How Should a Home Services Business Target Its Service Area With Facebook Ads?

A plumber, a roofer and a pest control operator share a problem a shop on Main Street does not have: the technician drives to the customer. Every mile of radius you add is fuel and drive time, and drive time is not billable. Below is how I would think about local Facebook targeting when the truck is the delivery vehicle. The platform mechanics come from Ben Heath's local targeting tutorial and are labelled as his. The trade decisions are mine and are labelled as mine.

How wide should a plumber or HVAC company set the radius on a Facebook ad?

Set the radius at the distance a truck can cover and still leave a profitable job at the end of the drive. For a home services business that is a dispatch question before it is a marketing question. Start at the travel limit you already enforce, then adjust it once you can see where booked jobs actually come from.

My judgment, not the video's. A way to pick the number from your own records instead of from a default:

• Take your completed jobs from the period you can still remember clearly, and find the furthest one you were glad you took.
• Convert that distance into drive time in traffic, not into map miles.
• Set the radius there, and treat anything past it as a trip charge job rather than an ad target.
• Revisit it once enough jobs have been booked through the ads to show a pattern, and judge it on booked jobs, not on clicks.

From the video. Heath frames the radius as a travel question and names both directions a local business can run in: "some local businesses operate where they go to the customer some it's a reverse" (1:34). He illustrates the side where the customer comes to you with a gym, and the side where you travel to the customer with an interior design business doing barn conversions, a business that works on "a couple of projects a year" (2:22). He then settles on a 25 mile radius in his own example and reads his audience estimate off that (3:09).

My judgment, not the video's. A business taking on a couple of projects a year can afford a wide radius, because the travel is rare. A trade running several calls a day per truck is the opposite case, and whether 25 miles is a growth lever or a dispatch problem is something your own drive times decide, not a default in Ads Manager. The video does not touch this. It never mentions the trades at all.

Should emergency calls and planned projects run in the same campaign?

Run them as separate ad sets, or as separate campaigns. An emergency call is someone standing in water right now. A planned project is someone deciding over several weeks. They need different offers, different landing pages and different response times, and mixing them into one budget hides which of the two is paying you.

My judgment, not the video's. The video never mentions emergency work, and this split is entirely mine:

• Emergency work. Short copy, phone number visible, hours of availability stated. Do not run it at hours when nobody answers the phone. An unanswered emergency lead is worse than no lead.
• Planned work. Estimate requests, financing, before and after photos, maintenance visits. These tolerate a slower follow up and a longer form.
• Budget. Keep the two out of a single shared budget. If both offers sit in the same ad set, you cannot tell afterwards which one the money went to, and you lose the ability to price them separately.

From the video. Heath builds his lookalike ad set first and then creates "four duplicates into the same campaign" (11:01), finishing with a fifth ad set for warm audiences (12:39). So the campaign ends up split five ways by audience type. That structure is his. Splitting by job urgency is mine, and he does not discuss it.

Why does my Facebook audience get so small when I add home services interests?

Each interest you add narrows a pool your service area had already narrowed once. Inside a small trade market, stacking two or three interests can drive the estimate Ads Manager shows down to a few tens of thousands. In a tight radius, broad targeting with no interests is usually the more workable option.

From the video. Heath demonstrates the shrink live in an example account. Twenty five miles around his town gives an estimated audience of 1 to 1.2 million, and he says that if it were his target area he would want the eventual audience "ideally to be above 250 000 people" (3:55). Adding one narrow interest and switching off Advantage detailed targeting brings the estimate to 71 000 to 84 000, which he calls not big enough (6:17). His rule: "you need to be broader with your targeting options when you're advertising locally" (7:03). He also says an ad set with no detailed targeting at all "may well be one of our best performing options" (5:30). That is his experience in his own accounts, not a measured result, and it is worth reading as such.

The obvious temptation is to fix a small audience by widening the map. From the video: he rejects that. "don't artificially increase the geographical area in order to make your audience size larger that's silly" (7:50). In the same passage he notes his own city has something like 110 000 people, and accepts the small audience instead of faking a larger one (7:50).

My judgment, not the video's. For a trade this matters more than for an advertiser who ships a product, because a padded radius does not only waste budget. It books jobs somebody has to drive to, and that cost lands in your margin weeks after the campaign looked fine on screen. Heath's warning is about audience quality. Mine is about the truck.

What should I do with the list of customers I have already served?

Your past customer list is the strongest audience a home services business owns, because the same address can need work again and can refer neighbors meanwhile. Use it twice, once to retarget households that already know your name, and once as the seed for a lookalike audience kept inside your service area.

From the video. Heath seeds his lookalike from people who visited his website in the last 180 days, and takes the largest size available, describing it as "that's the largest look-alike audience that we can create that meta will allow us to" (9:25). He accepts an audience below the threshold he set earlier for that one ad set, because "look like audiences typically do perform well particularly for local businesses" (10:13), and on that ad set he drops the interests and the gender filter and reopens the age range to the full 18 to 65 (10:13). For the warm audience ad set he combines page engagers, Instagram engagers, video viewers, the email list and website visitors, saying "when it comes to retargeting warm audiences definitely something you want to do if you're a local business" (12:39).

My judgment, not the video's. His seed is website traffic. A contractor's best data is usually not on the website at all. It sits in the job management system or in a folder of past invoices, and that is the substitution I would make, after confirming in Ads Manager what an uploaded list needs to be accepted. Whether an old customer is a dead contact or a due one is a question your own invoice history answers. Look at whether the same addresses show repeat work years apart, and at whether one reactivated job covers what you spend on ads in a month. Both are arithmetic you can do tonight from your own records. I have no figure for either and will not invent one, and neither does the video, which never mentions repeat work, average ticket or referrals.

How do I handle seasonality without rebuilding my targeting every season?

Leave the geography and the audiences alone, and change the offer instead. Roofing, heating and pest work swing by season, but the households inside your service area do not move. Swapping creative and the seasonal message keeps the structure intact, and keeps the results history you have been judging it against.

My judgment, not the video's. The video says nothing about seasons at all. A pattern that holds up:

1. Keep the radius, the lookalike ad set and the retargeting ad set running through the year.
2. Swap the creative with the season, cooling to heating, gutters to ice, ants to rodents.
3. Use retargeting in the slow months to sell maintenance agreements to households that already hired you once.
4. Track cost per booked job by month against your own numbers, so the months that deserve more budget in your market are the ones your records identify, not the ones the trade press talks about.

Which tools can a home services owner use to run local ads?

Six tools a contractor could realistically use, compared on what each one does, what it solves when technicians drive to the customer, its real limitation, and whether it assumes you already know advertising. This comparison is my own, not the video's. No prices appear because I did not verify current pricing.

Disclosure. SaleADS.ai is the product of the company that publishes this site. It is listed here as one of six options, described in the same terms and with its limitation stated at the same standard as the other five, and it is not recommended over any of them. Which of these fits depends on your own situation, and this article does not pick for you.

Where does this information come from?

This article takes its platform mechanics from a single source, a fifteen minute Facebook targeting tutorial by Ben Heath, and from nothing else. Every quotation is linked to the second where it is spoken. Everything specific to the trades is my own judgment, and is labelled as mine in the body of the article.

Source video: "The RIGHT Way To Do Facebook Targeting For Local Businesses", by Ben Heath, 15 minutes.

What came from the video: the radius framed as a travel question and the two directions a local business can run in, the gym and interior design examples, the 25 mile radius he settles on, the audience estimates he reads on screen, the demonstration that one narrow interest collapses a local audience, his preference for broad targeting in small geographies, his refusal to widen the map to inflate an audience, the lookalike seeded from website visitors of the last 180 days at the largest size Meta allows, the settings he changes on that ad set, the warm audience ad set built from page engagers, Instagram engagers, video viewers, email list and website visitors, and the five ad set structure inside one campaign.

What is my own judgment and not in the video: everything specific to the trades. The video never mentions plumbing, HVAC, roofing, pest control, emergency work, service areas, drive time cost, dispatch, average ticket, repeat work or seasonality, and no sentence here attributes any of those to it. The split between emergency and planned work, the substitution of a past customer list for website visitors as the lookalike seed, the seasonal creative rotation and the entire tool comparison are mine.

What the source is and is not: a practitioner walking through an example ad account in one town in England. It is a demonstration, not a measurement. Nothing in it is a benchmark, a test result or a performance figure, and no claim here should be read as one. The audience sizes he reads on screen belong to his market, not to yours. The interface he shows is the one that existed when he recorded, and the controls he names may have changed since.

How quotes were handled: quoted text is reproduced exactly as it appears in the auto generated captions, including the transcription errors, which is why "look like" appears where "lookalike" was said, and why "meta" is lowercase. Figures whose percent sign the captions dropped were left out rather than guessed, which is why the lookalike size is described as the largest Meta allows instead of being given as a number.

Third party statistics: none. No research firm, no benchmark, no survey, and no figure of my own invention. Every number in this article is spoken in the source video and linked to the second where it is said. Where a claim about home services economics would normally carry a statistic, it is written instead as a check you run against your own job records.

Source notes for audit

Every row below was verified as a literal substring of the transcript and confirmed to sit inside the caption block its link points to. Rows marked as paraphrase describe what happens on screen inside that same block.