You can find out what AI says about your treatment center by asking ChatGPT, Gemini, and Perplexity direct questions about your programs, location, and reputation, the same way a prospective client or their family member would. Whatever comes back, accurate, outdated, or missing entirely, becomes your to-do list for fixing what these tools rely on. This takes less than twenty minutes and requires no technical background.
Why a self-audit reveals your real AI presence
A self-audit means sitting down and asking AI engines what they know about your center, then recording the answers exactly as given. This matters because families searching for addiction treatment are increasingly typing questions into ChatGPT or asking Gemini for a recommendation instead of scrolling through search results. If you don't know what these tools say, you don't know what a family hears before they ever call you.
Unlike a Google search, where you can see your own website and reviews sitting right there, AI answers are generated from a mix of sources you can't fully see: your website content, directory listings, review platforms, news mentions, and licensing databases. The engine decides what to trust and what to repeat. A self-audit is the only way to see that output the way a searching family sees it, rather than assuming your polished website is what gets surfaced.
The questions to ask each engine about your center
The questions to ask are the ones a family member searching at 2 a.m. would actually type: "best addiction treatment centers near your city," "does your center name offer detox," "is your center name in-network with your insurance," and "what do reviews say about your center name." Ask each question in ChatGPT, Gemini, and Perplexity separately, since each pulls from different sources and can give different answers.
Run the same set of questions across all three engines rather than just one, because a treatment center might be described accurately in one and missing or wrong in another. Also try broader, non-branded questions like "how do I find a detox center that takes my insurance in your city" to see whether your center comes up at all when a family isn't already searching for you by name. If your name never appears in those broader answers, that's as important a finding as an inaccurate one.
Reading the answers for errors and omissions
Reading the answers means comparing every claim the AI makes against what's actually true about your center today: your levels of care, accepted insurance, accreditation status, admissions process, and location. Flag anything wrong, anything outdated, and anything a family would need to know that simply isn't mentioned. Omissions are often more damaging than errors, because a family won't know to ask a follow-up question about a service you offer that the AI never brought up.
Pay close attention to how the AI describes your specialties. Also check whether the AI cites your accreditation, licensing, or any specific clinical approach (such as medication-assisted treatment) by name, since vague or generic descriptions read as less credible to someone comparing multiple centers.
Prioritizing fixes that change the recommendation
Prioritizing fixes means acting first on the errors and omissions most likely to change whether a family chooses you or a competitor, not just the ones that are easiest to correct. An outdated insurance claim or a missing detox service sits higher on the list than a minor wording issue, because it directly affects whether a family in crisis calls you or moves to the next name on their list.
Start with anything that could cause active harm: wrong phone numbers, incorrect claims about levels of care, or outdated admissions information. Next, address gaps that make you look less complete than you are, such as missing specialties or accreditation. Lower priority items are stylistic, like an AI answer that's accurate but generic. Recheck the same questions after your website and listings have had time to update, since AI engines refresh their sources on their own schedule and changes don't appear instantly.
The real objection: "isn't this a waste of time if I can't control what AI says?"
You can't force an AI engine to say a specific sentence about your center, and that's the objection worth naming directly. But you don't need control over the exact wording, you need the underlying information to be accurate, current, and complete wherever these tools look: your website, your directory listings, your review profiles, your licensing records. When those sources are correct and consistent, the odds of the AI repeating them correctly go up. The self-audit isn't about micromanaging an algorithm. It's about finding out where the information feeding it is wrong or missing, so you can fix the source instead of guessing at the output.