Healthcare Marketing ROI: The Metrics That Actually Matter

Here is a report a clinic owner gets every month, in some form, from almost every agency: impressions up, clicks up, cost per lead down, 214 leads generated. It is accurate, it is well presented, and it answers a question nobody asked. The owner is trying to work out whether the money spent last month produced more patients than the money spent the month before, and none of those four figures can tell them.

The gap is not dishonesty. It is that lead-level metrics are the ones an ad platform can see, and patient-level outcomes are the ones a practice cares about. Bridging the two is the entire job of a marketing dashboard, and it is where most reporting quietly gives up.

Why lead count is the wrong top-line number

A lead is somebody who expressed interest. That is all it is. It contains no information about whether they were eligible, whether they could afford the program, whether they booked, whether they arrived, or whether they came back.

Optimizing for lead volume actively degrades the other four, and predictably: broader targeting and lower-friction forms both increase leads and decrease the share of them who become patients. A campaign can halve its cost per lead and lose money, and on a lead-count dashboard that shows up as a win.

The four numbers that do tell you something

1. Cost per booked appointment

Total spend divided by appointments actually on the calendar — not leads, not calls, not form fills. This is the first number that connects marketing spend to something the practice can act on, and it is usually a shock the first time a clinic sees it, because the ratio between leads and bookings is worse than anyone assumed.

Include everything in the numerator: ad spend, agency fees, tooling. A cost per booked appointment that counts only media spend flatters the channel and makes two campaigns with different management overheads look comparable when they are not.

2. Show rate

The share of booked appointments that are attended. This is the metric that turns a booking number into a revenue number, and it is the one most often missing entirely — because it lives in the practice management system rather than the ad platform, and nobody has joined the two.

Show rate also varies by channel in ways that change budget decisions. A source producing cheap bookings with poor attendance can be more expensive per attended patient than one producing fewer, better ones. If the number is low across the board, the problem is usually not the source at all — it is the sequence between booking and arrival.

3. Patient lifetime value

What a patient is worth across their whole relationship with the practice, not on their first visit. This is what sets the ceiling on acquisition cost, and without it every conversation about whether marketing is “too expensive” is guesswork.

It does not have to be sophisticated to be useful. Average revenue per visit, multiplied by the average number of visits a patient makes, is enough to make better decisions than no figure at all. The important discipline is segmenting it: in most practices one program type is worth several times another, and a blended average hides exactly the difference budget should respond to.

4. Pipeline velocity

How long it takes an inquiry to become an attended appointment, and how that time is distributed. Velocity is the diagnostic metric of the four — it does not tell you how you are doing so much as where the problem is.

A funnel with good conversion and slow velocity is fragile: it is working despite the delay, and a competitor who answers faster will take the difference. Speeding it up is also usually the cheapest improvement available, because it requires no additional traffic.

Reading them together

Individually each of these can mislead. Together they identify the problem, because the combination that is failing tells you which part of the funnel to look at.

That bottom row is the one most likely to be true and least likely to be reported, because a slow funnel does not look broken on any single metric — see why minutes decide who books.

The reports that mislead by construction

Three reporting habits produce confident, wrong conclusions often enough to be worth naming.

• Attribution that stops at the click. If the report ends where the ad platform’s data ends, every conclusion in it is about clicks. The join to booked and attended appointments is the whole point of the exercise.
• Monthly windows on a longer sales cycle. Where the gap between inquiry and appointment runs to weeks, a calendar month compares this month’s spend against last month’s patients. The trend looks like noise because the periods do not line up.
• Averages across programs with different economics. A blended cost per patient across a high-value program and an entry-level one describes neither, and the decisions it supports are wrong in both directions.

None of this requires new software so much as one system that can see the whole path. That is the argument for keeping campaigns, conversations, bookings and reporting in one place rather than four — the growth dashboards inside the platform exist because a number that lives in a different tool from the appointment it produced cannot be joined to it afterwards.

Where to start

Pick one month that has fully closed. Count the appointments actually attended that can be traced to marketing, and divide the whole marketing cost of that period by it. That single figure — cost per attended patient — will be less precise than a proper dashboard and more useful than the report you are getting now.

Then compare it against what an average patient is worth. If the second number is comfortably larger, the question is how much more you can spend rather than whether to keep spending. If it is not, you now know the actual problem, which is more than a lead count ever tells you.

A free marketing audit does this pass on your own numbers and shows you where the funnel is leaking. If you would rather see the system first, how the five-step growth framework works sets out how measurement and optimization fit together, and the plans and what each includes are published in full.