What Is AI SEO?

AI SEO is the work of making your practice easy to find on Google and easy for AI tools such as ChatGPT, Google AI Overviews, Google AI Mode and Perplexity to name when a patient asks a question. For a health practice, it means the patient who types "functional medicine near me" and the patient who asks a chatbot for a recommendation both end up seeing your name. It is the umbrella term I use for the work other people call answer engine optimization and generative engine optimization.

How AI SEO works

Patients now ask for help in several places, and each one picks practices differently. It helps to sort them into three groups.

• Search engines with AI built in. Google's AI Overviews and AI Mode write a short answer at the top of the results and link to pages that support it.
• Chat tools that search the web. ChatGPT and Perplexity look up live pages, then write an answer and show the sources they used.
• Models answering from memory. When a chat tool does not search, it writes from patterns it learned in training, which is frozen at some past date (see training data).

Inside each group, a practice can win at three different levels. You can rank, which means a classic listing or map result. You can be linked, which means your page appears as a source beside an AI answer. And you can be named, which means your practice name appears in the sentences of the answer itself. Most owners only track the first level. The patient who asks a chatbot does not see it.

Google is direct about the technical side. Its documentation says there are no additional requirements to appear in AI Overviews or AI Mode, and that a page must be indexed and eligible to show in Google Search with a snippet. It also says you do not need new machine readable files, AI text files or special markup. So the base layer is unchanged, which is pages Google can find, read and show.

What changes is how much weight lands on sources outside your own website. Three recent studies point the same direction, with a caution that all of them measure correlation, not cause.

• SE Ranking studied 129,000 domains and 216,524 pages and found that the number of sites linking to a domain was the strongest predictor of ChatGPT citations. Domains with up to 2,500 referring domains averaged 1.6 to 1.8 citations, and those with over 350,000 averaged 8.4.
• Ahrefs studied 75,000 brands and found YouTube mentions had a correlation of about 0.737 with AI visibility, with branded web mentions at 0.66 to 0.71. Backlinks alone were only weakly correlated.
• MGMA published an analysis of 33,340 ChatGPT citation events across 200 practices. It was written by the founder of a commercial AI visibility service, so read it as one vendor's data. It found the cited sources were spread across hospital staff rosters, practice websites, specialty association pages and national directories.

The pattern is that AI tools lean on what many independent places say about you, which is why AI SEO reaches beyond your site. The SE Ranking and Ahrefs results can look opposed on links. They measure different things, citations of pages versus brand visibility, so I read them as complementary. The methods split into two camps. Answer engine optimization shapes your own pages so a tool can quote them.

The other camp is generative engine optimization, which shapes how the wider web describes you so a model forms the right picture of your practice.

It helps to sort the work into levers you control and false levers. The real levers are your own site, your footprint on other sites, and your discipline about facts, meaning every claim is accurate, dated and the same wherever it appears. When independent places agree about you, that is corroboration. The false levers are fake mentions, bought reviews and mass-produced pages. They are manipulative, and in my judgment they also backfire on a practice whose name is its whole reputation. Ordinary search stays underneath all of this, so classic ranking and a technically healthy site still matter.

Last, a note on crawlers, the automated programs that read websites. OpenAI uses one bot to surface sites in ChatGPT search and a separate one for model training, and the settings are independent. A blanket rule that blocks every AI bot can remove you from the answers you want. I cover that in ChatGPT search.

Example: AI SEO at a health practice

This example is illustrative, not a transcript. Picture a four-provider functional medicine clinic in a Denver suburb. The clinic ranks in the Google map results for "functional medicine near me" and gets a steady flow of calls from that.

The owner runs the 60-second test. In Google, the clinic appears in the map. In ChatGPT, asked who treats Hashimoto's near Denver, the answer names two competitors and a hospital system. In Perplexity, the numbered sources include a state-level directory, a specialty association page and a regional magazine list. The clinic is on none of them.

That one test sorts the problem. The clinic wins at ranking and loses at being named. The gap is not the website. It is the set of independent pages the tools trust.

Here is how the MGMA data helps read that result. In its sample, hospital staff rosters made up 27.5% of cited sources, practice-owned websites about 23%, specialty association credentialing pages 22.1% and national medical directories 8.3%. Regional "Top Doctors" magazines were 2.7%. Those figures come from one study of ChatGPT through the API in May 2026, so they describe a snapshot, not a rule. But they suggest where to look. If the clinic's providers are missing from the association directory and the national directories, that is a concrete list of pages to fix.

The owner's action list is short. Claim and complete the missing directory entries. Make provider names, credentials, address and phone identical everywhere. Add a plain two-sentence answer to the top of the Hashimoto's page. Then rerun the same three questions next month and compare. No special code is involved, and the plan does not depend on any one tool.

How to check yours in 60 seconds

1. Open ChatGPT and ask, "Who are the best [your specialty] practices in [your city]?"
2. Ask the same question in Perplexity, then in Google. If you see an AI Mode tab, use it.
3. Write down whether your practice name appears, whether your site is linked, and which competitors are named.
4. Rephrase the question once and repeat.

Answers vary by wording and by day, so record the date and the exact wording you used.

What to do about it

1. Confirm your key pages are indexed in Google Search Console and load quickly on a phone.
2. Give each service and condition its own page, and answer the patient's question in the first two sentences.
3. Match your practice name, address, phone number and specialty exactly on every listing and directory.
4. Get your providers listed on specialty association pages, licensing board pages and national directories, and keep your Google Business Profile accurate.
5. Check that your robots.txt file does not block OAI-SearchBot, the OpenAI bot that surfaces sites in ChatGPT search.
6. Keep a monthly log of the same ten patient questions across ChatGPT, Perplexity and Google.

The complete method, step by step, is in AI SEO for Health Practices: The Complete Guide.