47.66° N 117.43° W Summit · Profitable Growth
First-party data · 8 healthcare accounts · Q3 2026
Healthcare Google Ads cost per lead, by practice.
Across 6 healthcare Google Ads accounts I manage with clean conversion tracking, the median cost per tracked lead in Q3 2026 (1 July to 30 September) was $39.54 (N = 6). In Q3 2026, healthcare cost per lead ran from $6.08 to $118.78, a 19.5x spread (N = 6). Phone calls made up most of the primary conversions in all 8 healthcare accounts in Q3 2026, a median 100%, and 5 of the 8 counted calls as their only lead (N = 8). The middle half of healthcare accounts paid $2.05 to $6.98 a click in Q3 2026 (N = 8), and the median conversion rate was 12.1% (N = 6).
Edition: Q3 2026 · Published · Next refresh January 2027
Window to , fixed · Read-only API data, anonymized · Download the CSV
In these accounts a healthcare lead in Google Ads is mostly a phone call from the ad. So the cost per lead depends on which calls the account counts. The cheapest figure here counts any call from the ad, which likely includes calls from existing patients. Two of the 8 accounts are out of the cost-per-lead median: one behaves like it isn't counting all of its calls, and one counts actions that aren't leads. Read the table with that in mind, and read the method before you cite a row.
- $39.54 Median cost per tracked lead, Q3 2026 (N = 6)
- 19.5x Spread from cheapest to dearest counted lead (N = 6)
- 5 of 8 Counted phone calls as their only lead (N = 8)
- 35.4% Median search impression share (N = 8)
01 Cost per lead, by practice
Eight practices,
one lead definition each.
Every row is one account, so every row is N = 1. Rows carry only a broad group, specialist practice or general and ancillary care, so no row points to one practice. Read a row as one practice's Q3 2026, and read the median and quartiles as the benchmark. The last column says what that account counted as a lead, because the definition changes from row to row. The last row is out of the cost-per-lead median, and so is one account that has no row here.
Cost per tracked lead by account, Q3 2026
One account per row. H8 is out of the median while its tracking is on review.
| Code | Practice group | N | CPC | CTR | Conv. rate | Cost per lead | What counts as a lead |
|---|---|---|---|---|---|---|---|
| H1 | Specialist practice | 1 | $1.31 | 13.0% | 21.6% | $6.08 | Calls from ads only |
| H2 | Specialist practice | 1 | $2.08 | 14.0% | 21.0% | $9.89 | Calls from ads and website calls, plus forms |
| H3 | General and ancillary care | 1 | $1.95 | 16.8% | 10.7% | $18.34 | Call extension calls only |
| H4 | General and ancillary care | 1 | $7.74 | 10.6% | 12.8% | $60.73 | Calls from ads only |
| H5 | General and ancillary care | 1 | $7.88 | 3.2% | 11.5% | $68.76 | Call extension calls only |
| H7 | General and ancillary care | 1 | $6.73 | 5.8% | 5.7% | $118.78 | Two call actions plus a lead form |
| H8 | Specialist practice | 1 | $3.36 | 6.8% | 1.6%, excl. | Excluded | Excluded from median: undertracking on review |
| Median | 8 / 6 | Not shown | Not shown | 12.1% | $39.54 | ||
| Lower quartile | 8 / 6 | $2.05 | 6.6% | 10.9% | $12.00 | ||
| Upper quartile | 8 / 6 | $6.98 | 13.2% | 19.0% | $66.75 | ||
Some accounts belong to clients I name elsewhere on this site. They count in every median and total here, but I don't show them as their own row, because a row would publish a named client's numbers. On this page that's one of the 8, so the table has 7 rows. Where a summary figure would let you work that account's number back out from the 7 rows shown, the cell says not shown. CPC and CTR summaries cover all 8 accounts (N = 8). Conversion rate and cost per lead summaries cover the 6 with clean tracking (N = 6). H8 and the account without a row are left out of those two, for the reasons below and in the method. Medians come from each account's unrounded figure and are rounded once. Quartiles use the inclusive method, the same as the account-wide benchmark.
Why the cheap end is cheap
H1 (N = 1) counts every call that comes from the ad. I couldn't confirm a minimum call length on that action, so a call from an existing patient counts the same as a call from a new one. That likely explains part of the 21.6% conversion rate and the $6.08 lead. H2 (N = 1) counts two call sources at once, calls from the ad and calls from the website, and lands at 21.0% and $9.89. Both work as volume signals. I'd read either as a new-patient cost only after checking call length.
Why two accounts are left out
H8 (N = 1) converted 1.6% of clicks, and its all-conversions column ran 10.2x its primary conversions. That reads as calls going uncounted, so I leave it out of cost per lead and conversion rate while its tracking is on review. The other one has no row here. It counted actions that aren't leads as primary conversions, so its cost per lead isn't comparable and it's out too. Both stay in every tracking finding.
02 Against the vendor panel
Cheaper on the headline,
not like for like.
The vendor benchmark I compare against is LocaliQ's healthcare search ads report, last updated 13 January 2026. It covers 3,542 search campaigns running between 1 October 2024 and 30 September 2025, across 16 specialties. LocaliQ says its "averages" are medians. Several practice types in my sample have no LocaliQ category, so I compare the headline medians only.
| Metric | This report, median (Q3 2026) | LocaliQ, median (3,542 campaigns) |
|---|---|---|
| Cost per click | Middle half $2.05 to $6.98 (N = 8) | $5.64 |
| Click-through rate | Middle half 6.6% to 13.2% (N = 8) | 6.07% |
| Conversion rate | 12.1% (N = 6) | 8.09% |
| Cost per lead | $39.54 (N = 6) | $66.02 |
My median cost per lead sits $26.48 under theirs, at $39.54 (N = 6) against $66.02. Both figures are medians, but theirs is across campaigns and mine is across accounts, the windows don't overlap, and their report doesn't say what each campaign counted as a lead. My accounts convert at a higher rate. LocaliQ's 6.07% click-through rate sits below the lower quartile of mine, and its $5.64 click sits inside the middle half of mine.
03 What the lead is · N = 8
A healthcare lead is
mostly a phone call.
Calls were the majority of primary conversions in all 8 accounts in Q3 2026, and the median was 100% (N = 8). Five of the 8 counted calls and nothing else. Only 3 of 8 counted any conversion other than a call.
-
7 of 8
left conversion value unused. Three report zero value and four carry a flat placeholder of 1 or less per conversion (N = 8). A ROAS figure from any of them means nothing.
-
4.4x
was the median ratio of all conversions to primary conversions, from 1.7x to 19.8x (N = 5).
-
100%
of Q3 2026 spend went to Search in all 8 accounts, across 27 campaigns. No Performance Max, Display, Video or Demand Gen (N = 8).
When the conversion is a call, the call settings decide the cost per lead: which call actions count, whether a short call counts, and whether a booked patient ever makes it back to Google. None of those settings shows up in the cost per lead column.
04 Fixing the definition
Count the new patient.
The fix I run on medical accounts starts with one primary conversion that means a new patient, then feeds booked appointments back to Google without sending any protected health information. That second step raises legal questions before technical ones. The medical practice marketing library covers it question by question.
- Can a medical practice use Google Ads conversion tracking without violating HIPAA?
- How to track which ads produce booked patients without sending PHI
- Do I need a BAA with my call tracking provider?
- How to track phone calls as conversions in Google Ads
The full architecture, from call to booked patient to the bid, is the Lead Quality Stack. The wider diagnosis library is Lead Quality.
05 Where the money stops
Budget caps seven
of eight before rank.
Every account lost more than 28% of its search impression share to budget in Q3 2026, and budget loss beat rank loss in 7 of 8 (N = 8). The exception is H8, which lost 40.4% to rank against 35.0% to budget (N = 1). Median search impression share was 35.4%, from 21.2% to 47.4% (N = 8).
Where each account's search impressions went
Impression share won, lost to budget and lost to rank, spend-weighted across each account's campaigns.
| Code | Practice group | N | Search IS | Lost to budget | Lost to rank | Zero-conv. term spend | Mobile share of spend | Keyword spend with no QS | Hours served |
|---|---|---|---|---|---|---|---|---|---|
| H1 | Specialist practice | 1 | 33.0% | 45.1% | 21.9% | 10.0% | 99.2% | 47.9% | 15 |
| H2 | Specialist practice | 1 | 37.9% | 36.8% | 25.4% | 15.3% | 91.8% | 28.9% | 24 |
| H3 | General and ancillary care | 1 | 46.7% | 41.7% | 11.6% | 30.8% | 96.5% | 24.8% | 18 |
| H4 | General and ancillary care | 1 | 47.4% | 29.0% | 23.6% | 22.3% | 96.2% | 96.9% | 15 |
| H5 | General and ancillary care | 1 | 21.2% | 53.7% | 25.0% | 31.5% | 100% | 100% | 22 |
| H7 | General and ancillary care | 1 | 29.6% | 37.9% | 32.5% | 41.0% | 90.7% | 57.5% | 24 |
| H8 | Specialist practice | 1 | 24.6% | 35.0% | 40.4% | 41.6%, excl. | 86.3% | 77.5% | 24 |
| Median | 8 | 35.4% | 37.4% | 25.2% | 26.5% (N = 6) | 94.0% | 52.7% | ||
| Lower quartile | 8 | 28.4% | Not shown | 23.2% | 17.0% (N = 6) | Not shown | 27.9% | ||
| Upper quartile | 8 | Not shown | 42.5% | Not shown | 31.3% (N = 6) | 97.2% | 82.4% | ||
Impression share is weighted by spend across each account's campaigns. Zero-conversion term spend is the share of spend visible in the search-terms report that went to terms with 2 or more clicks and no conversion. Visible terms covered 34% to 71% of each account's spend (N = 8). Its median and quartiles cover the 6 accounts with clean tracking (N = 6); the H8 value is shown and left out. Summary cells marked not shown would give back the number of the account without a row. One account's keyword report covered 68% of its spend (N = 1).
Searches for somebody else
A median 26.5% of visible search-term spend went to terms with two or more clicks and no conversion, from 10.0% to 41.0% (N = 6, clean tracking). In three of the general and ancillary care accounts, the main pattern was people searching for a different provider by name (N = 3). A fourth mixed competitor names with broad research searches (N = 1).
Quality Score goes unreported
A median 52.7% of keyword spend sat on keywords with no Quality Score reported (N = 8). Three of the 8 had 77.5% to 100% of keyword spend in that state. Where a score was reported, the median account had 0% of keyword spend in the 1 to 3 band (N = 8).
06 Hours and devices
Office hours convert
cheaper.
Five of the 8 accounts serve ads fewer than 24 hours a day, between 15 and 22 hours (N = 8). In the 6 clean-tracking accounts with 30 or more conversions, the cost per lead between 8am and 5pm was lower than off-hours in 5 (N = 6). Off-hours cost per lead ran a median 1.4x the business-hours figure (N = 6). The widest gap was one specialist practice, $20.73 off-hours against $9.39 in business hours, or 2.2x (N = 1). One general-care account went the other way, at $90 off-hours against $126 during the day (N = 1). I'd test why before acting on it.
Mobile took a median 94.0% of spend, from 86.3% to 100% (N = 8). Mobile converted at a higher rate than desktop in 5 of the 6 clean-tracking accounts (N = 6). The same account was the exception again, with desktop at 10.1% against mobile at 5.2% (N = 1). Part of the mobile edge comes from counting calls, which mostly happen on phones.
07 Quotable findings
Lift these as written.
Each line carries its N and window, so it survives being quoted on its own.
- The median cost per tracked lead across 6 healthcare Google Ads accounts with clean tracking was $39.54 in Q3 2026 (N = 6).
- Among healthcare Google Ads accounts with clean tracking, Q3 2026 cost per lead ran from $6.08 to $118.78 across accounts, a 19.5x spread (N = 6).
- Phone calls were a median 100% of primary conversions across 8 healthcare Google Ads accounts in Q3 2026, and only 3 of 8 counted any conversion other than a call (N = 8).
- All 8 healthcare accounts put 100% of Q3 2026 spend into Search campaigns, and none ran Performance Max (N = 8).
- Every healthcare account lost more than 28% of its search impression share to budget in Q3 2026, and budget loss beat rank loss in 7 of 8 (N = 8).
- A median 26.5% of visible search-term spend in healthcare accounts with clean tracking went to terms with 2 or more clicks and zero conversions in Q3 2026 (N = 6).
- Conversion value was unused in 7 of 8 healthcare Google Ads accounts in Q3 2026, so their ROAS figures carry no meaning (N = 8).
- Off-hours cost per lead ran a median 1.4x the 8am to 5pm figure across healthcare accounts with clean tracking in Q3 2026 (N = 6).
08 Method
How the numbers
were taken.
Source and window
Google Ads API reads on healthcare accounts I have managed, read-only. No client is ever named. Every query was a SELECT. Nothing was written to any account. The window is fixed at 1 July to 30 September 2026, evaluated in each account's own time zone. It won't move when you reload the page.
Inclusion and exclusion
- Included: healthcare advertisers with an enabled account and at least $300 of spend in the window. Eight accounts qualified, one per practice.
- Not in the sample: closed or cancelled accounts. The API returns nothing for them, including history from before they closed.
- Accounts without a row: Some accounts belong to clients I name elsewhere on this site. They count in every median and total here, but I don't show them as their own row, because a row would publish a named client's numbers. That's one of the 8 here, so the tables and the CSV have 7 rows.
- Tracking flags: an account is flagged for overcounting if conversions exceed 40% of clicks or exceed clicks, for undertracking if it spent more than $1,000 with fewer than 3 conversions, and for missing tracking if it recorded zero conversions on more than $500. A flagged account is excluded from cost-per-lead figures and still counted in the tracking findings. An account can also be excluded on review when its conversion setup shows it isn't counting leads.
- Result this edition: no account tripped a mechanical flag. On review, 2 of the 8 are left out of cost per lead, conversion rate, zero-conversion term spend, device conversion rates and the hourly comparison. H8 shows undertracking: a 1.6% conversion rate and all conversions at 10.2x primary. The account without a row counted non-lead actions as primary conversions. Both stay in every tracking finding (N = 8). The account-wide benchmark uses the same two exclusions.
- One caveat on comparability: one account changed its conversion actions in the last two weeks of the window.
What counts as a lead
- Lead, or conversion. Whatever the account counts in its Conversions column, meaning its primary conversion actions. That's the number bidding optimizes toward.
- Cost per lead. Cost divided by primary conversions, for the whole window.
- Call share. Conversions from actions Google classifies as phone call leads, as a share of primary conversions. Everything else counted as primary is a non-call conversion.
- All conversions. Every conversion action, counted in bidding or not. The multiple is all conversions divided by primary conversions, reported for the 5 accounts where my data carries it (N = 5).
- Business hours. 8am to 5pm in the account's time zone. Reported only for clean-tracking accounts with 30 or more conversions (N = 6).
- Zero-conversion term. A search term with 2 or more clicks and no conversion in the window.
How medians were taken
I computed each account's figure first, took the median across accounts, and rounded once at the end. With 8 accounts the median is the mean of the fourth and fifth values. With 6, it's the mean of the third and fourth. Quartiles use the inclusive method (linear interpolation between ranked accounts), the same method as the account-wide benchmark. Nothing is weighted by spend, so the largest account counts once like the others. N = 8 is a small sample and one account per row. Read a row as one practice and the median as the benchmark.
Spend, revenue and conversion values stay private. Every published figure is a unit price, a ratio, a share or a count of accounts. These rules are the benchmarks standard.
09 How to cite this
Quote it, link it.
Crowe, C. (2026). Healthcare Google Ads Benchmarks: Cost per Lead by Practice Type (Q3 2026). Q3 2026 edition, published 3 October 2026. https://connercrowe.com/benchmarks/healthcare-google-ads/
If you use a figure, please link to this page so readers can check the N and the window. Quote the edition, because the next refresh in January 2027 will publish new numbers and mark these superseded.
The anonymized rows are in the CSV: code, descriptor, N, window, every per-account metric in the tables, and a column marking which rows are in the conversion medians. It has 7 rows for the 8 accounts, for the reason in the method.
Data licensed CC BY 4.0, attribution to Conner Crowe with a link.
10 Questions
What practices ask.
What is the average cost per lead for healthcare Google Ads in 2026?
How much does a Google Ads click cost for a medical practice?
What is a good conversion rate for healthcare Google Ads?
Are phone calls the main lead source for medical practice Google Ads?
Should a medical practice use Performance Max for Google Ads?
Why is my healthcare cost per lead so different from published benchmarks?
More benchmarks
- Google Ads account benchmarks The same Q3 2026 window, across every vertical I manage.
- Meta ads CPM by placement What each placement costs per thousand impressions.
- Home services cost per lead by trade The service-business counterpart to this report.
- All benchmarks The index, and the rules every report is held to.
Want this run on your account?
I'll show you what your account
counts as a lead.
Add me as a read-only user and I'll write the audit myself: which calls your account counts, how short a call can be and still count, and what your cost per lead is measuring, 5 to 7 pages, back within 24 hours of access. No call needed. If you'd rather talk first, a call is open too.