit.
  2. Select a fixed set of recent website requests. This is a diagnostic trace, not a performance benchmark.
  3. For every request, record the landing page or known source, request time, requested service, permitted response channel, first response, qualification state, booking state, completed-job state, revenue state, job-cost state, membership outcome, and next owner.
  4. Allow `unknown` in every field. Do not turn a blank into zero, a missed disposition into lost, a scheduled job into completed revenue, or an unattributed customer into organic demand.
  5. Reproduce the mobile path yourself. Check whether the service promise, service area, availability language, call action, request form, and confirmation state are clear and accurate.
  6. Compare the website promise with real booking rules and capacity. Remove or qualify any “same day,” emergency, price, financing, or availability language the operation cannot consistently support.
  7. Trace calls, forms, chat, and SMS into one owned response view. Define allowed dispositions and an escalation path for requests that remain unacknowledged or unresolved.
  8. Find the earliest repeated unknown or failed handoff. Change one element there before changing channel spend: page clarity, form friction, call routing, response ownership, booking rule, job-type mapping, or completed-work attribution.
  9. Keep marketing stages separate. At minimum, count requests, qualified requests, booked jobs, completed jobs, and completed jobs with known economics. Show each numerator, denominator, and unknown count.
  10. Feed later outcomes back into the channel review only after consent, data handling, identifiers, deduplication, and conversion definitions are documented. Test imported outcomes before using them to automate bidding decisions.
  11. If reviews are part of the loop, ask every eligible customer for honest feedback through a direct link or QR code. Reward completion of a compliant request process, if anything—not positive sentiment.
  12. Scale traffic only when the affected handoffs have visible capacity, owned exceptions, and a measurement path that reaches a business outcome the company can actually verify.

Do not use this audit to justify invasive tracking, upload customer information without an approved basis, pressure a homeowner for a positive review, or automate consequential ad-budget changes from partial data. Measurement design needs the company's privacy, consent, access, retention, and financial-record rules.

Email version

Your HVAC website is part of the booking operation.

A clear page and an easy request form matter, but the handoff continues after the click. The request still needs a known response state, a qualification decision, a real booking, a completed-job outcome, and enough source and financial evidence to inform the next marketing decision.

Trace one service category through that path before buying more traffic. For each recent request, record the state, evidence, owner, and next action. Keep unknown separate from zero. Keep a form submission separate from a booked job, and a booked job separate from completed work.

Then fix the earliest repeated break—page clarity, form friction, call routing, response ownership, booking rules, or completed-work attribution.

The next useful count is qualified website requests with a known booking outcome, completed-job outcome, and source.

Where does website demand most often become unknown in your operation?

LinkedIn post

An HVAC website is not a brochure.

It is one handoff in the booking operation:

Search or referral → website action → office response → qualified request → booked job → completed work → known economics.

At every stage, record:

  • State
  • Evidence
  • Owner
  • Next action

Keep unknown separate from zero. Keep a lead separate from a booking. Keep a booking separate from completed revenue.

More traffic can scale a working path. It can also feed an unowned queue faster.

Before changing spend, trace one service category from request to completed work and fix the earliest repeated break.

Short post / thread starter

More HVAC traffic is not automatically more demand. Trace the path from website request to booked job to completed work. The first repeated unknown is a better starting point than cost per lead.

Community-answer suggestions

  • Ask which service category and market the operator wants to trace first.
  • Separate website friction from office response, booking capacity, and attribution gaps.
  • Ask for the count of requests with known qualified, booked, and completed outcomes before discussing more spend.
  • Keep gross profit and customer lifetime value unknown until the company's own records support them.
  • Treat review requests as a compliant operating process, not a positive-sentiment incentive.

Sources

One useful decision at a time

Get the next issue

Reader question

At which handoff does your current website demand most often become unknown?