From zero AI presence to cited in 60 days.
A newly opened behavioral health practice in San Diego had experienced clinicians and accepted major insurance plans. But AI had never heard of it. When people asked for help finding care, the answers named practices that had been open for years.
Practice
Behavioral health, CA
Category
Healthcare
Engagement
AI Audit + AI Starter
Duration
60 days
Representative AI Audit + AI Starter engagement. Figures are illustrative.
In this case file
01The before
A new practice AI had never heard of.
The practice opened with licensed therapists, a psychiatrist, and in-network status with several major insurers. It had the clinical team in place. What it didn't have was a history that AI could find.
That mattered because of how people now look for care. Many start privately, by asking an AI assistant a question they might not yet say out loud to anyone else.
When people looked for care, they asked questions like:
- Who are the best therapists for anxiety in San Diego?
- Which San Diego therapy practices accept insurance?
- What's the difference between intensive outpatient and regular therapy?
- How do I find a psychiatrist accepting new patients in San Diego?
- Which behavioral health clinics near me offer evening appointments?
These questions come at a moment when someone has decided to seek help. The practices AI names are often the only ones they contact.
We measured the practice across a fixed baseline of 20 high-intent prompts on six AI platforms. The practice was named in none of them. Its two nearest competitors scored 28 and 19 out of 100. The practice scored 0.
0 / 20
Practice named
11 / 20
Top competitor named
ChatGPT
0
Google AI
0
Gemini
0
Copilot
0
Perplexity
0
Claude
0
The problem was not quality of care. It was that no evidence existed yet.
The AI Presence Audit, delivered in seven days, showed exactly what was missing:
- The website had no structured data describing the practice, its services or its clinicians.
- The practice's name and address differed across four directories.
- No independent health publication had mentioned the practice.
- Services were described in clinical terms, not in the words people use when they ask for help.
- Insurance information was hidden in a downloadable PDF that AI systems rarely read.
The practice was ready for patients. It was not yet ready to be found by AI.
02The work
We built the evidence a new practice doesn't have yet.
A new practice can't rely on years of reviews and coverage. So the engagement focused on building clear, trustworthy evidence quickly, and doing it carefully, because healthcare claims are held to a higher standard.
We started with the audit findings and fixed the highest-priority gaps first.
1. We built a prompt map around how people seek care.
The 20-prompt map covered four stages:
- 1.Problem-aware prompts
Questions about anxiety, depression and stress, and when to seek help. - 2.Care option prompts
Questions comparing therapy, psychiatry and intensive outpatient programs. - 3.Provider search prompts
Questions asking for therapists and clinics in San Diego and nearby neighborhoods. - 4.Access prompts
Questions about insurance, availability and evening appointments.
Each prompt was checked across all six AI platforms every two weeks.
2. We clarified how the practice is described.
We wrote one description for every source to repeat: an in-network behavioral health practice in San Diego offering therapy and psychiatry for adults, with evening availability.
- Structured data added for the practice, its services and its clinicians.
- Plain-language service pages that answer the questions people actually ask.
- Insurance plans listed directly on the website, not in a PDF.
Every description avoided outcome promises and clinical claims. The goal was accuracy, not persuasion.
Prompt
“Which San Diego therapy practices accept insurance?”
Answer
Several San Diego practices accept major insurance plans, including Client, which offers therapy and psychiatry for adults with evening appointments. Check your plan's provider directory to confirm coverage before booking.
3. We earned placements in trusted health sources.
For a new practice, independent mentions are the fastest way to become credible to AI. The monthly work focused on earning them.
The work included:
- Five editorial placements every month on health and wellness publications.
- Each placement answering a real question, such as how to choose between therapy and psychiatry.
- Every piece reviewed for clinical accuracy, with no outcome claims.
- Every placement tracked and reported, with its live status and URL.
4. We corrected the third-party sources.
Four directories listed the practice with different names or an old suite address. Two showed it as not accepting new patients. We corrected all four.
That work focused on:
- Therapist and psychiatry directories.
- Insurance provider directories.
- Google Business Profile.
- Local health and business listings.
5. We used measurement to decide what to improve next.
Every two weeks, we re-ran the prompt map.
The first citation appeared on day 41, in Google AI Overviews. From there, we focused placements on the access and insurance prompts, where the practice had the clearest advantage.
No paid ads were added during the engagement. The work focused only on making the practice easier for AI to understand, verify and recommend.
03The result
From unknown to recommended in 60 days.
6 / 20
Prompts naming the practice
ChatGPT
0 →2
Google AI
0 →2
Gemini
0 →0
Copilot
0 →1
Perplexity
0 →1
Claude
0 →0
Visibility score
0 →31
Answer accuracy
— →78
At day 60, we re-ran the same 20 prompts across the same six AI platforms.
The practice was named in 6 of 20 prompts, up from 0 at baseline.
The broader movement mattered just as much:
- AI visibility score: increased from 0 to 31 out of 100, ahead of one of its two nearest competitors.
- Answer accuracy: 78 out of 100. AI now described the practice's services and insurance correctly.
- Directory errors: all four corrected.
- First citation: day 41, in Google AI Overviews.
AI answers began describing the practice the way it describes itself: in-network, adult-focused, with evening availability.
For insurance and access prompts, the practice moved from unknown to one of the first names recommended.
A new practice can't buy a history. But it can make sure the evidence it does have is clear, consistent and easy for AI to trust.
Engagement summary · AI Audit + AI Starter
A seven-day audit to find the gaps, then five placements a month to fill them, with every claim checked for accuracy.
Package
AI Audit + AI Starter
Investment
$499, then $1,099 / month
Duration
60 days