Somewhere right now, an analyst is building a model of a metro that looks a lot like yours. Housing stock age by zip code. Median income bands. Contractor counts, review footprints, owner ages. Your company is a row in that spreadsheet, and the analyst has never heard your hold music.
Market entry studies are a genre. Private equity platforms commission them before they pick a metro, and the genre has one consistent tell: every stat about your market gets read as someone else’s opportunity. We pulled the same data those models are built from, federal shipment numbers, Census housing data, and a year of M&A activity, and found something useful. Several things everyone “knows” about this market stopped being true in the last eighteen months, and most of the trend roundups published in January are still selling them.
Here is the state of the HVAC industry in 2026, stat by stat, read twice. Once as the market. Once as the target.
Your market is fragmented, and fragmentation is a strategy document now
The defining number in residential HVAC is that nobody owns it. The top 10 national players hold somewhere between 15 and 25 percent of the market. The rest belongs to roughly 120,000 heating and air contracting businesses, most of them independent and local. The systems market alone is $31.7 billion in 2025, headed for $54 billion by 2033 per Grand View Research, before you count the service and repair side. A market that size with no dominant brand is rare. Essential demand, recurring service revenue, a hundred thousand small owners. To you, that’s Tuesday. To an investment committee, that’s a thesis.
The thesis has been executing for years, and last year it accelerated. Private equity firms completed 32 add-on HVAC acquisitions through early June 2025, up from 17 in the same window the year before. An 88 percent jump. PE firms or their portfolio companies were behind more than half of all HVAC services M&A activity in that period.
But the deal count only measures one year of buying. Look at what’s already been bought, and what it’s worth. Apex Service Partners didn’t exist before 2019. Today it holds more than 100 local brands and nearly 8,000 tradespeople, and in late May, Apollo Global Management bought a minority stake in a deal valuing the company at $10 billion, with Apollo putting in $2 billion and Goldman Sachs running the process. Ten billion dollars, built in seven years, entirely out of companies that look like yours.
Wrench Group runs another hundred-plus brands at a reported $3 billion-plus in revenue. Behind them sit dozens of platforms in various stages of the same build: Sila, Redwood, Legacy, Any Hour Group, Turnpoint, NearU, a Blackstone-backed Champions Group as of February, and a long tail of smaller groups raising money to start. Add-on acquisitions at your size trade around 4 to 8 times EBITDA.
Here’s the part the market share stat hides. Almost none of those acquired companies changed their names. The trucks still say what the founder painted on them. The reviews, the phone numbers, the “family owned since 1987” on the website: all still there.
Between them, the platforms own well over a thousand local trade brands, and there’s a reason techs pass around wheel charts on Reddit mapping hundreds of familiar local logos to the platforms that hold them. Owners keep discovering that the competitor they’ve been benchmarking against for a decade has been reporting to Tampa for three years. Pull up your own competitive map and ask how sure you are about each name on it.
None of this means you should sell, and none of it means you’re about to get steamrolled. It means the spreadsheet exists, it’s been executing, and it’s wearing local plates. When a market study says “fragmentation creates clear entry opportunity,” you are the fragment. Every number that follows reads differently once you hold that thought.
The replacement wave is real, but the replacement math broke
By industry lifecycle estimates, roughly half the HVAC systems in American homes are 15 years old or older, and about a third went in before 2002. That’s the aging-stock story every market model leans on, and it’s directionally solid even if nobody has a perfect census of attics. Three million plus replacement opportunities a year, sitting in attics and side yards, getting older every season.
Here’s what the models built in 2024 didn’t price in: trade analysts put typical system prices at nearly double their 2019 levels. The A2L refrigerant transition added roughly 10 percent to install costs by most trade estimates, and new R-410A systems haven’t been manufactured since the start of 2025. EPA’s late-May rule change lets contractors install leftover pre-2025 inventory until it runs out, but once it’s gone, everything is new-refrigerant equipment priced accordingly. A homeowner who would have seen a $7,000 quote five years ago is now looking at five figures, and a whole-home heat pump install runs a median of about $25,000 by Rewiring America’s estimate.
So homeowners are doing what homeowners do when the quote doubles. They’re fixing instead of replacing. Housecall Pro’s platform data shows repair climbing from about 21 percent of HVAC revenue in late 2021 to over 33 percent by late 2025. You’ve felt this in your own board: the $900 repair winning against the $14,000 replacement, again.
The quote went public
The replacements that survive the repair filter are getting shopped harder. A homeowner who flinches at five figures doesn’t stop the project. He slows it down, collects more quotes, and does his research at the kitchen table before anyone rings his doorbell. That research now includes asking ChatGPT what a heat pump should cost, and the AI answers with real numbers whether your company participates or not. The pricing conversation is happening either way. The only question is whether your prices are in it.
This is why online pricing flipped from taboo to expectation. The old logic said publishing prices lets competitors undercut you. The new math says hiding prices removes you from the research phase entirely, which is where the 2026 homeowner makes most of the decision. And the tooling caught up to the moment. Instant-quote platforms like Contractor Commerce and HVAC Quote price dynamically off your own pricebook, using the inputs a comfort advisor would normally drive out to collect: square footage, system type, age of the existing unit. The homeowner answers four questions at the kitchen table and gets a real range built from your numbers, not a teaser.
Contractor Commerce reports one Phoenix client closing 80 percent of customers who saw pricing online before booking the visit. That’s a vendor’s number about its own product, so weight it accordingly, but the direction matches what transparent pricing does in every other category homeowners shop.
The same shift is reaching the in-home visit itself. Tools like Conduit Tech put a LiDAR scan and an ACCA Manual J load calculation in the tech’s hands in minutes: floor plan, sizing, duct assessment, proposal, all generated in the home while the customer watches. The quote stops being a number you defend and becomes math you show. Worth knowing: Conduit lists Sila Services, one of the larger roll-up platforms, as a partner. The platforms are already buying this presentation layer for their comfort advisors. The in-home visit becomes confirmation instead of discovery either way. The question is whose visit.
The “replacement-first revenue model” advice still circulating in industry content was written for 2021 economics. In 2026, the shop built only for replacement tickets has a demand problem. The shop that survives runs repair at a profit and converts it into replacements on the customer’s timeline. And when the homeowner goes looking for a price, its number is the one she finds. Replacement is still the engine. The path to it just runs through the repair call and the research phase more often than the straight sale.
The heat pump inflection didn’t happen in 2025. It happened in 2021.
You’ve probably seen the headlines announcing that heat pumps finally outsold gas furnaces. The announcement is about four years late. AHRI shipment data tracked by RMI shows heat pumps have outshipped gas furnaces every single year since 2021. In 2025 it was 3.6 million heat pumps against 3.2 million furnaces, a 12 percent gap. Anyone selling you the inflection as breaking news either hasn’t read the data or is recycling someone who hasn’t.
The 2026 picture is more interesting and less comfortable. Total shipments are down hard: January a/c and heat pump shipments fell 29 percent year over year, and the first quarter ran below 2025. But inside the decline, heat pumps are producing essentially all the unit growth that exists, while gas furnace shipments keep falling by double digits. Before you panic-read those headline numbers: a chunk of the decline is inventory correction after the refrigerant transition, not demand collapse. Distributor revenue held nearly flat through late 2025 while shipments fell, which tells you warehouses were full, not that phones stopped ringing.
Here’s why this belongs in an article about price. The homeowner staring at a five-figure quote is weighing repair against a straight AC-and-furnace swap against a heat pump against dual fuel, and the heat pump conversation is where price objections get answered with math instead of held off with charm. Operating-cost savings against the utility bill. Dual fuel as the hedge for the homeowner who won’t give up his furnace. Sizing that justifies the number on the proposal. In a market where every replacement gets shopped, the five-figure tickets go to the advisor who can run that math at the kitchen table.
And be honest about the PE angle, because it cuts against the independent here: the platforms know all of this. They run national training programs, and they’re buying the presentation tooling for their comfort advisors, which is what that Sila partnership from earlier is. Heat pump fluency is not your moat against them. It’s the entry fee. The good news is that it’s also the one weakness in the acquisition models you can close in a season: retraining six comfort advisors takes a quarter, while rolling a training program across a hundred acquired brands takes years.
Q4 2025 spent some of your 2026
If the first half of 2026 came in under your forecast, here’s a cause nobody puts on the marketing report: you sold a piece of this year last fall. Every contractor in America ran some version of “buy before the tax credit expires” through Q4, you included. The urgency was real and it worked. But the demand those campaigns converted had to come from somewhere, and a meaningful piece of it came from homeowners who would otherwise have replaced this year. A deadline promotion doesn’t create replacements. It moves them. That’s our read of the mechanics, not a published statistic, but anyone who’s run a deadline campaign knows where the volume comes from.
You don’t need us to tell you the 25C credit died on December 31; you watched it go. What’s worth saying is what hasn’t caught up. The “$3,500 to $8,000 per heat pump” figure that floated through sales scripts for three years was always sloppy, mashing the $2,000-capped federal credit together with state-administered, income-qualified rebates that exist in some states and not others. Plenty of scripts and website pages are still carrying versions of it. A comfort advisor quoting dead federal money to a homeowner who Googles it afterward doesn’t lose the sale. He loses the referral, the review, and the next three jobs on that street.
What’s left is a state-by-state patchwork: HEAR electrification rebates live in some states, paused or unlaunched in others, and already exhausted in pockets, like Colorado’s Front Range allocation, which closed in April. Incentive knowledge just shifted from a federal talking point everyone shared to a local capability almost nobody builds. The shop that can tell a homeowner exactly what’s available at their address, pulled from a resource like the DSIRE incentive database, has an edge that costs an afternoon to build. The shop still gesturing at “IRA incentives” is quoting a dead program.
What the people buying your competitors assume you’re bad at
Market entry studies justify their revenue projections by listing what incumbents do poorly. Read enough of them and the same four weaknesses appear every time: slow investment in heat pump training, clumsy rebate integration, a digital presence that exists but doesn’t compete, and single-trade limitation.
The uncomfortable part is that the list is usually right. That’s why the projections work. “Present on Google but not competitive” describes a lot of $10M shops: a Local Services Ads profile that’s live but ranking #6, organic rankings that stop at the company name, a Google Business Profile that hasn’t been touched since the photos were uploaded. The buyers aren’t betting they can out-tech you with something exotic. They’re betting you won’t fix the basics before they get there.
Here’s what their models can’t price: every weakness on that list is fixable by you faster than their weaknesses are fixable by them. A platform that acquires eight brands in a metro inherits eight CRMs, eight pricebooks, eight wrapped fleets that don’t match, and a general manager reporting to someone in another state. That thousand-brand wheel from earlier cuts both ways: every spoke on it is an integration project. You inherit nothing. You can fix your LSA profile this month. They’ll be untangling pricebooks into 2027.
And the consumer data sits on your side of the table. Survey after survey puts the same three things at the top of how homeowners pick a contractor, with roughly three-quarters rating them critical: a company that’s reliable and does what it says, a company they trust, pricing they understand. Not one of those is a scale advantage. All of them are local, earned, and slow to build, which is exactly why a buyer would rather purchase yours than build their own.
When a competitor sells, the math can move your way
We’re openly pro-independent around here, so take the bias as disclosed. But the honest case for the independent doesn’t require pretending the platforms are bad at this. It requires understanding exactly what they’re good at, because the opportunity lives in what’s left over.
Give them their due. The platforms professionalize. They bring real call center metrics, pricing discipline, recruiting pipelines, training budgets, and capital that doesn’t flinch at a slow quarter. The shop they buy usually gets better at the things its founder was too busy to systematize. Anyone telling you PE-backed competitors are pushovers is setting you up to lose to one.
Now look at what a $10 billion valuation has to be fed with. Apollo didn’t write a $2 billion check for the privilege of holding prices flat. Platform economics run on margin expansion, and the lever that moves fastest is price. Debt service doesn’t take the summer off. So the acquired shop’s pricebook climbs, the replacement quotas climb with it, and the techs who got into the trade to fix things start getting measured like a sales floor. Some of them leave. The founder, the person whose name is on the trucks and whose handshake built the review count, is usually gone within a few years of close.
Which is why a competitor selling can be the best thing that happens to your year, if you’re set up to collect. Three windows open. Their price increases build an umbrella over your market: you can hold price and take share, or follow them up and take margin.
Their unsettled techs become your recruiting class, and the best ones move in the first year, not the fifth. And their identity gets blurry at exactly the moment yours gets sharper, because “locally owned, the owner still picks up” stops being wallpaper and becomes the single clearest difference between you and the shop across town with the familiar name and the new playbook. You don’t have to say a word about who owns them. The homeowner notices on her own when the tune-up turns into a pitch.
None of this is automatic. An independent with 47-minute speed to lead and an unmanaged LSA profile collects none of it. The windows open either way. Whether anything comes through them is an operations question, which brings us to the playbook.
What the data says to do with the back half of 2026
So how does an independent beat doubled prices and platform competition in the same year? With the same playbook, because they’re the same fight: both come down to running a tighter operation than the math your competitors are stuck with. It has three moves, and none of them are glamorous.
Run repair like it’s a profit center, because it is one now. At a third of revenue and climbing, repair stopped being the loss leader you tolerate between installs. Dispatch efficiency, repair pricing discipline, and first-visit completion rates are margin drivers this year.
And with the federal urgency lever gone, the repair call is also the only honest replacement pipeline you have left. The tech standing in front of a 19-year-old condenser is your best marketing channel, if he’s equipped to start that conversation instead of just clearing the ticket.
Build your incentive answer sheet before the homeowner asks. One afternoon: list every state rebate, utility program, and financing option live in your service area, by city, with current numbers. Train it. Most of your competitors won’t bother, because it’s boring. Boring and accurate beats enthusiastic and wrong in every market we’ve ever looked at.
Choose density over distance. In a softer demand year with acquisition costs up, route density and repeat-customer economics beat planting a flag two counties over. The math on geographic expansion still works in specific situations, but it works because the operator went in with a plan, not because the map looked empty.
One more, whether or not you ever take a meeting: know your number. Add-ons trading at 4 to 8 times EBITDA means every point of margin you add this year compounds twice, once in the P&L and once in what the business is worth. Running the company like someone might buy it someday is indistinguishable from running it well.
The spreadsheet can’t see you
Everything in that analyst’s model is true. The housing stock really is aging. The incomes really do support premium tickets. The shipment curves point where they point. And none of it knows which dispatcher answers your phones at 6am, whose yard signs sit in which neighborhoods, or who gets greeted by name at the supply house counter.
The data describes the market. It cannot see the operator. That gap is the real state of the HVAC industry in 2026: the buyer is betting the data is enough, and you’re betting it isn’t. One of you is right, and it gets decided one service call at a time.
