Facebook ads for dentists

The split between the two apps is not a matter of taste for a dental practice, it is a matter of which procedures pay for the advertising. Cleanings are spread evenly across ages; the revenue is not. Implants, dentures, full-arch and crown-and-bridge work concentrate in the cohort where Pew's November 2025 survey puts Facebook at 74% of 50- to 64-year-olds and 57% of those 65 and over, against 40% and 19% for Instagram. If your growth plan is restorative, this is where the budget goes and the Instagram half is a supporting buy. If it is an orthodontic practice selling clear aligners to people in their twenties, read the Instagram guide first and treat this one as the supporting buy. Most general practices are somewhere in between and should run both, with the offer deciding which app each ad set sits on rather than a flat split.

The second thing Facebook buys you is the phone. Call ads are a mobile-only format, and Meta delivers them across Facebook Feed, Facebook Marketplace, Facebook Reels and Facebook Video feed as well as Instagram Feed and Instagram Stories — so two of the six surfaces that can ring your front desk are Facebook inventory with no Instagram equivalent. For an emergency line that is not a detail. It is also the one place in this whole playbook where the reporting is unusually honest: Ads Manager separates calls placed from 20-second calls and 60-second calls, and a 60-second call is the only number on the platform that behaves anything like a booked appointment. Optimise and judge on that, not on the click.

Third, shape. Facebook stopped being the landscape platform: Meta's own ads guide recommends 4:5 at 1440 × 1800 for Facebook Feed, the same portrait shape as Instagram feed, and that is where the bulk of a practice's creative should be rendered. The exception is 1.91:1, which in our own publisher routes to Facebook's right-hand column and search results and to no Instagram position at all. That is worth exactly one asset for most practices — the emergency ad — because somebody who has just typed a question about a tooth at six in the morning is on the other side of that surface and cannot be reached from the other app.

Finally, the audience you already own. A practice that has been on Facebook for a decade has a Page with engagers, reviewers and event attendees attached to it, and that is the warm pool the middle of the funnel runs on. It costs nothing, it does not depend on uploading anything about a patient, and it is the single most-skipped asset in local dental advertising — most practices spend their entire budget talking to strangers while the people who already followed them see nothing.

Ads Manager offers six objectives — Awareness, Traffic, Engagement, Leads, App promotion and Sales — and a dental practice has a real use for two of them, an occasional use for a third, and no use at all for the rest. The objective is not a label describing what you want; it is the instruction that decides who Meta shows the ad to, so choosing the flattering one is how an account ends up with 400 video views and no appointments.

Leads is the default and should be most of the budget. Seven of the ten ads on the examples page were written for it, and the reason is structural rather than stylistic: nothing in dentistry completes in the ad's own session. The most an ad can cause is a form, a call or a booking, which is precisely what Leads optimises toward.

The interesting one is Awareness, and it is only rational here because of the rule dentists escape. A practice's entire addressable market is the few thousand households that would plausibly drive to it — that is a shape you can actually buy when age and a three-mile radius are still available to you, and it is not a shape a housing advertiser can buy at all. An Awareness buy against a genuinely small, genuinely local audience is a different product from the Awareness buy a national brand makes with the same button.

This is the part of a dental campaign that is actually industry-specific, and it is where most accounts go wrong before a single targeting decision is made. A practice sells a dozen distinguishable things to a dozen distinguishable people, and they do not respond to the same argument. The reader who has been quoted $28,000 for full-arch implants and walked away is not weighing the same question as the reader who moved house last month and needs a cleaning; running one “new patient special” at both of them wastes the half of the budget spent on the first.

Each row below pins four decisions: who the offer is actually for, which angle family argues it best — every one of those is a page in the angle library, with the sub-angles spelled out — which objective buys it, and, where the examples page has one, a finished ad that shows the whole thing assembled. Twelve rows is not a checklist to run all at once. Pick two or three that match what your chairs are empty for this quarter.

One structural note before the table: the angle is a property of the offer, not of the practice. The same clinic advertising implants in March and year-end benefits in November should be making two completely different kinds of argument, and an agency that runs one house style across both is leaving the second one on the table.

Start from the rule that decides everything else. Meta's Marketing API accepts exactly four special ad categories — HOUSING, EMPLOYMENT, FINANCIAL_PRODUCTS_SERVICES and ISSUES_ELECTIONS_POLITICS — and health is not among them. A dental practice therefore keeps every lever a housing advertiser loses: age, a radius smaller than fifteen miles, lookalikes, exclusions, the lot. If a guide tells you to leave age open and buy a wide radius because “Meta restricts health advertisers”, it has confused you with a real estate agent.

The corollary is the part that catches people. Because the ad set can qualify the audience precisely, the creative does not have to — and the creative is where the enforcement actually lives. Meta's personal attributes policy prohibits an ad that asserts or implies a person's physical or mental health, medical conditions, or vulnerable financial status, among much else. Meta's own worked example is the pair every dentist needs: “Do you have diabetes?” is non-compliant, “New diabetes treatment available” is compliant. Now run the dental vernacular through it. “Missing teeth?” “Embarrassed to smile?” “Afraid of the dentist?” “No insurance?” — that last one is not even about teeth, it implies vulnerable financial status under a separate clause of the same policy. Every one of them fails on the second person. State the service, describe the patient in the third person, and the identical ad passes.

So the audience brief and the ad copy are two different documents, and the discipline is keeping them apart. “Adults who have not sat in a dental chair in a decade” is a superb targeting and writing instrument. It is not a sentence that may appear in the ad.

The one place this can go wrong outside the ad account is patient data, and it deserves stating plainly because the platform will let you do it. Whether you may upload a patient list at all is a HIPAA question for whoever handles your compliance, not something an ad platform can answer for you. And Meta's own Business Tools terms prohibit sharing data that is based on health information — including the names you choose for events, conversions and custom audiences. A custom conversion called “Sleep Apnea Enquiry” or an audience called “Denture Patients” is the violation, even when the underlying data is clean. Name events for what happened, never for who did it.

Meta gives you two Instant Form types and the difference is the whole decision. “More volume” is built to be filled and submitted in one thumb movement on a phone; “Higher intent” adds a review step that gives the person a chance to confirm what they entered. The pre-filled profile data is what makes the first one cheap and also what makes it worthless at the desk — a name and a number the reader never actually looked at is a lead that does not remember filling anything in when you ring back.

For most dental offers, use Instant Form on Higher intent. The review step is a filter you are buying deliberately: it will lower your lead count and raise the proportion of them that answer the phone, and in a business where the front desk is doing the conversion by hand, that is the trade you want. Add two or three qualifying questions for the expensive cases — which arch, insured or not, how soon — and accept the volume drop.

Send the click to a landing page when the offer needs substantiation that will not fit in an ad. A membership plan has terms, exclusions and a renewal date; a same-day crown needs the mechanism explained before anyone books it; a published fee needs the conditions attached to it. Those are pages, and the ad's job is to earn the read. Do not send them to the homepage — a reader who tapped an ad about implant-retained dentures and lands on a rotating hero of a smiling family has been asked to start their search over.

Emergency is the exception to all of it, and it is not close. Someone with an abscess at six in the morning does not want a form, and a form does not want them either. Run call ads: mobile-only, delivered on Facebook Feed, Marketplace, Reels and Video feed as well as the Instagram surfaces, and reported in a way nothing else on this platform is — calls placed, 20-second calls and 60-second calls, separately. A 20-second call is usually a wrong number or a hang-up. A 60-second call is somebody talking to your front desk. Judge the campaign on the second number and staff the phone for the hours the ad runs, because an emergency ad that rings out at lunchtime is the most expensive thing in this guide.

The rule that changed most recently is the one worth leading on. On 22 July 2026 Meta moved its Health and Wellness standards from product-based enforcement to claims-based enforcement, and before-and-after transformation imagery for cosmetic products, procedures and surgeries became permitted rather than auto-rejected, provided the ad is targeted to adults 18 and over. For years this was the single most-wanted and most-refused format in cosmetic dentistry. It is available again — and what is still prohibited is the thing that used to travel with it: statements of inferiority about physical appearance, imagery that works by making the viewer feel worse about how they look, sensational or exaggerated claims, and any claim to cure or eliminate a disease. In other words the image is no longer the violation. The caption is.

The second thing to get right is whose face is in the ad. A stock photograph of a model with veneers no dentist produced is the default in this category and it is worth less than a badly-lit photo of your actual operatory, because the reader is choosing a building they will drive to and a person who will be holding the instrument. Show the room, show the team, show the machine if the machine is the point.

Third, restraint with numbers. One figure per ad, and only one you will honour at the desk — the ADA's Code of Professional Conduct requires that fees not be represented in a false or misleading manner, and state boards enforce their own advertising rules on top of that. A comparative price claim about somebody else's product (“less than most insurance premiums”) is a claim you would be publishing under your own name and expected to substantiate.

Fourth, and this is the one that makes the length rules below binding rather than advisory: you cannot see the frame your ad will be shown in. Facebook has nine image surfaces — Feed, in-stream video, video feeds, Marketplace, Stories, ads on Reels, search results, Business Explore and the right-hand column — against Instagram's four, and their recommendations disagree with each other. Facebook Feed wants 4:5 at 1440 × 1800, 50–150 characters of primary text and a 27-character headline, and shows no description line at all. Marketplace wants a square at 1080 × 1080, 125 characters, a 40-character headline and a 30-character description. Leave Advantage+ placements on, which almost every practice does, and one ad fills all nine from a single export and a single headline. So write to the strictest surface and compose the portrait so it survives being cropped square — the placement you did not think about is the one holding your reader.

It is worth knowing how hard that is, because it is not a criticism of any one run. Five Facebook batches across this cluster have now been measured the same way — dentists, real estate, gyms, SaaS and apparel — and of the 250 headlines they produced, 19 fit 27 characters and 238 fit 40. Of the 250 primary texts, none fits 150 and the shortest written is 200 characters. Forty characters is about the length a headline comes out at when nobody is counting; twenty-seven is not. Look at what the ones that do fit have in common — “It was never the pillow.” at 24 characters, on the night-guard ad — and the pattern is the same on every page in the cluster. Twenty-seven characters is not enough room for a claim. It is enough room for a noun.

Three stages, and the third one is the stage most practices never build. Cold does the volume, warm does the conversion, and the patient list is both an exclusion and an audience — excluded from the acquisition offers so you stop paying to discount treatment for people who already come to you, and targeted directly for the campaigns that only make sense to someone already in the chart.

Keep the structure this simple on purpose. A practice's budget is small relative to the audience arithmetic Meta's delivery system wants, and every additional ad set divides the same evidence into thinner piles. Variation belongs at the ad level, inside one ad set, not at the ad-set level.

Here is the arithmetic that decides how well your ads will work, and almost nobody advertising a dental practice has done it. Meta's delivery system needs roughly 50 optimisation events per ad set per week to leave the learning phase; below that the ad set sits in “learning limited”, which is not a penalty but is an admission that the system cannot optimise with what you are giving it. So take your own cost per lead, whatever it actually is, and multiply by 50. That is the weekly spend at which delivery is working with full information on one ad set. Divide what you can genuinely spend by that number and you have an honest measure of how much of Meta's optimisation you are buying.

For most single-location practices the answer is: a fraction of it. That is not a reason to stop; it is the reason the funnel above has three ad sets and not nine. Every dollar you split into another ad set halves the evidence in both. One campaign, one cold ad set, several ads inside it, and let the offer be the variable that changes between ads rather than between ad sets.

Then leave it alone. Pausing an ad set, or changing its optimisation event, its audience or its creative, counts as a significant edit and restarts the learning phase; budget and bid changes can too, depending on the size of the change. An account that gets adjusted every morning is an account that never finishes learning, and the daily fiddling feels like management while being the single largest self-inflicted cost in a small ad account.

Seasonality is real in this trade and it is worth planning the year around three moments: late October through December, when annual insurance maximums are about to expire and the urgency is a fact rather than a device; January, when benefits reset and the deductible conversation starts over; and the back-to-school window for family and paediatric offers. Emergency runs flat all year. Whitening follows other people's calendars — wedding season, graduation, the run-up to the holidays.

Every one of these is common, most of them are expensive, and the first three are the reason a dental account gets rejections that a plumber's account never sees.