Tales From the Pink Swamp

The short answer: Facebook ads for local businesses work when they do one job: put a specific offer in front of people in your service area and turn their interest into a booked conversation. They fail when they are bought like a billboard — boosted posts optimised for likes, audiences drawn around a whole city, and leads that sit unanswered in an inbox for a day. The difference between those two outcomes is not budget. It is structure: a leads objective in Ads Manager, tight local targeting, one clear next step, and a follow-up that happens in minutes, not tomorrow. Get those four right and Facebook becomes a pipeline builder. Get them wrong and it becomes the most expensive way to collect compliments.

Start with the honest framing, because it saves more money than any targeting trick. Facebook ads are not Google ads, and treating them the same way is the first mistake. Google captures demand that already exists: somebody searches for an emergency plumber, a roof repair, or a dentist open on Saturday, and the ad meets them at the moment of need. Facebook creates demand: somebody is scrolling through photos of their cousin’s holiday, sees an ad for a kitchen refit or a lawn care plan, and realises they have been meaning to sort that out for months. One channel harvests intent. The other plants it. A local business that only runs Facebook ads for emergency work will be disappointed, because nobody scrolls Facebook hoping their boiler breaks. A local business that uses Facebook to fill its diary with planned work — refits, replacements, seasonal services, memberships, consultations — is using the channel for what it is actually good at.

That distinction decides almost everything else in this post: which businesses should run Facebook ads at all, what a lead should cost, what the ad should say, and how long you should wait before judging it. Keep it in mind as we go.

Facebook ads for local businesses: who they suit, and who they don’t

Facebook ads suit local businesses where three things are true. First, the job is worth enough to pay for a lead. Second, the decision is not purely an emergency — the customer can be persuaded a week or a month before they buy. Third, the business can actually answer and follow up, because a Facebook lead that waits is a Facebook lead that cools. Gyms, salons, clinics, dentists, home improvement firms, landscapers, cleaning companies, tutors, estate agents, and event venues all fit that shape. So do trades selling planned work: a plumber advertising boiler replacements and bathroom refits will do far better on Facebook than a plumber advertising emergency callouts, because the emergency customer is on Google, not scrolling.

Facebook ads struggle in the opposite cases. If your average job is worth very little and rarely repeats, the maths may never work — you cannot profitably pay $30 or $40 for a lead that produces a $60 one-off job. If your market is tiny — a village of a few thousand people — you will exhaust your audience fast, and frequency will climb until the same people see the same ad ten times. And if nobody in the business will answer leads quickly, Facebook will punish you twice: the lead goes cold, and Meta’s system learns that your ads produce low-quality outcomes and charges you more to show them. In each of those cases the right move is not a better ad. It is a different channel, or a different offer, until the maths changes.

Here is the two-minute test that settles it for any business. Work out what a new customer is worth to you in gross profit over their first year — not revenue, profit, and not lifetime value stretched over a decade, just the first year. Multiply that by the share of leads that actually become customers. The result is the most you can pay for a lead and still break even. Imagine a home improvement firm where a new customer is worth $2,000 in first-year gross profit and one in five leads buys. The ceiling is $400 per lead — generous headroom over any benchmark in this post. Now imagine a salon where a new customer is worth $400 in the first year and one in three leads books. The ceiling is $133 per lead — still workable, but the margin for sloppy targeting is much thinner. Run your own numbers before you spend anything. If your ceiling sits below your industry’s average cost per lead, fix the offer, the price, or the follow-up first. No amount of clever targeting rescues a business model where the lead costs more than the customer is worth.

What Facebook ads for local businesses cost in 2026

Costs first, because every other decision is priced against them. The most useful 2026 benchmark set for planning is WordStream by LocaliQ’s Facebook Ads Benchmarks 2026, based on 452 United States lead campaigns run between April 2025 and June 2026. Its headline figures are medians, which matters: medians from different campaign groups do not divide neatly into each other, so treat them as planning anchors, not promises.

Across all industries in that 2026 data, lead campaigns averaged a $1.80 cost per click, a 2.70% click-through rate, an 8.54% conversion rate from click to lead, and a $27.39 cost per lead. Traffic campaigns, optimised for visits rather than leads, were cheaper per click at $0.60 — which is exactly why beginners get fooled by them. A cheap click that never becomes a lead is not a bargain. It is a cost with no return. Judge Facebook spend by cost per lead and cost per customer, never by cost per click alone.

Industry swings are wide, and your industry’s number is the one that matters for your budget. In the same 2026 benchmarks, real estate leads were among the cheapest at $13.74, health and fitness sat near the average at $27.11, home improvement ran higher at $42.95, beauty and personal care higher still at $50.91, and dentists at the expensive end at $61.56. Those are national medians for the United States, not quotes for your town, and small samples sit behind some industries — but the ordering is instructive. A dentist and a roofer are not playing the same auction. Budget against your industry’s figure, then replace it with your own after the first month of real data.

Two more 2026 facts belong in any honest budget conversation. First, Meta’s own earnings calls showed the average price per ad rising 12% year on year in both the first and second quarters of 2026, while impressions grew 19% and then 14% — the auction is getting more crowded, and costs are drifting up even as more inventory appears. Second, a 2026 cross-platform benchmark put Meta’s cost per thousand impressions in the $14.50 to $22.00 range for most advertisers, with lead costs between $18 and $45. If your costs land well above those ranges, the cause is usually creative or targeting, not the platform. If they land well below, enjoy it — and check that the leads are actually qualified before celebrating, because cheap unqualified leads are the most expensive kind.

Translate the benchmarks into a starting budget the honest way: backwards from customers. Decide how many new customers you want a month, divide by the share of leads that become customers, and multiply by your industry’s cost per lead. Want four new customers a month, closing one in five leads, in home improvement at roughly $43 a lead? That is 20 leads at $43, or about $860 a month. Want two customers a month at the same close rate in real estate at roughly $14 a lead? That is 10 leads at $14, or about $140 a month — though in practice a budget that small may not give Meta enough results to learn from, so a floor of $300 to $500 a month for a first test is more realistic for most local businesses. The formula is the point, not the examples. Round-number budgets — $5 a day because a blog said so, $1,000 a month because it sounds serious — know nothing about your business. The backwards calculation does.

Boosted posts are not Facebook ads

This is the single most expensive misunderstanding in local Facebook advertising, so it deserves its own section. A boosted post is a simplified ad: you pick a post that already exists, pick a budget, and Meta shows it to more people. It is easy, and it is optimised by default for engagement — likes, comments, shares. A campaign built in Ads Manager is a different machine: you choose a leads or sales objective, you control the audience, the placements, the creative, and the destination, and Meta optimises for the outcome you actually want — a lead, a booking, a call.

The practical difference shows up in what you get for the money. Boost a post about your new van and you will get likes from people who will never buy, including friends, family, and strangers three counties away. Run a leads campaign with a tight service-area radius and a quote offer and you will get names and phone numbers from people who live where you work. Both cost money. Only one fills a diary. If you have boosted posts in the past and concluded that Facebook ads do not work, you have not actually tested Facebook ads. You have tested the boost button, and the boost button is designed to sell reach, not customers.

There is one legitimate use for a boost: amplifying a post that is already performing organically to an audience that already knows you — existing followers, past customers, website visitors. That is retargeting with extra steps, and it can work. What a boost cannot do is find cold strangers and turn them into leads efficiently. For that, use Ads Manager, every time.

The campaign structure that actually books work

A local Facebook campaign has three levels, and getting each level’s job right is what separates a working campaign from an expensive hobby. The campaign sets the objective. The ad set sets the audience, budget, and schedule. The ad sets the creative — the image or video, the headline, the body text, and the call to action. Most failed campaigns fail at the campaign level, because the objective was wrong, and no amount of good creative rescues an objective that optimises for the wrong outcome.

Objective: leads, or sales if the website can prove it. Choose the Leads objective if you want enquiries, calls, or bookings through a form. Choose Sales only if people buy or book directly on your website and the Meta Pixel can see that purchase or booking happen — otherwise Meta cannot learn what a good result looks like and will optimise blind. Never choose Traffic or Engagement for lead generation. Traffic finds clickers. Engagement finds likers. Neither finds customers, and both will happily spend your whole budget proving it.

One campaign, one ad set, to start. Beginners split a small budget across five audiences and wonder why nothing learns. Meta’s system needs results to optimise — the commonly cited figure is around 50 results a week per ad set to exit the learning phase, and a $300-a-month budget will not produce that across five ad sets. Start with one ad set, one audience, and two or three creative variations inside it. Let the system learn on a meaningful pile of data, then split winners out. Scale what works. Kill what does not. Complexity is a luxury for budgets that can afford to feed it.

Audience: geography first, everything else second. Draw a radius around your real service area — the towns and postcodes you actually serve, not the whole county because it felt ambitious. One 2026 local advertising guide reported that a roughly 15-mile radius delivered about 40% cheaper leads than city-wide targeting for local businesses; treat that as an agency-reported finding rather than a law of nature, but the direction is right and matches how the auction works. Every mile beyond your service area is money spent on people you cannot serve or will not travel to. Layer on age and homeowner or life-stage signals only where they genuinely describe your buyer — a roof replacement ad shown to renters is waste, however cheap the click. And leave room for Meta’s automation: broad targeting powered by good creative now outperforms rigid interest micro-targeting in most 2026 accounts, because the system has better data about who converts than your guesses do. Give it a sensible geography, a real offer, and creative that shows the work, and let it do the finding.

Creative: show the work, not the logo. The ads that win for local businesses are almost embarrassingly simple: a real before-and-after photo, a short video of the crew doing the job, a photo of a finished project in a home like the viewer’s. Polished stock imagery loses to real work every time, because the viewer is not buying a brand. They are buying proof that you can do their job, in a home like theirs, to a standard they would accept. Keep heavy text off the image — Meta pulls back reach when text covers much of the image — and put the offer in the headline and body text instead: what you will do, for whom, in which area, and what happens next. “Free roof inspection in [town] — book a 20-minute visit, no obligation” beats “Quality roofing solutions since 1998” by a distance, because the first one contains an offer and a next step and the second contains only adjectives.

Destination: a form or a landing page built for one action. A Facebook lead has two good destinations and one bad one. The good ones are Meta’s native lead forms, which collect a name and phone number without the person leaving Facebook, and a dedicated landing page with one offer and one form. The bad one is your homepage, where the visitor must work out what to do next among six menus, three services, and a blog. Our earlier post on lead magnets for local businesses covers the offer side of this in detail: the form must ask for as little as possible — name, phone number, and one qualifying question — because every extra field costs you leads. A form that asks for a life story gets a life story’s worth of nobody. Native lead forms tend to be cheaper per lead than website forms, because there is no page to load and no friction to lose people — one 2026 local guide reported $34.10 per lead for native lead forms against $45.80 for video-view campaigns, again agency-reported rather than universal — but cheaper is not automatically better. A lead form that produces unqualified names is worse than a landing page that produces fewer, better leads. Test both with the same ad, and judge by cost per booked job, not cost per lead.

The part most businesses skip: what happens in the first five minutes

Here is the uncomfortable truth about Facebook leads: they are colder than Google leads and they cool faster. A Google searcher has a problem right now. A Facebook lead has an interest right now, and an interest has a half-life measured in minutes. Our post on speed to lead covered the response-time data in full, and the principle applies with extra force here: the business that calls within five minutes books the job, and the business that emails the next morning books the memory of a lead. Build the follow-up before you build the campaign. That means a notification that reaches a phone, not just an inbox; a call or text within minutes during working hours; and a missed-call text-back for anything that arrives after hours, so a Saturday-night lead gets a Saturday-night reply instead of a Monday-morning apology.

The follow-up sequence for a Facebook lead should be short and specific. First contact: a call within five minutes, or a text within five minutes if the call fails, naming the offer they responded to — “Hi, it’s [name] from [business], you asked about a free roof inspection” — not a generic “thanks for your enquiry.” Second contact: a text the same day if the first fails, with a booking link. Third: a call and text the next day. Fourth: a final message a few days later offering a different next step, like a quote by photo or a shorter consultation. After that, the lead goes into a nurture sequence, not the bin — our post on nurture email sequences covers the longer game. The businesses that complain Facebook leads are junk usually have no sequence at all. They called once, got voicemail, and declared the channel dead. The channel was fine. The follow-up was the product, and it was missing.

There is a second follow-up job that costs nothing and doubles the value of every ad: retargeting. Most people who click a Facebook ad do not convert on the first visit. Retargeting ads for local businesses put a second, different ad in front of those visitors — a testimonial, a finished project, a reminder of the offer — and that second touch is often the one that books. Set it up from day one, with a small share of the budget, and exclude people who already converted so you stop paying to advertise to your own customers. Retargeting is the cheapest lead you will ever buy from Facebook, because the audience already raised a hand. Businesses that run prospecting ads without retargeting are paying full price for every lead and leaving the discount on the table.

Facebook ads for local businesses: a 90-day test plan that won’t waste money

Do not judge a Facebook campaign in a week, and do not fund it blindly for a year. Ninety days, judged in three stages, is the honest test.

Days 1–30: learn, don’t optimise. One campaign, one ad set, two or three creatives, one offer, one destination. Expect the first leads within the first two weeks at a modest budget, and expect the cost per lead to be noisy. Do not pause a creative until it has spent at least twice your target cost per lead — anything less is noise dressed as data. Do fix broken things immediately: a form that does not deliver, a phone number that rings nobody, a landing page that loads slowly on mobile. Those are not tests. Those are leaks.

Days 31–60: cut and double down. By now you have a real cost per lead and, if your follow-up works, a real cost per booked job. Kill the creative that produces clicks without leads. Kill the audience segment that produces leads without bookings. Put the savings into the winner. Add retargeting if you have not already, and add one new creative variation that changes the hook — a different first line, a different photo, a different offer framing — rather than a reskin of the same ad. Creative fatigue is real: when the same audience sees the same ad too often, costs climb and response falls, so fresh creative is a running cost of the channel, not a one-off.

Days 61–90: judge on customers, not leads. The only verdict that matters at day 90 is cost per customer: total spend, including any management fee, divided by customers won from the campaign. If that number sits comfortably below what a customer is worth in first-year gross profit, scale the budget gradually — 20% to 30% at a time, not a doubling overnight, because sudden budget jumps reset the learning and spike costs. If cost per customer sits above what a customer is worth, do not just spend more and hope. Change one thing: the offer, the audience geography, the destination, or the follow-up speed. If two of those changes fail, the honest answer may be that Facebook is not your channel right now, and the money belongs in search, reviews, or referrals. A failed test that you actually ran is valuable information. A failed test that you never measured is just a bill.

Tracking makes the verdict possible, and it is worth setting up properly from the start. Install the Meta Pixel on your website and the Conversions API alongside it, so Meta learns from real bookings and not just browser signals that keep getting weaker. If you also run call-heavy advertising, our post on call tracking for local businesses shows how to attribute phone bookings to the channel that produced them — essential when a Facebook ad produces a call rather than a form fill. Without tracking, you will judge the campaign on likes and leads, and both lie. With it, you judge on customers, and customers do not lie.

The mistakes that quietly eat a local Facebook budget

Targeting the whole city. Your service area is not your ambition. It is the postcodes you will actually drive to, profitably, this month. Every impression outside it is waste, and waste compounds.

Sending traffic to the homepage. A homepage answers ten questions. An ad should answer one and ask for one action. Build a landing page per offer, or use a native lead form, and never make a warm lead navigate.

Asking for too much on the form. Name, phone, one qualifier. That is the form. A ten-field form is a filter that filters out your customers.

Judging on likes and clicks. A campaign with a $0.30 click and zero bookings is not cheap. It is a loss with good branding. Cost per booked job is the only scoreboard.

Pausing too early. A week of data at a small budget is not a verdict. It is a weather report. Give the test its 90 days, fix the leaks as you find them, and judge at the end.

Running ads with no follow-up system. This is the big one, and it is not an ads problem. It is an operations problem wearing an ads costume. Fix the five-minute response before you raise the budget, or you will pay to generate leads you then donate to competitors who answer faster.

One creative, forever. Creative is the targeting now. One ad, shown until the audience is sick of it, is how a working campaign slowly dies. Budget for a new creative every month as a cost of the channel.

Facebook ads vs Google ads for a local business: how to split the budget

Most local businesses end up running both channels, just not at the same time or in the same proportion. The split follows the shape of your demand. If most of your work is urgent — burst pipes, lockouts, breakdowns — Google should take the larger share, because it meets existing demand at the moment it appears, and Facebook should take a smaller share aimed at planned work and retargeting. If most of your work is planned — refits, installations, memberships, seasonal services — Facebook can take the larger share, because it builds the pipeline weeks before the customer would ever search. Our post on Google Local Services Ads covers the search side of that split, and our cost-per-lead breakdown lets you compare the two channels on the same scoreboard: cost per booked job, not cost per click.

The practical rule for a first split: start with the channel that matches your most profitable work, prove cost per customer there, then add the second channel with the profits of the first. A business that tries to launch both channels, a new website, and a review campaign in the same month usually launches none of them well. Sequence beats simultaneity. Facebook first if your pipeline is the constraint and your work is planned. Google first if your phone is quiet and your work is urgent. Either way, the second channel’s job is to catch what the first one misses: Facebook retargets Google’s visitors, and Google catches Facebook’s audience when they are finally ready to search. Together they are a machine. Apart, each is just an ad.

Frequently asked questions

Do Facebook ads work for local businesses?

Yes, for many of them — particularly businesses selling planned work to local customers, like home improvement, salons, clinics, gyms, and landscapers. They work when a new customer is worth clearly more than it costs to win one, the campaign is set up in Ads Manager with a leads objective, the audience is limited to the real service area, and leads get a reply within minutes. They fail when they are boosted posts optimised for likes, aimed at a whole city, with no follow-up system behind them.

How much do Facebook ads cost for a small business?

WordStream by LocaliQ’s 2026 benchmarks put the median cost per lead at $27.39 across industries, ranging from about $13.74 for real estate to $61.56 for dentists, with home improvement at $42.95. Work your budget backwards from customers: customers wanted, divided by the share of leads that buy, multiplied by your industry’s cost per lead. A realistic first test budget for most local businesses is $300 to $500 a month, with a management fee counted separately if you hire help.

Are boosted posts the same as Facebook ads?

No. A boosted post is a simplified ad optimised for engagement — likes, comments, and shares. A campaign built in Ads Manager can optimise for leads or sales, with full control over audience, placements, creative, and destination. Boosts can amplify a post to people who already know you, but they are a poor way to win cold strangers as customers. For lead generation, use Ads Manager.

How long do Facebook ads take to work?

Expect the first leads within the first two weeks at a modest budget, a reliable cost per lead after roughly 45 days, and a real verdict on cost per customer after about 90 days. Do not pause an ad until it has spent at least twice your target cost per lead, and judge the campaign on customers won — not likes, clicks, or even raw leads — at the end of the test period.

Do I need a website to run Facebook ads?

Not strictly. Meta’s native lead forms collect names and phone numbers inside Facebook and Instagram without a website. A website with the Meta Pixel and Conversions API gives you better tracking, a place to send people who want more detail, and a landing page built for one offer — and it is the better setup for most businesses. What you cannot skip is a destination built for one action: a lead form or a single-offer landing page, never a general homepage.

Should I hire an agency or run Facebook ads myself?

Run them yourself for the first test if you can spare a few hours a week: one campaign, one ad set, one offer, and a fast follow-up habit. You will learn what a good lead looks like for your business, which is knowledge no agency can rent you. Hire help when the test works and you want to scale, or when you know you will not do the follow-up yourself — but keep the management fee separate from the ad budget, and judge the agency on cost per customer, not on reports full of clicks.

Sources

WordStream by LocaliQ, Facebook Ads Benchmarks 2026 (452 United States lead campaigns, April 2025 to June 2026): median $27.39 cost per lead, $1.80 lead-campaign cost per click, 2.70% lead-campaign click-through rate, 8.54% lead-campaign conversion rate; industry medians including real estate $13.74, health and fitness $27.11, home improvement $42.95, beauty and personal care $50.91, and dentists $61.56. Meta Q2 2026 earnings, as reported in a 2026 Coupler.io Facebook ads statistics roundup: average price per ad up 12% year on year in Q1 and Q2 2026, with impressions up 19% and 14% respectively. Hometown Digital, Meta Ads for Local Businesses guide, 2026 (agency-reported findings, not platform data): native lead forms averaging $34.10 per lead against $45.80 for video-view campaigns, and a roughly 15-mile radius delivering about 40% cheaper leads than city-wide targeting. Every figure in this post carries its source year above; where a figure comes from an agency guide rather than platform benchmark data, that is stated.