Google Ads are worth it when the new patients they bring pay back their cost, and your schedule answers that before any ad report does. One finished month gives a first verdict, and a phone line used only in the ads settles a close one.
Quick answer
Take the month before last. Divide its ad spend by what a new patient's first 90 days leave after the cost of those visits: your break-even patients. Compare them with the new patients who said Google and came in. More than all of them: no payback inside 90 days. Half or fewer: a provisional yes. In between: a phone line used only in the ads decides.
What do you count, and from which month?
I read the month before last because Google files each call or form the ads counted under the date of the click (Google, checked 2026-10-01), and late bookings land in the next month.
Illustrative: every number below is made up for an imaginary two-doctor foot and ankle practice, and none comes from a client.
| Row | What it counts | Where to find it | Made-up month |
|---|---|---|---|
| 1 | New patients who said Google and came in | Practice software, first visits marked arrived | 21 |
| 2 | New patients who said Google and booked | Same report, any status | 28 |
| 3 | Calls and forms the ads counted | Conversions column or agency report | 61 |
| 4 | Ad spend | Ad account or invoice | $5,130 |
Rows 1 and 2 take any source your desk records as Google: search, Maps or an ad, and "online" only if your desk uses that word for Google. Every cell is a count or a total, so the sheet can go to an agency or to me without naming a patient.
What is one new patient worth to you in 90 days?
No benchmark answers this; your records do. Take new patients whose first visit was four to six months ago and divide what their first 90 days collected (payments, not charges) by their number. Subtract what those visits cost that you would not have spent otherwise, such as supplies, lab fees and hourly staff; rent and salaried time stay out.
In the made-up practice, $655 collected minus $230 to deliver leaves $425, so $5,130 ÷ $425 gives 12.1 break-even patients; paying back takes 13 from the ads.
Which verdict does your month land in?
More break-even patients than row 1 is a no, unless your desk often leaves the source blank. Half of row 1 or fewer is a provisional yes, if at least half of your Google patients came through an ad. If your Business Profile (the free listing that shows on Maps) brings most of your calls, assume only a third and compare with a third of row 1. Anything between goes to the phone line below.
The made-up practice needs 12.1 of its 21 Google patients, about 57 percent: the middle, so the phone line decides.
If one patient more or less in row 1 would flip the verdict, the month is too thin, so add the month before it. Heavy spend on searches for your practice's own name flatters any verdict, since those people had already chosen you.
How do you tell ad patients from Maps and unpaid Google?
Your desk's "Google" mixes ads, Maps and unpaid search. A second number that rings at the front desk splits them. Put it only on the ads' call button (a call asset) and on one page that only the ads link to, and mark that page noindex, a tag that keeps it out of unpaid search (Google, checked 2026-10-01). Have the desk record "Google ad line" for every call on that number and every form from that page. With call reporting (Google's own call tracking) on, the ad shows a Google forwarding number that rings through to that line (Google, checked 2026-10-01).
After two finished months, count the new patients with that source who came in. It misses ad patients who dialed the main number, so it is a floor. If it alone covers the break-even patients for those two months, the ads are worth it with no assumption left; if not, they are unproven at this spend.
Why does the ad report count far more than your schedule shows?
My line is row 3 at more than twice row 1, and the made-up month has 61 against 21. Row 2 is a ceiling on what the ads brought, since the desk writes "Google" for Maps and unpaid search too, so at least 33 of the 61 brought no new patient who booked; read a gap of one or two as zero. Why the ad report shows more leads than your own records covers what fills it.
The gap matters most under Google's automatic bidding for the most conversions, Google's word for counted results (Google, checked 2026-10-01), which then buys more of row 3: an account bidding on calls and forms drifts toward cheap results. Two settings decide what lands in row 3. Calls and forms both counted in the Conversions column add up (Google, checked 2026-10-01), and a call from an ad counts after 60 seconds by default (Google, checked 2026-10-01). Time five booking calls at your desk: if the shortest runs well past a minute, that setting counts calls that ended before anyone could book.
Those settings sharpen what the bidding aims at; the target itself changes only when the account hears which clicks became patients. When the desk marks a first visit arrived, a small job sends Google the click id (a code Google adds to an ad's link on each click), a dollar figure and the arrival time, nothing from the chart. How that upload from your schedule to Google works has the mechanics; the loop is patient conversion tracking for medical practices, and how I built it for one practice shows the shape.
For the made-up foot and ankle practice: 12.1 needed against 21, a second phone line to order now, and a verdict in two months that the schedule, not the ad report, will give.
Tags
Frequently asked questions
What cost per new patient makes Google Ads worth it?
A finished month's ad spend divided by the number of new patients who said Google and came in should be at or below half of what a new patient's first 90 days leave the practice, net of what those visits cost to deliver. That is a provisional yes, assuming at least half of those patients came through an ad; above the full 90-day amount means no payback inside 90 days, and anything between needs a phone line used only in the ads.
What if my calendar is already full?
Then the schedule measures your capacity, not the ads. When every new-patient slot is taken, the count of new patients who said Google is capped by the calendar, and a yes may only mean the ads filled slots other new patients would have taken, so read a month that left new-patient slots open instead.
Can I run Google Ads without conversion tracking?
Yes. The ads still run, but the account can then bid only for clicks, visibility or video views, and with only call and form tracking it learns which clicks call or write in, not which become patients. A phone line used only in the ads counts ad patients on your schedule without any conversion tracking.
Do patients who found me on Google Maps count as Google Ads patients?
Not unless they came through an ad, and asking them will not tell you, because ads also appear on Maps and a patient may not know whether the listing they tapped was an ad. Give the ads a phone number and a landing page used nowhere else, then count the new patients who arrived through those: that count is a floor on what the ads brought.
What should I tell my agency if most counted calls never become patients?
Ask for every action that feeds the Conversions column and whether it counts one result per click or every one, a call length that matches how long a booking call takes at your desk, and ads judged on new patients rather than on cost per lead (spend divided by counted calls and forms). Those sharpen what automatic bidding aims at, but only sending back which clicks became patients changes the target itself.
Can my schedule show whether Google Ads pay off for long treatment plans?
Only partly. A 90-day window captures the start of a plan that bills over a year, so the schedule can say whether the first 90 days pay for the ads, and anything beyond that rests on your own forecast of what a patient is worth later.