Therapist Marketing · Clarity Library

How to Get Recommended by AI as a Therapist

A working checklist for showing up when ChatGPT, Gemini, or Perplexity gets asked to recommend a therapist, plus exactly how to test whether you already do.

Why this matters How AI decides The checklist Test yourself What matters less FAQ

Why this matters

A growing share of the people who need a therapist are getting an answer before they even see a typical search result link. Some of that happens right inside Google, where an AI Overview now sits above the regular results and can name two or three therapists without anyone clicking through. Some of it happens because people skip Google entirely and ask ChatGPT, Gemini, or Perplexity the question directly. Either way, if your practice isn't one of the names that answer brings up or recommends, you're invisible to that search, no matter how good your website looks to a person who is fortunate enough to discover it.

This is the same territory SEO has always covered, just with a different mechanism for who gets picked. Google used to pull ten search results (links) and let the person decide. An AI answer hands someone two or three names and narrows the decision for them. Earning one of those names takes a specific kind of groundwork, and most of it hasn't been done yet by most practices. If you have been wondering how to get recommended by AI as a therapist, we aim to simplify each step and highlight what's important.

How AI tools actually decide who to recommend

AI tools don't browse the internet fresh every time someone asks a question. They pull from sources they can verify quickly and quote cleanly. In practice, that comes down to four things working together.

Consistent identity. Your website, Google Business Profile, and every directory listing need to agree on your name, address, phone number, and specialties. Mismatches read as unreliable data and get skipped.
Extractable content. A page that opens with a clear, direct answer is far easier for an AI system to lift and quote than a page that builds up to the point over three paragraphs.
Structured data. Schema markup spells out, in code, who you are, what you treat, and where you practice, so nothing gets left to inference.
Outside corroboration. If your own website is the only place describing what you do, there's nothing for an AI tool to check it against. Directory listings, associations, and other mentions all function as evidence that what your site says is real.

None of these four stand in for the others. A beautifully structured site with no outside corroboration and a perfectly consistent listing with no answerable content will both struggle to get cited.

The checklist

Eleven items, organized into three groups. Tap a group to switch, then tap any item to open it.

Your practice name, address, and phone number need to be identical across your website, Google Business Profile, and every directory you're listed on (Psychology Today, GoodTherapy, TherapyDen, Inclusive Therapists, and any local directories). Even a small difference, like "St." on one listing and "Street" on another, reads as inconsistent data. Go through every listing you have and make them match exactly, down to how you format a suite number.

Category, every specialty, hours, service area, and a real description in your own words. Google discontinued the Q&A feature on Business Profiles in November 2025 and replaced it with AI-generated answers pulled from your profile data. An incomplete profile now gets an inaccurate AI-written answer standing in for it, so fill in every field you have access to.

Add JSON-LD MedicalBusiness schema (name, address, phone, hours, credentials) plus a Person entry for yourself with your registration or license details. If you have a genuine FAQ section, mark it up with FAQPage schema too (see Google's own FAQ schema documentation), since it's one of the formats AI tools pull from most reliably.

A basic starting example for a practice page:

{
 "@context": "https://schema.org",
 "@type": "MedicalBusiness",
 "name": "Your Practice Name",
 "address": {
 "@type": "PostalAddress",
 "streetAddress": "123 Main St",
 "addressLocality": "Your City",
 "addressRegion": "ON",
 "postalCode": "A1A 1A1"
 },
 "telephone": "+1-000-000-0000",
 "medicalSpecialty": "Psychotherapy"
}

This is technical enough that it's worth a developer or your website platform's built-in schema tools if you're not comfortable editing code directly.

Beyond your own website, list your practice on the directories your ideal clients actually use. Psychology Today, TherapyDen, GoodTherapy, and Inclusive Therapists are a reasonable starting set. Link your social profiles and directory listings back to each other wherever the platform allows it. Each additional accurate, consistent mention is one more thing an AI tool can use to confirm you're real and current.

A single "Services" page listing six specialties in one paragraph gives an AI tool nothing specific to quote. A dedicated page for each specialty, with its own URL, gives it something to point to directly when someone asks about that one thing specifically.

Before nuance or approach, answer the actual question in the first sentence or two. A page about anxiety therapy should say plainly, near the top, that you treat anxiety and who it's for, not open with a story or a general statement about mental health. AI tools tend to lift the clearest, most self-contained statement on a page, so make sure that statement exists and says what you'd want quoted.

Not three vague questions for the sake of having a section. Actual questions clients ask before booking: do you take insurance, what's a first session like, do you work with teens, what's your cancellation policy. Answer each in two or three sentences, then mark the section up with FAQPage schema (see Foundation, item 3).

Stale pages lose ground gradually and without much warning. Set a recurring reminder to review your specialty and FAQ pages every few months, confirm the details still hold, and add anything you've learned from questions clients have actually asked since you last touched the page.

Your registration or license, the body you're regulated by, and your actual qualifications should be easy to find, ideally on their own page or clearly stated in your bio. This is one of the clearest trust signals a regulated practice has, and AI tools weigh credibility markers heavily on health-related topics.

Real, unsolicited reviews on Google or a directory contribute to how trustworthy AI systems treat your listing. Many regulatory colleges restrict or outright prohibit actively soliciting reviews or testimonials, so check your own body's advertising standards before pursuing this one. Nothing here is worth risking your registration over.

A guest post, a podcast appearance, a quote in a local news piece, a listing in a professional association directory. Anything that mentions your practice on a site you don't control adds corroboration your own website can't provide for itself.

Not sure where you actually stand?

This is exactly what an AI visibility strategy is built to answer. See how we think about it, or reach out directly to talk through what it looks like for your practice.

How to test whether you're already showing up

The checklist above is the groundwork. This part is how you find out whether it's working, and it takes about ten minutes a month.

Which tools to test

Test ChatGPT, Google's AI Overview (the AI summary that now appears above regular results for a lot of searches), Gemini, and Perplexity if you have access. They pull from different sources and won't always agree with each other, so checking only one gives you an incomplete picture.

What to actually search

Skip your own name. Search the way a prospective client would actually ask, in full sentences if that's how you'd naturally phrase it:

"Who's a good therapist for anxiety in [your city]?"
"Trauma therapist near [your city] taking new clients"
"Do any therapists in [your city] specialize in eating disorders?"

Run the same handful of queries every time so you can compare results month over month. Use an incognito or logged-out window where you can, since a signed-in session can personalize results in ways a stranger wouldn't actually see. Run each query more than once if the tool allows it. AI answers aren't fixed the way a search ranking is; the same question can return a different answer on a different day, so one good result isn't a guarantee and one miss isn't a verdict.

What to look for

Are you named at all?
Is what's said about you accurate (address, specialty, whether you're taking new clients)?
If you're not named, who is?

What to do with what you find

Not mentioned

This usually traces back to the Foundation tab above. Start with NAP consistency and directory listings, since that's the most common gap.

Mentioned, but wrong

An old address, a specialty you don't offer, a closed status when you're open. Fix the source first (your Google Business Profile is the most common culprit), since correcting your own site alone won't necessarily update what's already cached elsewhere.

Competitor named instead

This is often the most useful result to find, because it's the most specific. When the tool shows its sources (Google's AI Overview and Perplexity both do this), look at what it's citing. That tells you exactly where that competitor is visible that you aren't, which turns a vague sense of falling behind into a concrete, fixable list.

Keep a simple log: date, query, tool, and what came back. Checking five to ten priority queries once a month is enough to catch real movement without turning this into a full-time project.

What matters less than you'd think

Two things worth knowing so you don't spend limited time in the wrong place.

llms.txt, a file some sites now add specifically to guide AI crawlers, currently appears to have little measurable effect on whether you actually get cited. It's easy enough to add if your platform supports it, so there's no harm in having one. It just isn't where the real gains come from, so don't let it become the afternoon project that pushes the Foundation items above further down the list.

Keyword stuffing doesn't help the way it might have with older SEO. AI tools are built to extract a clean, quotable answer, and a page written for old keyword-matching logic often reads as exactly the kind of unclear content these tools skip over.

Frequently asked questions

No. The two overlap almost entirely. Structured data, a complete Google Business Profile, and clear, well-organized content help you rank on Google and get cited by AI tools at the same time. Nothing on this checklist works against traditional SEO.

There's no fixed timeline, and it depends how far your current online presence is from the Foundation items above. Directory and profile fixes can be reflected within weeks. Content and structured data changes tend to take longer to fully register, so treat this as ongoing maintenance rather than a project with a single finish line.

Not necessarily. Many website platforms, including WordPress with an SEO plugin, offer built-in or plugin-based tools for adding schema without touching raw code. If yours doesn't, this is a reasonable one-off task to hand a developer rather than attempt manually.

That's exactly what the testing section above is built to catch. If you find inaccurate information, correcting it at the source, usually your Google Business Profile or a directory listing, is the most direct fix.

Where to start

None of this replaces good clinical practice, and nothing here promises a specific outcome. What it does is make sure the people already looking for what you offer can actually find you when they ask. Start with the Foundation tab. It's the part every other section depends on, and it's the part most practices haven't touched yet.

If you've already decided whether to build your practice around a specialty, that decision affects several items on this checklist, especially the specialty pages. See our piece on whether to niche your therapy practice if that's still an open question for you.