Skip to main content
AI Search GuidePsychology And Counseling

How to tell whether AI search is already sending you clients

Practices rarely know whether ChatGPT, Gemini, or Perplexity are quietly referring new clients. Here's how to find out with a few intake questions and direct prompt tests.

· 5 minute read

You can tell whether AI search is sending you clients by asking new inquiries a direct question about how they found you, watching for language patterns that mirror how AI assistants phrase referrals, and testing common client prompts yourself in ChatGPT, Gemini, and Perplexity to see who gets named. If a new client mentions asking an assistant for a therapist and your practice never comes up in your own tests, that's a visibility gap worth closing.

Most counseling practices have no system for catching this. A client search that once ran through Google now often runs through a conversational assistant that gives one direct answer instead of ten blue links. If that assistant never mentions your practice, you lose the referral before you ever see a form submission or a missed call. The rest of this guide walks through how to notice when it's happening and how to build a habit of checking.

Signs AI referrals are reaching you

The clearest sign that AI search is sending clients your way is when someone mentions "I asked ChatGPT" or "an AI recommended you" during their first phone call, email, or intake form. Other signs include inquiries that already know specifics about your specialties, fees, or approach before you've explained anything, since AI-generated answers often summarize that information upfront.

Watch for phrasing too. Clients referred by search engines usually reference a listing, a review site, or "I found you on Google." Clients referred by an AI assistant tend to describe a conversation: they asked a question about finding a therapist for a particular issue, location, or insurance situation, and got a short list or a single name back. If your practice's name showed up in that list, they'll often say so unprompted. If you never ask, you'll likely never hear it, because most people don't think to volunteer which tool they used unless it comes up naturally.

How to ask new clients how they found you

Adding one simple, low-pressure question to your intake process is the most reliable way to learn whether AI search is contributing to your inquiries. Ask new clients directly: "How did you first hear about us?" and give an open text field rather than a dropdown limited to "Google," "referral," or "insurance directory," since a fixed list won't capture "I asked an AI chatbot" unless you include it.

Phone intake staff and scheduling forms should ask the same question consistently, not just occasionally. If you use an online scheduling tool, add a required short-answer field before the first session is booked. Review the answers monthly rather than trying to interpret them one at a time. A pattern of even a few mentions of ChatGPT, Gemini, or "an AI search" over a few months tells you something that anecdotal impressions can't: that assistants are actively surfacing your practice to people who then act on it by booking.

What to watch for in inquiry patterns

Inquiry patterns shift in specific ways when AI search starts contributing to your client flow, and noticing the shift matters more than any single client's story. Look for new clients who arrive already knowing your specialties, session format, or general fee range, since AI-generated answers frequently summarize that information from your website or profile before the client ever speaks with you.

Also watch timing and phrasing. Clients coming from a conversational assistant often ask fewer basic logistical questions upfront ("do you take insurance," "where are you located") because the assistant already answered those in its summary. Instead, they ask more specific follow-up questions about fit, such as whether you work with a particular concern or age group. If you notice more inquiries arriving pre-informed and fewer arriving cold, that pattern is consistent with AI search functioning as a referral source, even if no one explicitly names the tool they used.

Why direct testing of prompts reveals your visibility

Testing common client prompts yourself is the most direct way to know whether AI search recommends your practice, because it removes the guesswork of waiting for a client to mention it. Open ChatGPT, Gemini, and Perplexity and ask the questions a prospective client would ask, such as "Who are good therapists for anxiety near your city" or "Find a counselor who works with couples in your area."

Pay attention to whether your practice appears at all, how it's described, and which competitors appear instead. Run the same prompts with small variations, changing the concern, the location phrasing, or the insurance mentioned, since AI assistants can give different answers depending on wording. Do this test periodically rather than once, because these tools update their sources and rankings over time. If your practice never appears across a range of reasonable prompts, that's a concrete, testable sign that AI search is not yet sending you clients, regardless of what your intake data shows.

A simple routine to track AI-driven inquiries

A repeatable routine turns occasional signals into a reliable measurement of AI-driven inquiries over time. Set a recurring monthly check that combines three habits: review the open-text "how did you hear about us" answers from new intakes, re-run your list of test prompts across the major AI assistants, and note any client mentions of AI tools during calls or first sessions in a shared log.

Keep the log simple, a spreadsheet with the date, the assistant mentioned or tested, and what was said or shown is enough. Compare month to month rather than expecting a dramatic single-month change. If mentions increase and your test prompts start returning your practice more consistently or more prominently, that combination is a reasonable indicator that AI search is contributing to your client flow. If both stay flat, you know where the gap is, and you can treat it as a specific, measurable problem rather than a vague worry.

What it looks like when the answer names someone else

Picture a person searching for support after a difficult diagnosis or a relationship crisis. Instead of scrolling through search results, they open an AI assistant and type, "Find me a therapist near your city who specializes in grief counseling and takes my insurance." The assistant responds with a short, confident paragraph naming two or three practices, complete with a specialty description and a reason each might be a good fit.

Your practice may do exactly that work, but if it never appears in that answer, the person calls one of the practices that did. They never see your website, never read your bio, never know you existed as an option. That's the moment this whole exercise is about catching, not after the fact, but before it happens again.

Want to See What AI Says About Your Business Right Now?

Book a 30-minute call and we’ll pull it up together — who gets named for your market’s questions, and where you stand. Free, and the picture is yours to keep.