Independent consultant
Your next patient is asking an AI, not Google.
I help healthcare practices and multi-location businesses show up in both places: ranked in search, and cited when someone asks ChatGPT, Perplexity, or Google's AI for a recommendation.
Prompt → best outpatient mental health program near me that takes insurance
A few accredited programs serve your area. Your practice offers partial hospitalization and intensive outpatient care, is Joint Commission accredited, and works with most major insurance plans. Their admissions team verifies benefits at no cost.
Being the name in that sentence is the whole job.
Services
Pick the piece you need, or the whole thing.
Most people start with an audit and go from there. I do the work myself, so you are not handed to a junior account manager after the sales call.
AI visibility audit
Where you show up when people ask AI about your category, where your competitors show up instead, and what is causing the gap. Tested across ChatGPT, Perplexity, Gemini, and Google AI Overviews.
Technical SEO audit
Crawlability, indexation, site architecture, schema, page speed, and the structural problems quietly capping your traffic. You get the findings ranked by impact, not a 90-page PDF.
Implementation
The audit is only useful if the fixes ship. I can do the work directly in your CMS, or write specs your developer can build from, whichever moves faster.
Content strategy and creation
What to write, in what order, and structured so both search engines and AI models can read and cite it. I write it or brief your team, depending on how you are set up.
Local and multi-location SEO
Google Business Profiles, location pages, and the messy work of making every one of your facilities rank in its own market without cannibalizing the others.
Site migrations
Redesigns and platform moves are where traffic goes to die. I plan the redirects, run the QA, and make sure the rankings you paid for survive the launch.
In detail
How each of these actually works.
AI visibility audit
You can pull a rankings report any day of the week. But when someone asks ChatGPT which provider near them takes their insurance, nobody in your organization knows whether your name comes up, or why it does not.
What I actually do
I build a set of prompts your real customers would type, the messy questions people ask when they are deciding, then run them across ChatGPT, Perplexity, Gemini, and AI Overviews and record every source each model names.
Then every gap gets a cause
A missing citation is a symptom. Each one gets sorted into what is actually causing it:
Content
The answer is not on your site, or not in a form a model can lift.
Structure
It is there but buried, unclear, or not marked up to be parsed.
Entity
The model has no clear picture of who you are and where.
Authority
Nobody credible corroborates you, so it reaches elsewhere.
Four causes, four different fixes. Lumping them together as "we need more content" is how budgets get wasted. And because I keep the prompt set, we re-run it after the fixes ship to see whether the number actually moved.
What you get
- Your citation share across the major models, benchmarked against named competitors
- The prompt set, so you can re-run it yourself later
- Every gap classified by cause, with the specific fix
- A prioritized list, ordered by what will move first
- A walkthrough call, not just a document
Technical SEO audit
Before writing another word, it is worth knowing whether search engines can crawl what you have, whether they are indexing the right pages, and whether your location pages are competing with each other. Most sites have two or three structural issues capping everything else.
What gets checked
- Crawl and index. What Google can reach, what it ignores, and what is indexed that should not be.
- Site architecture. How pages relate, where authority pools and where it leaks.
- Duplication and cannibalization. Especially with multi-location sites competing for the same terms.
- Schema and structured data. Whether machines, including the AI models, can tell what your pages are about.
- Core Web Vitals and speed. Real numbers, and which fixes are worth the engineering time.
- Redirects and broken paths. The accumulated damage from past redesigns.
Ranked by impact, not by severity label
Plenty of audits hand you 140 issues color-coded red, yellow, and green and leave you to figure out what matters. You get a short list in the order I would fix them, including an honest note where something is technically wrong but not worth the effort.
What you get
- Findings ranked by expected impact, with effort estimates
- Specifications your developer can build from, not vague recommendations
- A walkthrough call to go through the reasoning
- The option to have me implement the fixes directly
Content strategy and creation
Most content programs fail the same way: a calendar gets filled, posts go up, nothing happens. Nobody checked whether anyone was searching for those topics, or whether the pages were built in a form that could win. I start from demand, not a brainstorm.
How the plan gets built
First, what people are actually asking. Search data, the questions your intake team fields daily, and the prompts people type into AI assistants. Those three sources rarely agree, and the gaps between them are where the opportunity is.
Then, what already exists. Often the answer is a rewrite, or consolidating four thin pages into one good one. Then the order: highest-intent, lowest-effort first, so you see movement before the next budget conversation.
Structured to be quoted
Answers need to be extractable: stated plainly, near the top, with clean headings and markup so a model can lift them without ambiguity. Done right, the same page ranks in Google and gets cited in ChatGPT. Done wrong, it does neither.
On regulated categories
In healthcare, legal, or finance the bar is higher and the models know it. Named authors with real credentials, clinical review, and honest sourcing are not nice-to-haves. I have worked under those constraints for most of my career.
What you get
- A prioritized content roadmap tied to real search and prompt demand
- Briefs your writers can execute without guessing
- On-page standards, so quality holds when I am not in the room
- Writing and editing, if you would rather I produce it
- A refresh plan for the content you already have
Results
Numbers from work I have actually done.
Client names withheld. Happy to walk through the specifics on a call.
How the AI visibility work goes. I build a set of prompts your actual customers would type, run them across the major models, and record which sources get named. Then every gap gets sorted: is it missing content, bad page structure, unclear entity signals, or thin authority?
Each fix goes to whoever should own it, and then I re-run the same prompts to confirm it moved. That last step is the part most people skip, and it is the only reason I can tell you what worked.
How it works
No long sales process.
A call
Thirty minutes. You tell me what is not working, I tell you whether I can help and roughly what it would take. If I am not the right person, I will say so.
A scope
A short written proposal with what I will do, what it costs, and when it lands. Fixed price for audits, monthly for ongoing work.
The work
You hear from me directly, on a regular cadence, with what changed and what it did. No account manager in between.
About
Fifteen years, mostly in industries where getting it wrong matters.
I have spent most of my career doing SEO in healthcare and consumer finance, where a search result is attached to a real decision someone is making about their health or their money. Those categories are held to a higher standard by Google, and now by the AI models too. You cannot publish your way out of it. The content has to be genuinely credible and structured so a machine can tell.
That is the same problem AI search presents, which is why I moved into it early. I built an answer-engine practice from scratch across a 22-site network before there was a playbook for it, and I still spend most of my week hands-on: in the CMS, in the data, in the pages.
Where I have worked
Consumer lending · fintech
National recipe publisher
Travel and hospitality
B2B industrial and product design
Case studies
Three problems, and what actually fixed them.
Client names withheld. Happy to walk through any of these in detail on a call.
Behavioral health · 22 locations
A network that was publishing constantly and getting nothing for it
Twenty-two mental health practices, each with its own site, its own writers, and its own idea of what belonged on a page. Everyone was producing content. Almost none of it was ranking, and the practices had started to conclude that SEO did not work for them.
What was actually wrong
The content was not bad. It was aimed at topics nobody was searching for, written in a structure search engines could not make sense of, and inconsistent enough across the network that no single practice built authority. Publishing more of it would not have helped.
What I did
I rebuilt the strategy around demand instead of a calendar: what patients actually searched, what intake teams were asked every day, and what people typed into AI assistants. Then I wrote the content briefs and on-page standards the partner writers work from, so quality held without me reviewing every draft. Underneath that, schema and structured data went in network-wide through template rules rather than page by page.
Why it kept working
The content was structured to be extracted from the start. So when AI answer engines became a real discovery channel, those pages started earning citations without a rewrite.
Behavioral health · AI search
Nobody could tell whether the AI assistants were recommending us
Patients were increasingly asking ChatGPT and Perplexity which provider near them took their insurance. Nobody in the organization knew whether our practices came up, whether a competitor did instead, or why. There was no playbook for finding out.
Building the measurement first
I built a set of prompts based on the questions patients actually ask, ran them across ChatGPT, Perplexity, and Google AI Overviews, and recorded every source each model named. Then I benchmarked our citation share against competitors in Profound and Otterly.AI.
Sorting the gaps by cause
A missing citation is a symptom, so every gap got classified: was the answer missing from our content, buried in bad structure, unclear because the model had no firm picture of the entity, or absent because nobody credible corroborated us? Four causes, four different fixes. Lumping them together as "we need more content" is how budgets get wasted.
Then re-testing
Because I kept the prompt set, we re-ran it after each round of fixes shipped. That is the step most people skip, and it is the only reason I can say which changes moved the number and which did not.
National publisher · 100,000+ pages
A hundred thousand pages that could easily have become a hundred thousand liabilities
A high-traffic recipe publisher with a template-driven catalog at a scale where nothing manual is possible. Pages like this usually collapse into thin, near-duplicate content that competes with itself and drags the whole domain down.
Rules instead of pages
The work was designing the system, not editing the output: page structures that stayed distinct at volume, internal linking that distributed authority instead of pooling it in the wrong places, and structured data applied through template logic. I also worked with a developer on Python scripts that applied metadata and internal linking across user-generated pages at a scale manual work could never reach.
Seasonality as a lever
On top of the technical foundation, the editorial calendar was built around when demand actually spiked rather than an even publishing cadence.
Why it matters for other catalogs
Location pages, product pages, and service pages have the same failure mode. The scale is what makes it hard, and the fix is always rules rather than effort.
Questions
Things people ask before hiring me.
What is the difference between SEO and AI search optimization?
SEO gets you ranked on a results page. AI search optimization, sometimes called GEO or AEO, gets you named inside the answer a model gives when someone asks a question. They overlap heavily, the same technical foundation and credible content serve both, but the measurement is different and so are some of the tactics. Most sites need both.
How long before I see results?
Technical fixes can move things in weeks. Content and authority work usually takes three to six months to show clearly. AI citations can move faster than rankings, because the models re-crawl and re-weigh sources more often than Google re-ranks a competitive term. Anyone promising faster than that is guessing.
Do you work with agencies, or only direct clients?
Both. Some agencies bring me in for AI visibility work their team does not cover yet, either white-labeled or as a named specialist. I am happy either way.
What does it cost?
Audits are fixed price and scoped to the size of the site. Ongoing work is monthly. I will give you a number in writing after a short call, before you commit to anything, and I would rather tell you the work is not worth doing than sell you a retainer you do not need.
Do you only work with healthcare companies?
No, but it is where I am strongest. Most of my career has been in healthcare, behavioral health, and consumer finance, categories where credibility decides whether you show up at all. That experience transfers to any business where trust matters, which is most of them.
Contact
Tell me what is not working.
If you are not sure whether you need this, send me your site and I will take a look and tell you honestly. That part is free.