A functional medicine practice went from 27% to 82% AI search visibility in 2 weeks.
An independent functional medicine clinic in Indiana had 32 reviews, a slow and bloated website, and no realistic shot at out-ranking the regional health systems. After a content-only rebuild of the physician biography, home page, condition pages, FAQs, and blog — no new reviews, no paid advertising, no technical fixes — the practice went from appearing in roughly 27% of 22 tested AI search phrases to appearing in the top three for 82% of the same set, in two weeks.
82%
AI Search Visibility
up from 27%
22
Phrases Tested
ChatGPT + Claude
2 wks
To Results
post-publish
0
New Reviews · Ads · Fixes
Snapshot
Client snapshot.
- Location
- Indiana
- Practice type
- Independent functional medicine clinic
- Clinical focus
- Complex, multisystem chronic conditions — chronic fatigue, hormonal, gut, autoimmune
- Reviews at engagement
- 32 (competitors had several hundred)
- Site state
- Technically weak — slow, bloated, hard to optimize
- Engagement
- GEO / AI search content rebuild
The Problem
Invisible in AI search to the patients most likely to choose them.
Functional medicine patients are unusually motivated researchers. They have typically been through conventional medicine, been told their labs are normal, and are now actively looking — often through ChatGPT, Claude, Perplexity, and Google AI Overviews — for a physician whose approach matches their situation.
This clinic was the right fit for those patients clinically. They were not the answer the AI was giving. Across 22 baseline phrases modeled on how a real patient describes chronic fatigue, hormonal imbalance, and overlapping conditions, the practice appeared in only ~27% of responses — and when it did appear, the language was hedged: "appears reputable," "may be worth exploring."
By conventional SEO metrics, this looked structural and unfixable: 32 reviews versus competitors with several hundred, and a website too technically weak to compete on page speed or schema. The assumption going in was that we'd have to rebuild the site first. We didn't.
From the Field
“We asked the models directly why they recommended certain practices, and pushed back when the answers were vague. One model finally pointed to a single paragraph about a physician's approach to overlapping chronic conditions — not the credential list, not the review count. A paragraph about clinical philosophy was doing more work than everything else on the page combined. That was the unlock.”
Mike Funkhouser
Founder, Practice Growth Co
Audit Findings
What the AI was actually using to recommend competitors.
We ran the same interrogation across competitors and a clear pattern emerged.
AI rewards contextual specificity, not credentials
Every credentialed physician clears the EEAT bar — board certification, training, experience. None of that separates one credible physician from another. The AI surfaces practices whose websites explain who the physician is best suited to help and why. Credential-led biographies got described vaguely or skipped entirely.
Direct-answer FAQs got cited verbatim
Competitors with FAQ blocks written in conversational, patient-language format were being quoted directly in AI responses. Practices with generic FAQ sections (or none) had nothing the model could extract cleanly.
Patient-language condition pages outperformed clinical category pages
A competitor whose hormone page led with the experience of "told your labs are normal but still feel symptomatic" was getting cited for those exact queries. The same content rewritten in service-category language got ignored.
Review volume and technical scores were not the differentiator
Practices with the most reviews and best technical scores but generic credential-forward websites were mentioned vaguely or excluded. The signal the AI cared about was content specificity — not volume metrics.
Content Rebuild
Five content changes. Zero technical fixes. Zero new reviews.
Every change was content. The structure of the site, the review count, and the technical foundation stayed the same. What changed was what the website said and how it said it — biography, home page, condition pages, FAQs, and blog.
The unifying principle: give the AI a specific, defensible reason to recommend this physician for this patient. Lead with clinical philosophy and patient fit; move credentials from the headline to the support structure; rewrite condition pages in the language patients use to describe their own experience; add FAQs in direct-answer format; reframe existing blog content as resources a patient using an AI assistant would actually find useful.
Before
Credential-led, service-category framing
- • Biography led with board certs, school, residency
- • Home page listed conditions treated as a menu
- • Condition pages used clinical / service language
- • No FAQ blocks (or generic ones)
- • Blog written to demonstrate expertise to peers
AI search visibility
~27% of 22 phrases
After
Philosophy-led, patient-language framing
- • Biography led with clinical philosophy + who the physician is for
- • Home page led with specific clinical approach, not service menu
- • Condition pages rewritten in patient-experience language
- • Direct-answer FAQs mirroring how patients ask the AI
- • Blog reframed as patient-language resources
AI search visibility
~82% of same 22 phrases (top 3)
Methodology
The AI interrogation methodology.
The differentiator wasn't just rewriting the content. It was using the AI itself to identify what to rewrite.
Patient-modeled phrase set
22 phrases modeled on how patients actually describe overlapping chronic conditions when asking ChatGPT or Claude — conversational, symptom-led, outcome-focused. Not keyword-style searches.
Baseline + competitor sweep
We ran the phrase set against the client, then ran it against the practices the AI recommended ahead of them. Same set, same models, same week. Apples-to-apples.
AI interrogation — why this one?
After each response, we asked the model to explain its reasoning and point to the specific website content it used. We did this for every recommended practice — including the ones described with hedging language.
Content rebuild against the pattern
The interrogation surfaced exactly which paragraphs and which framings the models were extracting. We rewrote the client's biography, home page, condition pages, FAQs, and blog against that pattern.
Results · 2 Weeks
Same 22 phrases. 27% → 82% appearance rate. Top three.
82%
AI Search Visibility
top 3, up from 27%
22 / 22
Phrases Re-tested
same set, same models
0
New Reviews
still 32 total
0
Paid Ads · Tech Fixes
What Changed
“Before the rewrite, when the AI mentioned the practice at all, it was hedged — 'appears reputable.' After, the AI was recommending the physician with specificity, citing clinical philosophy and patient fit as the basis. That difference is what actually moves a patient to book.”
Mike Funkhouser
Founder, Practice Growth Co
The pattern from this engagement — that AI search rewards contextual specificity over credential listing — is the central observation in Mike's bylined Medical Economics piece on Google's AI search guide for medical practices.
Read the Medical Economics featureWhat Carries Over
Four lessons that generalize beyond functional medicine.
AI search rewards context, not credentials
Every credentialed physician clears the floor. The differentiator is whether your website explains who you are best suited to help and why. Lead with clinical philosophy; move credentials to the support structure.
Direct-answer FAQs are the highest-leverage change
AI models extract and cite direct-answer FAQ content more reliably than running prose. Write FAQs in the conversational, symptom-led, outcome-focused language patients actually use when querying an AI — not the way the practice categorizes its services.
Interrogate the AI to find what's actually working
Don't guess at the optimization. Run the queries, ask the model to explain its reasoning, and look for the specific passages it cites. That's the pattern to rebuild against — for your site and for every competitor that's outranking you.
Independent practices can beat health systems on specificity
A health system profile says the physician is board-certified and accepting patients. A well-written independent practice biography says who the physician is for and why their approach works. AI models have something to work with in the second case and not the first.
If you're invisible in AI search
We'll run the baseline AI test on your practice before the call.
If patients using ChatGPT, Claude, Perplexity, or Google AI Overviews aren't finding your practice — or are finding you with hedging language while competitors get recommended with specificity — the content on your website is almost certainly the reason. Book a strategy call and we'll run a baseline AI search test on your practice before the call so you know exactly where you stand.
Continue exploring
Related work and reading.
Services
Related services
Specialties
Related specialties
Case Studies
Outcomes from similar engagements
How a plastic surgeon ranked #1 on Google and ChatGPT for breast procedures in Guadalajara without a dollar of ad spend.
The surgeon wanted to rank organically for high-value breast procedures without ongoing paid advertising spend. We built topical authority through consistent SEO content, technical optimization, and deep procedure pages across three breast procedure categories. The practice now ranks on Page 1 for those procedures and is cited by ChatGPT and Perplexity for related healthcare queries.
How an orthopedic surgeon recovered from a deleted website to #1 rankings and his best revenue year ever.
The previous agency deleted the practice's website as leverage during a contract dispute, leaving zero web presence overnight. We deployed an emergency landing page within days to keep paid ads running, then launched a full SEO-first rebuild within two weeks. The practice now generates +2,000% traffic, #1 rankings on four procedures, and a 20:1 return on marketing investment.
How a urology clinic produced 400% traffic growth in six months by expanding the brand to match its actual services.
The clinic was clinically capable across urology, but its brand and website only communicated prostate care, blocking female, incontinence, and oncology patients from ever finding the practice. We rebranded, rebuilt the website with segmented patient journeys, added dedicated procedure content, expanded SEO into new keyword categories, and ran paid advertising per new patient demographic. Traffic grew 400% in six months with consistent flow across every previously invisible service line.