HVAC Marketing Plan: A 12-Month Template With Numbers

The table below is the template specification. Build it in a spreadsheet with one tab per section, or in a single sheet with these eight blocks stacked. The "Where the number comes from" column is the part that separates a plan from a wish list.

Sections 1 through 3 are the plan. Sections 4 through 6 are the execution. Sections 7 and 8 are what stop the plan from being fiction by June. Contractors who skip section 7 discover in the third quarter that they cannot answer the one question the plan exists to answer — which channel produced the booked jobs.

One row deserves naming because the industry keeps skipping it. Cornerstone's framework reports that database marketing to lapsed customers delivers $8-12 return for every dollar spent, compared to $3-4 for new customer acquisition — and that most contractors nonetheless spend 70-80% of budgets chasing new customers. That is an agency's figure from an agency's webinar, not a measured study, so treat it as a recommendation with a direction rather than a return you can bank. The direction is worth acting on anyway: section 6 exists so the cheapest audience you own gets a line in the plan instead of an afterthought in October.

The search side follows. The table below is our own [VENDOR — DataForSEO] Google Ads search-volume pull of 2026-08-17 (United States, location_code 2840, English), showing the twelve monthly values each keyword returned for the window 2025-08 through 2026-07. All five keywords returned the same window, so the columns line up.

Three readings matter for a calendar.

Repair demand peaks at the temperature extremes. "Air conditioning repair" peaks at 673,000 in 2026-07 and troughs at 165,000 in 2025-12 — a four-fold swing. "Furnace repair" runs the mirror image: it peaks at 450,000 in 2026-01 and troughs at 60,500 in 2026-06, a seven-fold swing. "Emergency ac repair" is the sharpest line on the sheet, at 60,500 in 2026-07 against 5,400 in 2026-02.

Replacement demand peaks earlier than repair demand. "Ac installation" does not peak in July with the heat. It peaks at 90,500 in 2025-09 and again in 2025-10, and troughs at 27,100 in 2026-01. "Furnace installation" peaks at 49,500 in 2025-10, three months ahead of the furnace-repair peak in 2026-01. Replacement intent lags the breakdown that triggers it, so it lands in the shoulder season — the weeks between the two temperature extremes when neither repair curve is near its maximum. A calendar that runs replacement offers only inside the peak repair months is arriving after the search happened.

March and April 2026 are the dual-shoulder months. They are the only two months in this window where both repair curves sit in the bottom third of their own ranges at the same time — air conditioning repair at 301,000 and 246,000 against a 673,000 peak, furnace repair at 165,000 and 110,000 against a 450,000 peak. April is the deeper of the two: adding the two returned values gives 356,000 in April against 466,000 in March, and April is the lowest combined figure of any month in the window (our own arithmetic across those two columns). Those are the months your plan has to have something else in it.

Cornerstone's recommendation, in the recap's own words, is to "front-load 60-70% of marketing spend into peak 4-6 months when customer acquisition costs are lowest, not evenly across 12 months." Two cautions attach to it. That is a marketing agency's recommendation, not a measured result. And "peak" is defined by your local degree-day curve, not a national table — the EIA notes the West North Central division generally has the most heating degree days and the West South Central the most cooling degree days each year, so a Minneapolis calendar and a San Antonio calendar are not the same document.

The skeleton those readings produce runs in five phases. Adjust the month labels to your own climate division; the sequence is what transfers.

Two mechanics belong in the calendar rows themselves. Google Ads offers seasonality adjustments to tell Smart Bidding about an expected conversion-rate change, but its own documentation bounds them tightly: "Use seasonality adjustments only if you expect major changes to conversion rates, because Smart Bidding already manages seasonal events," and they are "ideal for short events of 1–7 days" and "may not work as well" beyond fourteen. A three-day promotion qualifies; a four-month cooling season does not. And if Local Services Ads carries budget, Google states plainly that "after you've reached your monthly max, your ad won't appear for the remainder of the month" — a cap set in February switches you off in July unless a calendar row moves it.

The chain has four steps, and it runs in this order:

1. Revenue target ÷ average ticket = jobs required. Use your own ticket by department, not a blended figure.
2. Jobs required ÷ close rate = appointments required. Close rate is sold jobs divided by presented opportunities.
3. Appointments required ÷ book rate = leads required. Book rate is the share of leads that become a booked appointment — SearchLight Digital's published definition, and the one to standardize on.
4. Leads required × cost per lead = channel spend. Do this per channel, because cost per lead is not one number.

The worked example below runs that chain on published figures so you can see the mechanics. Every conversion input is a real published number with a source, but none of them is your number — the point of the table is the formula, not the result. Substitute your own row values before you commit a dollar.

Our own arithmetic over the sourced inputs above, rounded up at each step. The chain is reproducible; the inputs are not yours.

Read the last cell against the first. That $158,994 is 15.9% of the $1,000,000 target — well above the 6% of annual revenue that ACCA and Farmington measured as the average across more than 1,000 HVACR contractors, above the 10% of gross revenue Cornerstone recommends for growth, and inside the upper half of the "10% to 20% of sales" that ServiceTitan's blog attributes to industry experts. The chain lands high because every choice in it is the conservative one: the ticket is a target floor, the lead price is the HVAC-specific row rather than the cheaper blended figure, and no step gives credit for the leads that arrive from channels you do not pay per lead for. The gap is still the point. In the same ACCA study, "contractors who invest at least 12% see their net profits jump from 5% to 9%" — a correlation in a survey, not a proven return, but a correlation that runs in the same direction as the arithmetic.

Three cautions on those inputs, because this is where plans get built on sand. ServiceTitan's $7,700 and its 25-35% marketing-lead close rate are recommended targets published in a software vendor's playbook with no sample, no dataset and no date — the page itself says "This is just a sample of some important KPIs & benchmarks," and the install ticket is published as a floor rather than a central estimate. SearchLight's figures come from one PPC agency's own client accounts in a single month, January 2026, which is a heating month, so its cooling rows are off-season observations — and its campaign-type table blends HVAC and plumbing contractors, which is why the chain above uses the HVAC General service-category row instead of the $149 non-branded average. And the install close rates ACHR News reports from that ACCA study cover install jobs across all lead sources; no independent source publishes HVAC close rate broken out by marketing channel, which is exactly the number a plan most wants. If paid search carries the lead target, model the spend line before committing it with the paid-search cost model for HVAC campaigns.

Three process rules then decide whether section 5 survives contact with the year.

Pick one allocation axis and record which. Cornerstone's published split of its 10% is 25% to prospect and customer direct mail, 35% to digital, 20% to traditional media and brand building, and 20% to strategy and production support — though the body of the same page says 15%-20% for strategy and production, an inconsistency worth knowing before you quote it. ServiceTitan's playbook cuts on a different axis entirely: 25% branding, 50% call to action, 25% internal. These are not two estimates of one quantity, and a plan can only use one. Use the axis your reporting already produces, because a split you cannot report against is a split you abandon by April.

Give every row a constraint and a tracking mechanism. A channel line with dollars but no cap and no measurement is the row that quietly consumes the year. The grid below is what section 5 looks like filled in — the published constraint in each cell is what stops the row from being re-argued every quarter. Vendor figures in it are labelled: the cost-per-lead figures come from one PPC agency's own managed client accounts in January 2026, not from Google.

Fund the assets steadily and the campaigns seasonally. The first two rows are assets rather than campaigns — they compound when funded every month and decay when switched off in the shoulder season — while the paid rows are the ones the front-loading rule applies to. That distinction is what keeps a 60-70% peak-season concentration from switching off the profile work in March.

One row most 2026 plans still lack: AI search visibility. BrightLocal's Local Consumer Review Survey 2026 finds the share of consumers using AI to find local business recommendations has climbed from 6% in 2025 to 45% today, making AI the third most popular source of recommendations behind Google and Facebook — across all local categories, not HVAC specifically, from a panel of 1,002 US adults. That justifies a measured line, not a guaranteed outcome. Cornerstone's version is the right shape to copy: allocate 5-10% of the digital budget to testing new platforms and AI tools as they emerge. Anyone selling a guaranteed placement inside an AI answer is selling what Google warns about in writing: "If they guarantee you that their changes will give you first place in search results, find someone else." If you want the search and AI-visibility rows of a plan like this run by a team working one trade only, that is a one-trade agency built only for HVAC — audits, retainers and monthly measurement, priced in public.

Run each review in this order:

1. Reconcile actual revenue against section 1, by department. A miss in replacement and a miss in service have different fixes.
2. Recalculate section 2 from actuals. Your close rate and book rate move. When they move, every lead target downstream is wrong until you re-run the chain.
3. Report cost per booked job by channel, not cost per lead. SearchLight's own January 2026 dataset makes the case in figures that cover its HVAC and plumbing accounts together: non-branded search cost $149 per lead but $804 per paying customer, while branded search cost $34 per lead and $104 per paying customer. The agency publishes no cost-per-paying-customer figure for its HVAC-only rows, so use the ratio as a warning about lead price, not as an HVAC benchmark. A cheap lead that does not book is not cheap.
4. Check the tracking plan still works. Call tracking numbers get changed, form endpoints break, and a CRM lead-source field with 40% "unknown" makes the whole review guesswork.
5. Keep, cut, or double one line. Decide, write it down, and re-cut the next two quarters of the calendar to match.

Step 3 is where most plans fail quietly, and the reason is usually not the plan. Angela Kiel of BDR put the constraint plainly on BDR's marketing-budget page: "You could have the best website in the world, but if your internal staff doesn't know how to answer the phone, upsell accessories, or follow up with happy calls, you're driving your profits down the line. The phone call isn't the revenue, it's the relationships and actions after the call that generate true growth." A quarterly review that only examines media spend will keep buying leads into a broken conversion floor.

Which raises the question a plan can point at but cannot answer on its own: once the calendar, the targets and the budget are written, what determines whether those leads turn into booked jobs? The answer sits downstream of marketing, in call handling and in the offer. ACCA and Farmington's study of 1,000+ contractors measured the size of it: four or more proposal options raise close rates by 10% and shift premium equipment sales from 26% to 42%, and close rates rise 11% when financing is offered. As TJ O'Connor, president of Farmington Consulting Group, told attendees: "If you don't offer something to a customer, you have zero chance of making that sale." None of that is a marketing line item, and all of it changes what your marketing line items return.

Channel-by-channel scope, and what a combined program actually includes, sits on our how the channels fit together.

• Which HVAC marketing channels produce booked jobs — the channel-by-channel comparison this plan assumes you have already made.
• Budget tiers for 1-3, 4-10 and 10+ truck shops — the arithmetic behind section 5 at each company size.
• How HVAC demand behaves across peak and shoulder seasons — the seasonality depth behind section 4, including what to sell in the dual-shoulder months.