You turned the ads on. The money is going out.
A few calls that went nowhere. A lot of cleanings when you wanted implants. Chairs that still have gaps in them on a Tuesday afternoon.
If that sounds familiar, you’re not alone. In our experience most dental campaigns don’t fail because the ads are bad. They fail for reasons that sit outside the ad account, which is why spending more rarely fixes them.
Here are the ten we see most, and how to test each one against your own numbers.
What a dental ad is actually buying
A paid ad is a placement you rent. You pay for the click whether or not the patient ever sits in your chair.
When someone is buying a product online, that’s a clean trade. The click and the sale happen in the same session, so you know by the evening what the ad was worth.
Dentistry doesn’t work like that. The ad buys an enquiry. What follows is a phone call, a consultation, a treatment plan and a decision about money, spread over weeks, and almost none of it reaches the ad account. Your dashboard covers the cheap part and goes quiet for the expensive one.
That blind spot is where the money goes.
Key takeaways
- A large group of patients who need work aren’t searching for you. They’re avoiding you.
- If a competitor could paste your ad onto their own site unchanged, it isn’t an ad.
- Cost per lead can’t tell a good campaign from a bad one. Cost per accepted case can.
- High-value dental decisions take weeks. A three-day test tells you nothing.
- How fast you call back decides who gets the conversation. Who calls back decides whether it becomes a case.
Reason #1: Advertising to everyone with teeth
The first thing we look at in a clinic’s account is usually one campaign, one budget, pointed at “people near here who might want a dentist.”
Then the owner wonders why every enquiry is a cleaning.
The reason is structural. The audience and the bid are set at ad set level. Put a full-arch case and a hygiene visit in the same ad set and they share one audience and one bid, so whichever produces cheap enquiries takes the delivery.
Splitting them into separate campaigns instead carries its own cost. A clinic’s conversion volume is small, and spreading it thin leaves every campaign with too little data to optimise on. Keep the campaign count low and do the separating below it.
- Give each treatment its own ad set. That’s where the audience and the bid live, and a treatment you can’t aim separately is a treatment you can’t control.
- Work out what you can pay per enquiry, per treatment. Case value multiplied by the share of enquiries that become cases. Use your own acceptance rates.
- Read your own search terms report before splitting by intent. People say words like cost and price mark a ready buyer. Don’t take it on faith. Tag which terms produced accepted cases, and split on what you find.
Stop buying “dental.” Start buying the case you actually want.
Reason #2: Expecting search to find the patient who is avoiding you
About one adult in nine has high dental fear (Journal of Dentistry, 2021, pooled across 31 population studies worldwide). Not a small niche, and not a cheap one: the same literature describes a vicious circle, where avoidance leads to worse oral health and eventually more invasive treatment.
Search only shows your ad to someone who typed something. A patient who has been putting this off for six years types nothing, until pain or a broken tooth forces it.
- Don’t expect search alone to fill high-value treatment. It reaches that patient only at the very end, against every other clinic bidding on the same moment.
- Look for the anxiety words in your search terms report. Nervous, anxious, sedation, scared. If they produced enquiries and you never bid on them, you found a gap. If they’re absent, don’t build a campaign on our say-so.
- Compare channels on accepted cases. An enquiry from someone who wasn’t looking for a dentist will look worse on every dashboard metric and can still be worth more.
Reason #3: Writing ads about yourself
Gentle. Caring. Painless. State of the art. Family friendly.
Every clinic in your city is running that ad. That’s the definition of invisible.
None of it is untrue. That’s the problem. A claim every competitor can also make doesn’t help a patient choose between you, and you’re paying by the click to make it.
Some of those words are worth bidding on. Gentle and painless are things a frightened patient types. They stay weak as a headline, because every clinic claims them.
- Name the situation, not the category. Someone missing a back tooth doesn’t see themselves in “quality dental care.” Whether that lifts your numbers is testable, and reason #6 is how you’d test it.
- Use your own photographs. Your team, your chairs, your cases. A stock model with perfect teeth is a picture of no clinic in particular, including yours.
- Lead with what a competitor can’t copy. How many of these you place a year. The equipment you have. Your hours. The languages your team speaks. The complicated version of the case others refer out. Claim only what you can substantiate, because these are the claims patients check.
Ask one question about your current ad. Could a competitor paste it onto their own site tomorrow without changing a word? If yes, it isn’t an ad. It’s wallpaper.
Reason #4: Letting the landing page kill the case
The ad buys the click. Everything after belongs to the page, and clinic pages break in the same three places.
- You send them to your homepage. They clicked an implant ad. Send them to an implant page.
- You give them six things to do. One treatment, one action. Book, or call. Not book, call, subscribe, follow, download and browse.
- You make them hunt for the phone number. One thumb, at night, on a phone.
Don’t take that on trust. Split the traffic between a dedicated treatment page and the homepage, leave it one full conversion cycle, and compare enquiries. If the treatment page wins, your ads were never the problem. If it doesn’t, you’ve learned something about your market that this post can’t tell you.
Reason #5: Judging the campaign before the cycle ends
Is your campaign failing? Or have you just not let it finish?
Days are the wrong unit. What matters is the conversion cycle: how long it takes a click to become a booking (platform bidding guidance). Judge anything before one full cycle and you’re reading noise.
Nobody prices a full-arch case on Tuesday and pays for it on Wednesday. They read. They compare clinics. They talk to their partner. They wait for money. Your own records hold the real number: measure the gap between first contact and accepted case, by treatment, and make that your review date.
Once a cycle has passed, the numbers point somewhere:
- Clicks fine, enquiries low? Look at the page and the offer.
- Clicks low, cost per click high? Look at the message and the audience.
- Enquiries fine, cases low? The problem is on the phone. See reason #10.
Reason #6: Changing four things at once
Most clinics change the headline, the image, the budget and the audience in the same week, watch the number move, and learn nothing about which change did it.
That’s not testing. That’s guessing with extra steps.
- Change one thing at a time. Otherwise the result can’t be attributed to anything.
- Test big differences, not small ones. Arithmetic, not preference: the smaller the true effect, the more traffic you need to see it. Clinic volumes can’t detect a small lift.
- Run every test past one full conversion cycle. Ending early doesn’t measure the campaign, it measures your patience.
Reason #7: Retargeting you might not be allowed to run
Every generic ads guide says the same thing. Upload your patient list, build a lookalike audience, chase the people who didn’t convert.
Check the rules first. Health advertising policy covers invasive procedures, surgery and injections, and for those ads it switches off uploaded customer lists, lookalike audiences, audience expansion and advertiser-built segments (health advertising policy). The policy doesn’t name dentistry. It names surgery, and implant placement is surgery.
The logic is easy to see. A list of everyone who visited your implant page is a list of people with a health condition.
- Don’t build audiences from treatment pages. That’s the exact thing the rules exist to stop.
- Read the policy against your own treatment mix. Where your work sits relative to “invasive” is the platform’s judgement, not ours. Confirm before you build on it.
- Keep what still works. Location, in-market segments and demographics are untouched by that policy.
Reason #8: Never sending your results back
Left alone, the system optimises toward whatever you told it to count. Usually a submitted form.
So it gets very good at finding people who submit forms. Wrong area, wrong treatment, no intention of turning up: it doesn’t know, because you never told it.
Offline conversion import closes that loop. You send back what actually happened in your practice after the click (importing offline conversions).
- Feed back attendance. The target stops being “filled in a form” and becomes “turned up.”
- Feed back accepted cases with values attached. Same mechanism, one step further down.
- Expect a project. It needs your practice management system and your ad account connected, which is why almost no single-site clinic has it.
Reason #9: Building the budget backwards
You’ll read that your weekly budget should be fifty times your target cost per acquisition. Nobody repeating it explains why fifty. Whatever its origin, it’s a rule about how much data an algorithm wants, not about what a dental case is worth, and most single-site clinics will never produce the weekly volume it assumes.
Build from your own case values. What’s a case worth, how many do you want a month, and how many enquiries does it take to get one? If you don’t know that last number, that’s the first job.
Then stop managing cost per lead. Here’s why, in arithmetic you can check against your own account:
- Clinic A pays four times the going rate per enquiry and accepts one case in four. Cost per case: 4 × 4 = 16.
- Clinic B pays the cheap rate and accepts one in thirty. Cost per case: 1 × 30 = 30.
Clinic A runs at roughly half the cost per case. Clinic B has the better-looking dashboard. Cost per lead cannot tell those two apart, which is the whole problem with managing to it.
Reason #10: Answering too slowly, and with the wrong person
This is the leak we see most, and it shows up in no ad account.
In business-to-business data, calling a web enquiry within five minutes instead of thirty makes you about 100 times more likely to reach the person at all, and 21 times more likely to qualify them. Inside the first hour alone, the odds of making contact fall by more than ten times (MIT Lead Response Management Study, 15,000 leads and 100,000 call attempts).
Later work by some of the same researchers, published in Harvard Business Review, found firms responding within an hour were roughly seven times more likely to qualify an enquiry than those responding later (Harvard Business Review). Treat the two as one body of work rather than two independent confirmations.
Neither was run on dental patients, so the exact multiples belong to those studies. The scale is what carries over, and so does the obvious part: a patient who contacted more than one clinic is having the conversation with whoever calls back first.
Speed only wins you that conversation. It doesn’t win the case. That part is decided by who picks up.
- Call back inside the window the data points at. The steepest measured drop sits between five and thirty minutes, so build the rota around minutes. “When reception gets a moment” is not a policy.
- Give it to a named person. Shared responsibility for callbacks means nobody owns the number.
- Listen to a week of your own enquiry calls. You’ll hear the real first question patients ask and whether your team can answer it. Better evidence about your practice than anything published.
- Confirm before the appointment. New patients miss appointments more often than returning ones (Cureus, 2024), and almost every patient your ads produce is a new patient.
Every patient your ads produce sits in the group most likely to miss the appointment. The confirmation call is part of the cost of the campaign.
FAQ
How fast should we call an ad enquiry? Minutes. The measurement behind that comes from business leads, so treat it as a direction and time your own.
Our cost per lead went up. Is something broken? Check accepted cases before you touch the bid. Cost per lead rising alongside cases means the campaign found better patients.
Which treatment should we advertise first? The one where you have capacity, margin and the clinical confidence to take volume. Advertising a treatment your diary can’t absorb creates enquiries you’ll answer badly, which costs more than not running the ad.
Are we even allowed to advertise? Check your dental council before your budget. Rules on testimonials, before-and-after images and price claims vary widely, and some regulators restrict dental advertising heavily. We’re not lawyers, so confirm yours before you spend.