HVAC Facebook Ads: Homeowner Targeting Is Gone

If you came here looking for the homeowner targeting checkbox, the honest answer is that it does not exist any more.

Meta removed direct homeowner targeting as part of complying with fair housing law. Then it went further: detailed targeting exclusions came out of ad sets on 31 March 2025, out of boosted posts on 10 June 2025, and on 15 January 2026 any ad set still relying on removed targeting options stopped delivering entirely.

If an HVAC campaign of yours went quiet in mid-January and nobody could work out why, check that date before you check anything else.

The useful news is that for this trade specifically, what replaces it is better than what was lost.

The short answer

Stop trying to describe your customer and start describing your service area and your offer. Target the postal codes where the housing stock is old enough to be failing, upload your customer list and build a lookalike from it, and let the creative do the qualifying. That combination outperforms interest targeting for HVAC because it is anchored to the one thing that actually predicts demand, which is equipment age.

Why losing it barely matters here

Homeowner targeting was always a poor proxy in this trade, and it is worth understanding why before replacing it with something equally weak.

Homeownership tells you someone can authorise the work. It tells you nothing about whether they need it. The variable that predicts an HVAC job is the age of the equipment, and the best available signal for that is the age of the house.

The median owner-occupied home in the US is 41 years old. A neighbourhood built in 1978 has a housing stock full of systems on their second or third replacement cycle. A subdivision completed in 2021 has almost none, and every one of those households would have shown up as a homeowner in the old targeting panel.

So the interest checkbox was giving you a demographic when you needed an inventory. Geography gives you the inventory.

Lever one: geography, weighted by housing age

This is the strongest replacement and it is fully available.

Map your service area by build era. Municipal open data, assessment records and simple local knowledge will tell you which neighbourhoods went up when. You already have a version of this in your head from driving the routes.

Weight budget toward the 25 to 45 year old stock. Old enough that systems are at end of life, recent enough that the homes are worth investing in rather than gutting.

Cross-reference with your own job history. Pull the last two years of replacements and plot the postal codes. That map is the highest-value marketing document in the business and almost nobody makes it. It shows where the work actually came from rather than where you assumed it would.

Keep radii tight enough to be honest. Drive time is a real cost and a real constraint, particularly through October when trips, drive time and miles per vehicle all peak. Advertising into an area you serve reluctantly produces leads you convert badly.

Lever two: your own customer list

This is the single best audience available to an HVAC contractor, and most have never uploaded one.

A list of your past customers already encodes homeownership, income band, housing type, neighbourhood and the fact that they were willing to pay you. No interest segment ever approximated that, and no competitor can copy it.

Two things to build from it:

Custom audiences, for reaching your existing base directly. Maintenance plan offers, seasonal reminders, and replacement conversations with customers whose systems you already know the age of. This is the cheapest advertising you will ever run.

Lookalike audiences, for finding more people like them. A lookalike built from a few hundred past customers outperforms any hand-built interest stack, because it is derived from behaviour rather than self-declared interests.

Two refinements worth the effort. Build a lookalike from your replacement customers only rather than all customers, so it optimises toward the segment carrying most of the revenue. And build an audience of everyone who visited a service page or started a quote and did not finish, since that is the closest thing Meta has to search intent, and it is usually the cheapest source of leads in the entire account.

Lever three: creative as the filter

The part of targeting that survived the changes intact is the part that was always doing most of the work.

An advertisement is a qualifier. Show a furnace in a basement, speak about your home rather than your place, mention a system that is fifteen years old, and renters, apartment dwellers and people with new equipment simply do not engage. You did not pay for their attention, and you did not need a checkbox to exclude them.

What qualifies well in HVAC:

• Name the symptom. A specific noise, a room that never gets warm, a bill that climbed. People self-identify against symptoms in a way they never do against a service description.
• Name the age. "If your system is over twelve years old" does more filtering than any demographic setting.
• Show the actual work. A real technician, a real rooftop unit, a real duct. Home services is one of the few categories where the job itself is watchable, and it reads as a real company rather than a stock library.
• Use local specifics. Naming a neighbourhood or a recognisable landmark filters and builds trust in the same sentence.
• Keep the offer price-free. A fixed-scope check gives someone a reason to act without starting a price comparison you did not want.

And the practical constraint that matters more than any of it: refresh creative every four to six weeks in-season. Creative age and season drive the $25 to $200 cost per lead spread in HVAC far more than audience settings do. Rising costs on stable targeting is fatigue, and no bid adjustment fixes fatigue.

The special ad category question

Worth understanding, because getting it wrong is expensive and getting it right takes a minute.

Special ad categories exist for housing, employment, credit and social issues, and declaring one strips detailed targeting, interest and behaviour segments and most demographic options, while constraining geography, with a minimum 15 mile radius applied in the US under the housing category.

Standard HVAC service advertising is not in any of these. Housing means property sales, rentals and housing finance. A furnace repair ad is a home services ad.

The place to be careful is equipment financing. If you run a monthly-payment offer, it is worth confirming with Meta how that ad is classified before building the campaign, because consumer credit offers can attract the credit category and the targeting restrictions that come with it. Financing-led replacement offers are among the better things to run on this platform, so it is a question worth resolving early rather than after a campaign underperforms for reasons that look like bad luck.

Audit checklist for an existing account

Ten minutes, and it catches most of what breaks:

1. Any ad set that stopped delivering around 15 January 2026. Rebuild it rather than repairing it.
2. Detailed targeting exclusions still referenced anywhere. They no longer function.
3. Customer list uploaded and refreshed within the last quarter.
4. A lookalike built from replacement customers specifically, not just all customers.
5. A retargeting audience of site visitors who did not convert. If this is missing, turn it on first.
6. Creative age. Anything older than six weeks in-season is a suspect.
7. Geography checked against your replacement job map, not against your aspirational service area.
8. Landing pages matching the ads. An ad promising a tune-up that lands on a generic homepage loses most of what it paid for.
9. Lead notifications reaching a phone, not an inbox nobody opens between jobs.
10. Median first-contact time on Meta leads. This one decides whether any of the above matters.

That last item is not a targeting issue and it is still the largest determinant of whether the account works. Meta leads are colder than search leads by construction, only 11% of HVAC businesses reply to a new lead within an hour, and contacting within five minutes makes you 21 times more likely to qualify than waiting thirty. A perfectly targeted lead answered the next morning is a wasted lead, which is the argument in why your HVAC leads are not converting.

Building the replacement map

Lever one depends on a document most contractors do not have, so it is worth spelling out how to make it. It takes an afternoon and it outlasts every campaign you will ever run.

Pull two years of completed jobs from your CRM or invoicing system. Job type, postal code, value, and date. If the data is messy, take replacements only; they are the ones that matter most and the ones least likely to be miscategorised.

Plot them by postal code. A spreadsheet pivot is sufficient. You are looking for concentration, not precision.

Add three columns: number of jobs, total revenue, and average value. The three do not rank the same way, and the difference is where the insight is. A postal code producing many low-value repairs is a different asset from one producing few high-value replacements, and they deserve different offers rather than different budgets.

Compare against build era. Municipal open data or assessment records will give you approximate construction dates by area. Overlay them.

What usually falls out of this exercise, in roughly this order of frequency:

• One or two postal codes produce a disproportionate share of replacement revenue, and nobody had noticed because the reporting was by month rather than by area.
• A neighbourhood the business considers core turns out to generate volume but little margin, typically repairs on systems too old to be worth maintaining and homes whose owners will not replace.
• An adjacent area with matching housing stock produces almost nothing, not because demand is absent but because no one has ever advertised into it. That is the growth area, and it is invisible without the map.

The third finding is the valuable one. It is the only reliable way to identify where to expand that does not involve guessing, and it converts a targeting problem into a geography decision you can actually verify afterwards.

Refresh it annually. Housing stock ages, neighbourhoods turn over, and the map from three years ago is describing a market that has moved.

What to do when volume is too low

A common failure after the targeting changes: contractors replaced interest stacks with tightly drawn geographic audiences, and delivery collapsed.

Meta's system needs enough people and enough conversion events to optimise. A single postal code in a small market may not clear that bar, and the symptom is an ad set that spends slowly, costs a great deal per result, and never stabilises.

The fixes, in order of preference:

1. Widen the geography before widening anything else. Add adjacent postal codes with similar housing age rather than loosening other constraints.
2. Consolidate ad sets. Three narrow ad sets competing for the same audience learn more slowly than one combined ad set, and they bid against each other.
3. Optimise for a more frequent event. If booked appointments are too rare to teach the system anything, optimise toward form completions or calls initiated, and hold the quality line through creative and follow-up instead.
4. Accept a broader audience with sharper creative. This is usually the right answer in small markets. Let the advertisement do the filtering, which it does at no cost, rather than the targeting panel, which does it by restricting delivery.

The instinct to narrow feels like precision and usually behaves like starvation. In a trade where the addressable audience is every household in a service area with equipment over a decade old, broad plus specific creative beats narrow plus generic creative almost every time.

What this changes strategically

The targeting removals push HVAC advertising toward what it should have been doing anyway.

Interest targeting encouraged contractors to guess at who their customer was. Geography, customer lists and qualifying creative force you to answer a better question: where is the failing equipment, and what would make its owner act before it fails?

That question has an operational answer rather than a marketing one, and it connects to the rest of how this platform earns its place. Meta cannot capture emergency demand; it creates demand ahead of failure, which is why it belongs to the pre-season windows rather than the peak, and why the offers that work on it are tune-ups, maintenance plans, indoor air quality and financing. The full case is in HVAC Meta ads that actually work, the timing in the best time of year to advertise HVAC, and where Meta sits against every other channel in the best lead sources for HVAC companies.