The proof
What we refuse to promise
Medicine runs on evidence. So do we. A studio cannot control what Google shows or what patients decide, so we will not sell you a number we cannot stand behind. We promise the system, the work, and the measurement.
We will not promise
- Traffic.
- Rankings.
- Revenue.
Each one is refused for a reason, and each one has something we measure in its place.
The refusal set
No number we cannot stand behind.
Three things every agency offers to promise. Here is why we will not, and what we put on the report instead.
Traffic.
Search demand belongs to patients and the season, not to a studio. We build every surface patients check, then count what actually arrives.
Measured instead
Inquiries, clicks, and bookings, tracked to their source
Rankings.
Google takes no instructions from agencies. A guaranteed position is a guess with your name on it.
Measured instead
Where you appear, checked live on the call
Revenue.
Revenue is made in the exam room by medicine, payers, and capacity. We claim the pipeline, not your practice.
Measured instead
Cost per booked appointment, in plain arithmetic
PulseTrack methodology
Every number carries its method
- Metric
- Booked appointments
- Cohort
- New patients
- Timeframe
- Last 30 days
The headline number
Booked appointments lead the page, with the cohort and timeframe printed beside them. Never a raw traffic curve pretending to be an outcome.
Where bookings came from
Every appointment request and booking click is tied to the surface it arrived from: search, maps, ads, or direct. You see which work is working.
Cost in plain arithmetic
Spend sits next to what it produced, and the division is printed so you can check it: what was spent, what it booked, what each booking cost.
The work and the plan
What shipped this month and what happens next, in plain language. The report is a record of work, not a mood board of charts.
Build standards
Standards every build ships with
- 3facts on every number
Metric, cohort, timeframe. A number missing one does not get printed.
- 5standards, each checkable
Every row below tells you how to check it yourself.
- 0numbers we invent
No composite figures, no anonymous quotes, no borrowed case studies.
- AAWCAG level we build to
Tested on every page, in both themes, before launch.
- Accessibility to WCAG AA
- What every build ships withVerified with automated testing and a full keyboard pass on every page, in both themes. Contrast is measured, never eyeballed.
- How you check itTab through this page. Every control shows a visible focus ring and nothing traps you.
- Speed, measured fairly
- What every build ships withWe time the old site and the new site the same morning, on the same throttled connection and device profile, and report both numbers with the method attached.
- How you check itAsk on the discovery call. We run the timing in front of you.
- The stack, named plainly
- What every build ships withNext.js 14 and TypeScript. Modern, maintained, and yours: you own the repository, the domain, and every design asset.
- How you check itView the page source. The stack is visible to anyone who looks.
- Structured data on every page
- What every build ships withEach page carries a machine-readable summary that tells search engines and AI assistants what it is. The technical name is schema; the point is being legible to machines.
- How you check itPaste any page of this site into a public schema validator.
- The studio site is the exhibit
- What every build ships withWe hold TallBunnyMD.com to the same public bar we sell: Lighthouse 100 on Accessibility, Best Practices, SEO, and Agentic Browsing, and Performance 100 on desktop and on mobile with DevTools throttling. Zero axe-core violations at WCAG 2.1 AA. Lab captures, with the method named.
- How you check itOpen Chrome DevTools on this page, run Lighthouse 13. Desktop, or mobile with throttling set to DevTools, is the method. Compare the five categories.
Live verification
We measure your practice in front of you
Nothing on the discovery call is prerecorded. We run the numbers on your current presence while you watch, and you keep them whether or not you hire us.
30 minutes. Plain language. A roadmap, not a pitch.
Measured live, on screen
- Your current site, timed on a throttled mobile connection.
- Your Google Business Profile, exactly as patients see it.
- Your local search results for the care you provide.
- What AI assistants answer when asked who to see nearby.
Patient data
What we can promise, and what we cannot
What we engineer
The parts we control
Attribution
Attribution is designed to work without conditions, diagnoses, or medical records. We measure where an inquiry came from and whether it became an appointment. Nothing else is needed for that, so nothing else is collected.
Calls
PulseTrack records where an inquiry came from and whether it booked. It never touches your phones, and there is no field anywhere in it for a patient's name.
Advertising
We do not build advertising audiences from people who read about a condition, and we do not upload patient lists to advertising platforms. Campaigns target the care, not the reader.
Placement
We do not place advertising or analytics code inside patient portals or logged-in scheduling. Marketing code stays on marketing pages.
In writing
Every tracking script on your site is inventoried in writing, with what it sends and where it goes, so your privacy officer does not have to take our word for anything.
Vendors
Where a company would handle information about your patients on your behalf, it signs a business associate agreement first, or we do not use it.
What we do not claim
The limits, said out loud
No certification
There is no HIPAA certification for software or for marketing agencies. No such program exists, so we do not claim one, and neither should anyone selling you a badge.
Not compliance
We cannot make your practice HIPAA compliant, and no vendor can. Compliance belongs to your whole operation. What we can do is build the parts we control so they do not put you at risk, and document them so you can check.
Not attorneys
We are not attorneys and none of this is legal advice. Your counsel or privacy officer should review anything that touches patient data, and we give them everything they need to do it properly.
No outcomes
We do not promise rankings, traffic, or patient volume. We commit to inputs and terms. Outcomes are measured, not promised.
Attribution needs a source and an outcome. It does not need a person, so we build it without one.
There is nothing sensitive in it to protect, which is the whole point of designing it that way.
Case studies
No case studies yet.
Client case studies publish here as practices approve them, with the practice named and the numbers pulled from PulseTrack, not from memory. We do not publish invented results, anonymous quotes, or composite figures. Until a practice says yes, this space stays empty, and empty is the honest state.
An empty shelf is a fact you can check. A borrowed one is not.
Reserved for the first approved case study
- Practice
- Not published yet
- Metric
- Not published yet
- Cohort
- Not published yet
- Timeframe
- Not published yet
- Method
- Not published yet
One practice, one metric, one cohort, one timeframe, the method linked.
What we find
Fifteen practice websites, opened one at a time
Fifteen practice websites in a single South Carolina metro area, audited systematically in July 2026, read first the way a patient reads a site and then the way a search engine or an assistant reads it.
5
had no working way to book online
of fifteen
A main Request an Appointment button leading to a page that no longer exists. A booking page whose entire content was a placeholder sentence, listed in the site's own map as a real page. A contact page instructing patients to use a form that was not on it. One site had no form of any kind anywhere, including its contact page, leaving a phone line and a consumer email address as the only ways in.
Quoted verbatim from a live practice website: “This is just a placeholder”
Check your ownOpen your own site on a phone and try to book an appointment at nine on a Sunday night.
7
published hours that were missing or disagreed
of fifteen
One practice published no office hours anywhere: not the footer, not the contact page, not the machine copy. Another published three different sets, two of them on the same page, thirty minutes apart. Several told search engines one set of hours while printing another for patients. One sent search engines a closing time that is not a time.
Quoted verbatim from a live practice website: “12:3012:30”
Check your ownSearch your practice name and compare the hours Google shows to the hours in your own footer.
9
were unreadable as a medical practice
of fifteen
No machine-readable record that the practice is a medical practice with an address, opening hours and named doctors. Two had none at all. One had a single generic block that would suit any business on any street. In a group of five board-certified dermatologists, not one of the five existed as a record a search engine or an assistant could read.
Check your ownPaste your homepage into a public structured data validator and see whether it knows you are a medical practice.
5
showed no patient review anywhere
of fifteen
Not a quote, not a rating, not a star. Several of those practices have public reputations running to hundreds of reviews. Elsewhere in the batch, one site declared a star rating and a review count in its code that had not moved since around 2021, and one had a page called Testimonials containing no testimonials, only a poster asking for one.
Check your ownSearch your own practice name on your phone and compare the rating you see to what your patients have been saying lately.
5
had recorded no visitor data since 2023
of fifteen
The only measurement tag on the site was Google's retired Universal Analytics, which stopped processing data on 1 July 2023. Three years of nothing. One of those practices was buying Google ads against it. Another was still loading the tracking script of a previous vendor, pointed at a server that no longer exists.
Check your ownAsk whoever runs your site when your analytics last recorded a visit, and ask them to show you.
The exception, which matters more than the list
One site in the fifteen was a genuinely good build: actively maintained, recently refreshed, with machine-readable markup better than most of what we see. Its problem was the one nobody can see by looking. Measured on a phone connection, the main image took over sixteen seconds to appear. It looks perfect on the office desktop. That is the argument for measuring rather than judging, and it is why an opinion about a website is a hypothesis until something times it.
None of the above is a matter of taste. They are a dead link, a wrong number, a fact stored in two places, an hour nobody checked. All of them are cheap to fix, all of them are checkable by the practice itself in an afternoon, and every one of them sits directly on the path a patient takes to reach a doctor who was recommended to them.
Next step
See what your practice’s system would look like
A 30-minute call in plain language. You leave with a roadmap for your practice, whether you hire us or not.
Or write to hello@tallbunnymd.com