Google Ads vs SEO vs Referrals: Cost Per Patient (2026)

$212.63 is what one agency's dental accounts paid for a single Invisalign lead, and a lead is not a patient. It's a phone number attached to a human being who may never pick up when you call back. That figure comes from Delmain's 120-day benchmark across its own dental ad accounts, which also clocked branded search at $7.64, general dentistry at $50.60 and implants at $107.63 per lead. A 25x spread inside one channel, and every one of those numbers is a cost per lead, not a cost per booked patient. The agency says so plainly. Most practice owners quoting those numbers do not.

Here's our thesis, and we know it will annoy a few vendors: for most single-location and two-doctor general practices, the SEO retainer is the weakest line on the marketing budget, and that money does more work funded into referrals. Not because SEO doesn't matter. Because what it actually does is cheaper than what you're paying for it.

$50.60 Buys a Ringing Phone. That's All.

The gap between a lead and a patient is where marketing budgets go to die. A lead becomes a patient only after someone answers the phone, offers a time that works, and the person shows up. Three conversion steps, each one leaky, each one run by your front desk rather than your agency.

Run it. At $50.60 per general dentistry lead, a $5,060 quarterly ad budget buys 100 leads. Say your team books 60 of them, which is a decent answer-and-book rate for cold inbound. Say 75% of those keep the appointment. That's 45 new patients. Your cost per acquired patient is $112.44. Add a modest $500 a month in agency management and the quarter costs $6,560, which is $145.78 a patient.

That lands inside the $100 to $300 Google Ads range that DentX compiled across the industry, alongside $100 to $200 for organic and $0 to $50 for referrals. Worth saying out loud: DentX discloses those are compiled ranges, not one measured dataset. Treat them as a sanity check on your own numbers, not as a target.

Implants change the shape of the deal but not the logic. At $107.63 a lead, $10,763 buys 100 implant leads. Book 40 consults, hold 30 of them, close 40% and you have 12 cases. That's $896.92 in acquisition cost per case. On $4,500 of case value, you spent 19.9% of revenue to get the chair filled. That's defensible if your close rate holds. Drop the close rate to 25% and you're at $1,435 a case, and the math stops working.

81.9% Chose Their Dentist Because a Person Told Them To

A peer-reviewed survey of 520 patients across four clinics found that 81.9% picked their clinic on a personal recommendation, against 6.9% via the internet and 2.5% via advertisement. Those numbers are not close. They're not even in the same category of number.

Now hold that next to the ADA's 2024 Communications Trend Report, which surveyed 1,008 consumers in March 2024 at a ±3% margin of error and found 73% naming Google as their most trusted source when selecting a dentist. Those two findings look contradictory until you notice they're measuring different moments. One measures where the name came from. The other measures where the decision got confirmed.

Google isn't where patients find you. It's where they decide whether to believe the person who told them about you.

That reframe should change your budget. If Google's real job is confirmation, then you need a complete Google Business Profile, a steady flow of recent reviews, correct hours, a site that loads on a phone and pages that answer the three questions a referred patient actually has: do you take their plan, when can they come in, and does the office look like a place they'd bring their kid. That is a defined, finishable scope of work. It is not a $2,500-a-month indefinite retainer.

What Most Owners Get Wrong: Pricing Leads Like Patients

The specific mistake, and we see it most often in owners who are two to seven years into ownership and running ads without an in-house marketing person: quoting cost per lead as if it were cost per patient, then comparing it to a referral.

It goes like this. An owner reads $50.60 and mentally files Google Ads as a $50 patient. A referred patient, at the DentX range, costs up to $50 too. So ads look like parity, and parity at scale looks like growth. They raise the budget.

But $50.60 was never the patient price. The patient price in that same scenario was $145.78 with management fees in, which is nearly three times what the owner thought. Worse, the referral number they compared it to was real, because a referred patient is already past the trust step and shows up at a materially higher rate than a cold ad click.

The second mistake inside the first: nobody is costing the front desk. Those 100 leads generated something like 180 call attempts, 40 voicemails and a pile of after-hours forms. That labor is real and it lands on people you already pay, which is exactly why it never shows up in the cost-per-patient calculation. Our view: if you aren't tracking answer rate and booked-to-kept rate per channel, your cost per patient is a guess with a decimal point on it.

$8,008 and 180 Referred Patients

Here's the comparison we'd actually run. Take a two-doctor practice that funds referrals like a channel instead of hoping for them. Budget: $3,600 a year for a part-time hour a week of relationship work, $1,800 in handwritten notes and small thank-you gifts, $1,200 for review request automation, $1,408 in specialist and physician relationship costs including lunches and printed materials. Total $8,008.

That program yields 180 new patients in a year. Cost per patient: $44.49.

Now run Google Ads to the same 180 patients. At $145.78 each including management, that's $26,240 a year. Same patient count, $18,232 more.

Add the SEO retainer at $2,500 a month, $30,000 a year. If it delivers 200 new patients, that's $150 each. If it delivers 120, it's $250 each. Most owners genuinely do not know which of those two numbers they're living in, because organic and referral traffic both arrive as "found us online" and get attributed to whoever sends an invoice.

Across all three channels the practice spends $64,248 for roughly 520 new patients, a blended $123.55 each. The blended number looks fine. The blend is hiding that one channel is doing the same job for a third of the price.

Our call: cut the SEO retainer to $12,000 a year for a defined scope, the profile, the reviews, the site speed, the location pages, and move the $18,000 into referral and reactivation work. Even if the referral channel gets three times more expensive as you scale it, to $133.47 a patient, that $18,000 still buys about 135 new patients. We would take that trade every time, and we wouldn't take the retainer.

Rising Cost Per Patient: Which Kind Is Yours?

Acquisition cost goes up. The question that decides whether you manage it or restructure around it is our Drift vs. Erosion test.

Drift is your cost per lead creeping from $50.60 to $53 over a year because two more practices entered the auction. You respond operationally: tighten match types, cut the dayparts your phone isn't staffed for, lift the booked-to-kept rate from 75% to 82% and you've clawed the whole increase back.

Erosion is structural. A DSO moves into your zip with a budget that doesn't need the campaign to pay for itself this quarter. Your cost per lead goes to $60 and keeps going. At $60 a lead with the same conversion, cost per patient moves from $112.44 to $133.33 before management fees. No bid tuning fixes that, because the other bidder isn't solving for the same thing you are.

The percentage-of-spend agency fee is erosion by design. Your fee rises automatically with your spend, and your spend rises automatically with the auction, so you pay more per patient precisely when each patient is getting more expensive. Treating that as drift and waiting it out is the expensive mistake. Renegotiate to a flat fee or walk.

Do This Before Your Next Ad Invoice

This week, pull 90 days of call logs and count three numbers per channel: leads, booked, kept. Divide total channel spend, including management fees, by kept patients only, and put that number next to $44.49. Then ask your SEO vendor, in writing, for the specific scope they're delivering each month and the new patient count they're claiming, and if neither answer is concrete, convert that retainer into a defined project and redirect the difference. Fund the referral program on Monday, not after the next quarter's numbers come in.

If you want to see how your marketing spend sits inside total overhead before you move money around, run your own numbers through our free overhead calculator. Marketing is one of the few overhead lines you can genuinely reprice in a quarter, and knowing whether you're at 3% or 7% of collections changes which of these channels you can afford to be wrong about.

What we're tracking next: whether branded search cost per lead holds near $7.64 as more DSO money enters local auctions, because branded search is the one ad spend we'd defend without argument. It's the cheapest way to make sure the patient your hygienist's neighbor referred actually lands on your site instead of a competitor's. If branded cost per lead starts climbing past $15 in a market, that's a signal the confirmation step is being bought out from under you, and we'd rewrite this entire recommendation.