PPC
You can buy patients this month. The question is what each one costs.
Paid search is the fastest thing we sell. With the tracking connected, an account can be buying booked appointments inside its first month. Nothing else on this site moves at that speed.
Most practices running ads can give you two numbers: what they spent last quarter, and roughly how many enquiries came in. The figure between those two decides whether the quarter was worth it, and it is usually the one missing.
There is a catch, and it belongs on the first screen rather than in month four of a contract. Paid search stops the day the budget stops. What survives is the tracking and the pages the ads pointed at, so we build those on the way through.
What healthcare PPC buys, once you take the dashboard away.
Position at the moment somebody searches, for as long as you keep paying for it. That is the entire product. Everything else is an argument about which searches are worth buying and what an appointment ends up costing.
Most healthcare PPC gets sold and reported on the wrong end of that argument. Impressions, clicks, click-through rate, cost per lead. The first three are plainly vanity metrics. Cost per lead is the one that fools good operators, because it looks like an outcome and behaves like one, right up to the moment somebody asks how many of those leads sat in a chair.
We run an account where we cannot answer that question. It is further down this page, with the figure and the reason. Closing that gap is tracking and attribution work, and it happens before a bid does.
Paid search also gets confused with the things it sits next to. Organic search takes quarters and then keeps working. Ranking in the map near each location is its own discipline, run per address. Getting quoted inside an assistant’s answer is a fourth thing again. Ads can sit on top of all three, and none of them substitutes for another. The services overview shows how we group them.
What we run
Three things, and the first one is not a campaign.
The conversion is the booked appointment, and it is set before we touch a campaign
An account buys whatever you tell it is a conversion. Tell it a click and it will buy clicks all day, cheerfully. So the first job is making the appointment a trackable event: calls tied to the campaign that produced them, forms tied to the page that received them, the booking system counted where one exists. Where a practice cannot measure that yet, we say so, and we do not let the reporting imply more than it knows.
Campaigns split by location, by procedure, and by language
One campaign averaging across five sites and nine procedures hides the two that are losing money. We split the account along the lines that move the price. In a dermatology account we run, the split that mattered turned out to be language. Ads written in Spanish returned a 9.2% click-through rate. The English versions of the same ads returned 4.3%, and cost more per lead. Nobody had run that group on its own before, so nobody could see it.
The landing page, which is usually the cheapest thing in the account to fix
Plenty of the accounts we run today arrived paying a fair price for the click and losing the visit. The ad promises one procedure and the click lands on a home page. Booking takes three more clicks. Lifting a landing page's conversion rate does the same thing to your cost per appointment as cutting the cost per click, it is generally faster, and it does not have to be bought again next month.
Three, rather than a menu. Bid strategies, negative keyword lists and ad copy tests all matter, and they matter perhaps a tenth as much as the three above. Most accounts we inherit have been tuned obsessively at the level that matters least.
What changed in two accounts.
Both are ours. We do not name clients, so the descriptions are as specific as they get. The figures are exact and dated, and each one carries the caveat it came with.
$122.36 → $8.80
- Appointments booked through ads
- 49,320
- Ad impressions across the period
- 3.1M
- Extra appointments per $1,000/mo, at the end
- ~114
We ran the paid side from January 2024, alongside search and analytics. The conversion was the booked appointment from month one, which is the only reason the cost could be driven anywhere at all. You cannot optimise toward a number nobody is counting. We cut that cost by changing what the account was told to buy, and not by hunting cheaper clicks.
By the end of the period the account was limited by its budget rather than by its performance. That is a rare place for a paid account to sit, and it changes the conversation entirely. At $8.80 an appointment, another $1,000 a month bought roughly 114 more visits. The question stopped being whether the ads worked and became how much of them the group wanted to buy.
One account, at this scale, once. The figures describe January 2024 to April 2026. There has been a softer stretch since that review, so we quote the arc and never this month’s number. The arc is the version we can stand behind.
And a Manhattan dermatology practice, where the average lied
Between 15 February and 20 May 2026 we ran paid search for a Manhattan dermatology practice. It produced 628 leads on $6,495 of spend, $10.34 apiece blended, with cost per lead falling 45% across the period: $19.80 in February, $10.86 in May.
The blended figure is true and it hides most of what we learned. One ad group, the Spanish-language one, returned 9.2% click-through against 4.3% for its English equivalent, at $9.68 a lead. One ZIP code produced 499 of the 628 leads. Nearly four fifths of what the account produced came out of a few square miles, and no report set to show an account average would ever have surfaced it.
In April the cost per lead rose 27% over March on flat spend. Nothing had broken. We reported the rise, changed nothing, and it came back down in May. Rebuilding an account around one bad month is a good way to end up with a worse account.
To be exact about it: leads, not booked patients. Nothing down-funnel is connected in that account, which makes $10.34 a media-efficiency figure and nothing at all about the cost of acquiring a patient. Connecting it is the defined next step. We would rather draw the distinction than let you assume the better number.
Both accounts are told in full, caveats included, on the results page.
Who it's for
Practices where the searching and the deciding happen close together.
- Urgent care
A patient who decides in minutes. Paid holds the slot at the top while organic catches up underneath it.
- Behavioral health
Admissions, not enquiries. The distance between those two decides whether the spend was worth anything.
- Med spas
The ads are the easy part. Money leaks at the booking step, after the click.
- Cosmetic dentists
High ticket, low frequency. One implant case can pay for a quarter of clicks.
- Dermatology
Two buyers in one account, medical and cosmetic, and they should not be sharing a budget.
- Plastic surgeons
Months of research, then one consultation. The held consultation is the conversion.
Our paid proof sits in two of those six: the urgent care group and the dermatology practice. For the other four we have the method and the economics and no paid case of our own to point at. Their pages say so themselves, which is a habit we would rather keep than explain away on a call.
The boundary
What paid search doesn’t do.
- It stops the day the budget stops
- Paid search is a faucet. Turn it off and the appointments it was buying end that week, with nothing left behind except whatever you built while it ran. That is why the tracking and the landing pages are part of the work rather than an upsell. Those survive a pause. The campaigns do not, and an agency that lets you think otherwise has rented you an asset and called it one.
- It cannot outbid a bad landing page, or a phone nobody answers
- Bidding buys the click. What happens after the click is your site and your front desk, and no bid strategy compensates for either. In the accounts we run, enquiry volume climbing while booked appointments sit still is common enough that we look for it before we look at the bidding. When it happens, the cause is downstream of anything we can buy, and spending more only widens the gap.
- A cheaper lead is not a cheaper patient
- Cost per lead is easy to move. Broaden a match type, chase a cheaper term, and the number on the report improves while the quality behind it falls. With nothing connecting the form to the chair, that is indistinguishable from progress. One of the two accounts we run sits in exactly that position, which is also why we set the conversion before we set the budget.
- It comes with no promised cost per acquisition, from us or from anyone
- An agency that quotes you a cost per appointment before it has seen your account has invented the number. Your market, the practices bidding against you and your own close rate set that price, and we know none of the three yet. What we will commit to is the structure, the reporting, and saying when a search is not worth buying.
Questions
What practices ask before they hand over a budget.
- How much should we be spending on Google Ads?
- Work backwards. Decide how many new patients a month you want, divide by the rate at which enquiries turn into patients, then multiply by what an enquiry costs in your market. That gives you a budget you can defend instead of one somebody guessed. If the middle numbers are unknown, finding them is what the first month of an engagement is for.
- What should a patient cost us to acquire?
- That depends on your market, the practices bidding against you and what you charge, so treat any published figure as a reference point and not a target. The number worth arguing about is what a booked appointment costs. For a multi-location urgent care group we ran, that figure fell from $122.36 in January 2024 to $8.80 by spring 2026. Urgent care is high volume and low ticket, and your number will look nothing like it.
- How fast does paid search work?
- Faster than anything else we sell. With conversion tracking in place, an account can be buying appointments in its first month. Give it about ninety days before you judge the price. In the accounts we run, the early weeks go on buying information: which searches convert, which ones only look like they do, and which of your pages the traffic should be landing on.
- What happens if we stop?
- The appointments the ads were buying stop that week. Paid search does not decay gently the way organic search does. What you keep is the conversion tracking, the landing pages and the knowledge of which searches actually produce patients in your market. That is worth having. It is not worth the media spend on its own.
- Do we need separate landing pages, or can the ads point at our website?
- Usually separate, and not because agencies enjoy building them. A practice site is organised around the practice. An ad is a promise about one procedure, and the page has to keep that promise inside a screen. Where the existing site already has a page that does the job, we point the ads at it and spend the money somewhere it is needed.
Start with the analysis.
Send us the account. You get back what you are paying for a lead and, where the data allows it, for an appointment. Which campaigns are carrying the account and which are being carried. What is going to searches that were never going to convert. You get that in writing, and a contract is a separate conversation.
If your question is arithmetic rather than diagnosis, the ROI calculator works backwards from a patient target to the spend it implies.