HVAC Seasonality by Climate: How to Plan for Peak Season and Survive the Slow Season

Before you change anything, pull the last 24 months out of your field service platform or accounting system. You want four columns by month: total revenue, inbound calls, jobs completed, and revenue split into service and repair, replacement and install, and maintenance agreements. Two years matters because one year can be a fluke - a hot summer or a mild winter will make a single year look far more or less seasonal than your market really is.

From that, calculate three numbers. The first is your peak-to-trough ratio: revenue in your best month divided by revenue in your worst month, averaged across both years. The second is trough length: how many consecutive months sit below about 75 percent of your average month. The third is how many weeks of fixed costs you have in the bank on the first day of your trough.

As rough orientation, a flat market like South Florida often runs a peak-to-trough ratio around 1.3 to 1.5. Four-season markets with a strong maintenance base land around 1.5 to 2. Single-peak desert and heating-dominant markets commonly run 2 to 3, sometimes higher for shops that are mostly replacement. But the published averages do not matter much. Your own ratio does, and it is shaped as much by your revenue mix as by your weather - a shop with 1,200 active maintenance agreements has a much flatter curve than a replacement-heavy shop down the street.

Split the curve by revenue type and you will usually see where the swing comes from. Service and repair calls follow the weather almost exactly. Replacement lags the weather by a few weeks and spikes after extreme events. Maintenance follows whatever schedule you set, which is the one part of the curve you control completely. That is the lever most of this guide is about.

Single-peak cooling markets - Arizona, Nevada, and most of Texas. Summer is brutal and long, emergency same-day AC repair is a major revenue category, and demand falls off hard from roughly November through February. The business problem is two-sided: surviving a four-to-five month summer at maximum capacity without losing calls or burning out techs, and then carrying full payroll through a winter where heat pump heating calls do not come close to filling the board. Texas adds freeze events, which is why dual-fuel and emergency heat work has grown in the northern half of the state.

Year-round cooling markets - Florida, and to a lesser degree the Gulf Coast. There is effectively no off-season, so the problem is not the trough, it is sustained capacity and fatigue. Hurricane season layers event-driven replacement spikes on top of a baseline that never really drops. Maintenance agreements are unusually strong here because of the retiree and snowbird population, and that recurring base is what lets Florida shops staff for twelve months instead of eight.

Humid Southeast heat pump markets - Georgia, the Carolinas, Tennessee. AC service drives most revenue, but mild winters with occasional hard cold snaps mean heat pumps carry real winter load. The trough is shallower and shorter than in the desert, and it gets punctured by cold-snap spikes where auxiliary heat and defrost problems flood the phones for three or four days. Spring humidity creates a shoulder-season opportunity in dehumidification and indoor air quality that desert shops do not have.

Four-season dual-peak markets - Ohio, Illinois, New Jersey, Michigan, Pennsylvania. Cold winters and hot humid summers give you two peaks, with two short shoulders in roughly April through May and late September through October. This is the most forgiving pattern financially, because each shoulder is exactly when maintenance visits belong. A shop with a strong agreement base can nearly erase its troughs. A shop without one feels two slow periods a year instead of one.

Heating-dominant markets - Minnesota, Massachusetts, the rest of New England, and much of Colorado. January is peak, and no-heat calls are safety calls, not comfort calls, which makes after-hours coverage non-negotiable. Summer is the softer season, though AC penetration keeps growing. The best trough fill here is policy-driven: heat pump conversion programs like Mass Save and cold-climate heat pump incentives create a steady install pipeline that can be scheduled into the months furnaces do not need you.

Mild and marine markets - coastal California, Oregon, and Washington. This pattern is smaller today but growing fast. Demand is low-amplitude most of the year, and then a heat dome sends call volume up several times over in homes that were never built with AC. The planning problem is surge capacity for a peak you cannot predict weeks in advance, plus electrification retrofit work that has become the growth engine in California. Check your own state page for climate, local software trends, and licensing notes - they are linked at the bottom of this guide.

In peak season you are not short of demand. You are short of hours, trucks, and people who can answer the phone. That changes what matters. A missed call in March is a delayed job. A missed call in July is a permanently lost one, because about 85 percent of callers who reach voicemail never call back and most call a competitor instead.

Run the numbers for an eight-tech shop. Say you take 600 inbound calls in a peak month and miss 30 percent of them, which is typical during peak - shops miss 25 to 40 percent when every CSR is already on a call. That is 180 missed calls. If a third of those were bookable work, that is 60 jobs. At a $450 average service ticket that is $27,000 of revenue in a single month, and some of those calls were replacement opportunities worth $9,000 each. An AI receptionist runs $49 to $199 a month at the Goodcall and Rosie tier, around $249 and up for Numa at that volume, and $300 to $800 for Avoca at call-center scale. This is the easiest purchase decision in the business.

The second leak is dispatch. Peak season is when drive time quietly eats a tech a day. Routing by proximity and skill rather than by who is free next is worth one to two billable hours per tech per day, and in peak that time converts directly to additional completed calls rather than to idle time. Set booking capacity limits by job type in your platform so the phones cannot book 14 hours of work into an 8-hour day, and triage deliberately: members first, no-cooling and no-heat before noise complaints, and replacement leads to a comfort advisor the same day.

Third, move work out of the peak. Tune-ups and non-urgent maintenance should not be happening in July in Phoenix or January in Minneapolis. Every maintenance visit scheduled in peak is a repair call you could not take. If you have agreement customers due in the peak window, move them into the shoulder and tell them why - priority scheduling when it breaks is the benefit they bought.

Last, the overtime question. Owners flinch at time-and-a-half, but do the math on one extra hour: roughly $55 to $70 fully loaded for a mid-level tech, against a repair call averaging $450 at 50 percent or better gross margin. Planned overtime during peak weeks is profitable. Unplanned overtime every day for four months is how you lose the tech in September. Set a cap, rotate on-call, and protect days off.

Maintenance agreements are the only part of your revenue curve you fully control, which makes them the most important seasonality tool you have. The mistake most shops make is scheduling them on a generic spring-and-fall template regardless of climate.

In four-season markets that template is roughly right: 60 to 70 percent of maintenance demand wants to land in about eight weeks in spring and eight weeks in fall, which is exactly when the board has room. The work is to pre-book those windows before they arrive - send the scheduling campaign four to six weeks early, let members self-book online, and fill the calendar before the first warm week instead of reacting to it.

In single-peak cooling markets the visit that matters is pre-cooling, and it belongs in February through April, in the trough, not in May when the phones are starting to ring. The second visit is lower value and can spread across October through January. In Florida, stagger visits by the month each member signed up rather than cramming everyone into spring, because there is no quiet window to cram them into. In heating-dominant markets, the pre-heating visit is the critical one and it belongs in August through October.

The other reason agreements flatten the curve is what they produce. A maintenance visit is where techs find the failing capacitor, the cracked heat exchanger, and the 16-year-old system on its last season. Scheduled in the shoulder, those findings turn into repair and replacement work that also lands in the shoulder. Scheduled in peak, they turn into a note nobody follows up on.

Watch renewal rate as closely as signups. Below 70 percent you are refilling a leaking bucket every year. Between 70 and 85 percent is a working program. Above 85 percent, where auto-pay shops tend to live, the base compounds on its own and your trough gets shallower every year without any other change.

There are two ways to deal with the slow season, and serious shops do both. Fill it with work that does not depend on the weather, and fund the gap that is left with cash you set aside on purpose.

Start filling with your own customer database, because it is the cheapest work you will ever book. Pull every customer with equipment 12 years or older and run a pre-season replacement campaign with financing options shown as monthly payments - replacement decisions that are not emergencies are exactly the kind that can be made in January. Add past customers who have not had service in 18 months, and customers who declined a repair recommendation. Database marketing typically costs a fraction of what a new lead costs, and it is work you can schedule.

Then add counter-seasonal lines that fit your climate. Heat pump and electrification retrofits are the biggest in markets with strong incentive programs, like California, Massachusetts, and New Jersey. Indoor air quality and dehumidification fit humid markets. Duct sealing and airflow correction can be scheduled anytime. Commercial preventive maintenance contracts are billed monthly and serviced on your schedule, which is why so many residential shops add a small commercial PM book specifically to carry winter.

Now fund the rest. The reserve formula is simple: take your monthly fixed cash out - the payroll you are keeping, trucks, rent, insurance, software, debt service - and subtract the cash you realistically bring in during a trough month after materials. Multiply the gap by the number of trough months. For an eight-tech shop with about $115,000 a month in fixed cash out and about $75,000 of trough-month cash in, that is a $40,000 monthly gap across three months, or roughly $120,000 before any surprise. Build it during peak by moving a fixed percentage of every peak-month deposit into a separate account you do not touch.

Open or expand your line of credit during or right after peak season, when your trailing revenue and bank balances look their best. Banks lend on recent performance, and applying in January with three soft months on your statements is the worst possible time to ask. The line is a backstop for a bad year, not the plan for a normal one.

The instinct in a slow January is to cut technician hours or let someone go. For good techs that is almost always the wrong trade. A replacement tech costs $75,000 to $110,000 once you count recruiting, burdened pay during ramp-up, and the callbacks while they learn your systems - and the only time you will be hiring is the spring, when every other shop in town is hiring too. The tech you let go in January is working for your competitor in June.

Instead, give trough hours a job. Use them for training - structured online programs like Interplay Learning run $35 to $150 per user and are far easier to schedule when the board is thin. Rebuild truck stock and standardize inventory so peak-season trips to the supply house drop. Update the price book before the season rather than in the middle of it. Push vacation into the trough and block it out of the peak, and say so in writing when you hire.

Consider guaranteeing a base of hours year-round in exchange for flexibility in peak. Techs care more about a predictable paycheck than about where the hours fall, and a guarantee is a real retention advantage over shops that cut hours every winter. It is only affordable if you have done the reserve math above, which is one more reason to do it.

For the office, the surge is on the phones, not the dispatch board. Hiring and training a seasonal CSR for a four-month peak rarely works - by the time they are good, the season is over. Covering overflow and after-hours with an AI receptionist scales up and down with call volume automatically, which is exactly the shape of a seasonal business.

The most common seasonal marketing mistake is spending the most when you can serve the least. Paid search costs climb during heat waves and cold snaps because every competitor is bidding, and the leads you buy land on a board that is already full. If you cannot answer the phone or dispatch the job within a day, those dollars are subsidizing your competitors.

A better calendar has three phases. Four to six weeks before your peak, run pre-season campaigns: tune-up offers, agreement signups, and replacement financing for old equipment, aimed at getting work booked before the rush. During peak, throttle paid top-of-funnel spend to what you can actually serve, and keep spending on the channels that bring higher-value work, like replacement leads. In the trough, shift budget to database marketing and to counter-seasonal services, where it fills real gaps.

Reviews follow volume, which means peak season is when you generate the most of them - if you ask. A shop completing 150 jobs a month in peak that asks manually might get 5 percent, or eight reviews. An automated request after every completed job tends to get 20 to 30 percent, or 30 to 45 a month. Those reviews are what lift your map ranking going into the next season. Review tools like NiceJob at $75 to $149 a month are the right tier for most residential shops.

Map the calendar to your pattern. Desert shops push pre-season in February through April and database replacement campaigns in November through January. Four-season shops run two pre-season pushes and two trough campaigns a year. Heating-dominant shops run their pre-season push in August through October and summer heat pump conversion campaigns. Florida shops focus on steady agreement growth and hurricane-season readiness rather than seasonal pushes.

You do not need new software to plan around seasonality. You need the platform you already have set up to answer seasonal questions. Four settings matter most.

First, reporting by month and by revenue type, going back at least two years, so you can see your curve without exporting to a spreadsheet. If your job types are not clean enough to split service, replacement, and maintenance, fix that first - it is the foundation for everything else. Second, booking capacity limits by job type and day, so peak-season scheduling cannot overbook the board. ServiceTitan and FieldEdge have mature capacity planning, which is part of why large Texas and Arizona shops lean toward them; mid-market platforms like Housecall Pro handle simpler versions.

Third, membership scheduling campaigns that fire automatically on your climate-specific schedule, not a default template. Most platforms can send due-date reminders; the setting that matters is when they send and whether customers can self-book. Fourth, call tracking tied to booked jobs, so you can see missed-call rate and booking rate by month and know exactly how much capacity you lost in last peak.

Then add the one or two tools that match your specific seasonal leak. For most shops that is an AI receptionist to cover peak-season overflow and after-hours no-heat or no-cooling calls. For shops whose curve is driven by replacement, it is database marketing and review automation. For shops trying to hold onto techs through a long trough, it is structured training that makes slow weeks productive.

Two months before the trough: finalize your reserve number, open or expand the line of credit, schedule trough-season training, and build your database replacement and agreement campaigns so they launch on day one of the slow period rather than three weeks into it.

During the trough: run database and counter-seasonal campaigns, schedule the critical maintenance visit for your climate, rebuild truck stock, update the price book, take vacations, and review last peak for missed calls, booking rate, and drive time. Decide on AI receptionist coverage, staffing, and overtime rules now, not in the first week of the rush.

Six weeks before peak: launch pre-season campaigns, finish maintenance visits for members, confirm capacity limits and triage rules in the platform, stock common peak failure parts, and set the overtime cap and on-call rotation.

During peak: protect capacity. Answer every call, route by proximity and skill, move non-urgent work out of the window, throttle paid spend to what you can serve, automate review requests, and move a fixed share of every deposit into the reserve account.

After peak: open the shoulder campaigns, follow up on every repair and replacement recommendation techs logged during the rush, and pull the month-by-month numbers again. Each year of measurement tightens your plan, and each year of a stronger agreement base makes the curve itself a little flatter.