You measure whether AI search is bringing clients to your speech therapy clinic by combining three checks: asking new clients directly how they found you, watching for unusual upticks in direct or branded website visits, and periodically typing real client questions into ChatGPT, Gemini, and Perplexity to see whether your clinic appears in the answer. No single number proves it, but these three signals together give you a clear, honest picture.
Asking new clients how they found you
The single most reliable signal for whether AI search tools are sending clients to your practice comes from asking them directly during intake. A short question on your intake form or in the first phone call, phrased as "how did you hear about us," will occasionally surface answers like "I asked ChatGPT for a speech therapist near me" or "Google's AI summary mentioned your clinic." These direct mentions are rare compared to referrals or insurance directory searches, but they are unambiguous when they happen.
Train front-desk staff or intake coordinators to listen for phrasing beyond the usual options of "Google," "referral," or "insurance list." Parents and adult clients researching speech-language pathology services often describe their search process in more detail than expected, especially if they used a conversational tool to compare providers or understand diagnoses like apraxia or aphasia before calling. Add an open text field rather than only checkboxes so these details aren't lost. Over several months, patterns emerge even without formal tracking software.
Watching for direct and branded searches
Branded and direct search traffic, meaning visits where someone typed your clinic's name directly into a browser or search engine rather than a generic term like "speech therapist near me," is a strong secondary indicator that AI tools are surfacing your practice by name. When an AI engine answers a question by naming your clinic specifically, the person often goes to a search engine next and searches your name to verify or find your contact details, rather than clicking a link inside the AI tool itself.
Check your website analytics platform for a rising trend in direct traffic or searches containing your clinic's name over time. This won't tell you which specific AI engine sent someone your way, since most AI tools don't pass detailed referral information the way a typical website link does. But a steady increase in people searching your clinic by name, especially from areas outside your usual referral network, suggests something is introducing your practice to new audiences. Compare this trend against periods when you know referral sources or advertising were flat, to isolate what's changed.
Testing prompts in the major engines yourself
Running the same questions a prospective client might ask directly through ChatGPT, Gemini, Perplexity, and Google's AI Overviews shows you exactly what these tools currently say about your clinic and your competitors. Try phrasings like "speech therapist for a toddler with a language delay in your city" or "best pediatric speech-language pathologist near your neighborhood" and read the full response, not just the first sentence.
Pay attention to whether your clinic is named at all, what details the engine includes (address, specialties, insurance accepted), and whether that information is accurate. Note which competitors appear alongside you or instead of you. Because these tools generate responses based on current web content, testing the same prompts every few weeks lets you see whether your visibility is improving, stable, or slipping, and whether the details being shared about your practice are correct or outdated. Keep a simple log of the date, the prompt, and what was said, so you're comparing consistent snapshots over time rather than relying on memory.
Deciding what to adjust based on what you see
The value of tracking these signals is in using them to decide where to focus attention, not simply to collect data. If intake conversations reveal that clients are describing symptoms or conditions in AI tools before finding you, your website and any public profiles should clearly address those same conditions in plain language, since that's the content AI engines draw from when constructing answers. If branded search traffic is rising but AI prompt tests show your clinic missing from responses, the disconnect suggests people are hearing about you through word of mouth or other channels, not AI search specifically, and it may be premature to credit AI tools for that growth.
If prompt testing shows an AI engine describing your clinic with outdated hours, an old address, or missing specialties like feeding therapy or accent modification, that's a direct, fixable problem: outdated or incomplete information published anywhere online, including your own site, is likely being repeated by the AI tool. Correcting it on your website and any directory listings gives these engines more accurate material to draw from next time they're asked. Treat each signal as a prompt for a specific, small correction rather than a verdict on whether AI search is "working" overall.
Checking your own progress without waiting on anyone else's report
You don't need to depend on a third party to tell you whether this is working. Set a recurring reminder, every few weeks, to do three things yourself: glance through recent intake forms for any mention of AI tools or unusual search phrasing, pull up your website analytics to see whether direct or branded traffic is trending up or flat, and run the same handful of AI prompts you'd expect a prospective client to ask, comparing the answers to your log from the previous check. Keep this log in a simple document or spreadsheet you control, so you always have a record to compare against, even months later. This routine takes a small amount of time each month and gives you a direct, current answer, based on what clients tell you and what the AI tools themselves are saying, without needing to interpret anyone else's summary of it.