Why isn't my treatment center showing up in ChatGPT?

First, confirm what the machines actually say

Before you fix anything, document the problem. Open ChatGPT, Gemini, and Google's AI Overviews in a logged-out or fresh session and ask the questions a family would ask: best rehab near your city, detox centers that take a common payer in your state, whether your brand name is legitimate. Run ten to twenty variations and record three things for each: were you named, was your website cited, and who showed up instead.

This baseline matters because AI answers move slowly. You will use the same prompts every month to prove progress, which is the core of measuring AI search visibility. If competitors and directories dominate every answer, you now know exactly which queries you are fighting for.

The four reasons AI engines skip your center

When a center never appears, the cause is almost always one of four gaps, and usually several at once.

• Entity weakness. Your name, address, and phone number differ across your website, Google Business Profile, state licensing records, and directories. The model cannot confirm you are one real organization, so it leaves you out.
• No citable sources. Your pages bury answers in marketing copy. There is no paragraph an engine can lift that cleanly answers what this center treats, where, and for whom.
• Thin content. A five-page website with a homepage, a generic about page, and a contact form gives a language model almost nothing to work with.
• No third-party corroboration. Nobody else says you exist: no accreditor profile, no SAMHSA locator listing, no press, no directory profiles, few reviews. Models trust what multiple independent sources repeat.

These are the same signals we break down in how AI engines choose which rehabs to recommend.

The fix sequence, in order

1. Clean up the entity first. One exact name, address, and phone everywhere: website footer, Google Business Profile, state license listings, accreditor directories, every citation you control. Add Organization and MedicalOrganization schema so machines can read who you are.
2. Publish answer-first pages. One page per real question: each level of care, each location, each payer topic. Open every page with a complete three-to-five sentence answer. The full method is in structuring content for AI citations.
3. Earn corroboration. Confirm your listing on findtreatment.gov, keep accreditor profiles current, and pursue association membership and local press. Every independent mention makes you more real to a model.
4. Build the review corpus. A steady, ethical review-request process gives engines recency and volume to draw on.
5. Re-test monthly. Same prompts, same log, every month, so you can see the ladder being climbed.

What to expect and when

This is a compounding play, not a switch. In the first thirty days you can complete the entity cleanup and ship your first answer-first pages. By sixty to ninety days, expect your brand-verification queries (is this center legit) to improve first, because those are the easiest for a model to resolve once your entity is coherent. Competitive discovery queries (best rehab near my city) typically take one to two quarters of consistent publishing and corroboration before you appear.