Readable by Google and by AI
AI assistants and Google's AI features build answers from pages they can crawl, index and trust, so most of the work that makes a practice visible in AI search is the same work that makes it visible in Google. Nobody outside the AI companies knows exactly how assistants choose which practice to name. This series sticks to what the companies document and what you can check yourself.
What is known, and what is guessed
AI search attracts more confident claims than almost any topic in marketing. It helps to sort them.
Documented by the companies themselves: Google says its generative AI features are rooted in its core Search ranking and quality systems, that there are no additional requirements to appear in AI Overviews or AI Mode, and that a page must be indexed and eligible to show a snippet to be used as a supporting link. OpenAI, Anthropic and Perplexity each publish the names of their crawlers and say which ones power search in their assistants and which collect training data. Google says it does not use llms.txt files, and that structured data is not required for its AI features.
Not documented anywhere public: how any assistant weighs one practice against another, how much reviews or directories count, or why the same question can produce different practices on different days. Anyone who claims to know these weights precisely is guessing, and Google warns that no third party tool has access to its internal ranking or AI systems.
The guides in this series say plainly which of the two each claim belongs to.
Start here
If you are new to the topic, read these four first. Together they explain how assistants find practices, how to see what they say about you, and the two settings that can quietly remove you.
- How ChatGPT, Gemini and Perplexity decide which practice to recommend separates what is documented from what is inference
- How to check what AI says about your practice today gives you a repeatable test you can run in half an hour
- robots.txt and AI crawlers explains which crawlers to allow and which you can safely block
- Your practice as an entity covers keeping your facts consistent across Google, directories and your own site
Then work through the rest of the list below, from schema markup to plain language.
What this means for a practice
The practical takeaway is reassuring. You do not need a separate AI strategy, a special file or a new kind of content. You need what patients have always needed from your site, done well:
- Pages that crawlers can reach and read without running scripts
- A clear page for each service, condition, provider and location
- Facts that match everywhere: name, address, phone, hours, insurance and providers
- Content written or reviewed by real clinicians, saying things only your practice can say
- A current Google Business Profile, which Google says can help businesses appear in AI responses
The practices that do these things are easier for any system to find and to describe correctly, whichever assistant a patient happens to use.
Every guide in this series
- 1What is GEO, and how is it different from SEO?
GEO, short for generative engine optimization, is the effort to have AI assistants mention and cite your practice when they answer a patient. Google says optimizing for its generative AI features is still SEO, and most of the work overlaps: crawlable pages, clear facts, real expertise and a consistent presence across the web. What differs is how answers are shown and measured.
- 2How ChatGPT, Gemini and Perplexity decide which practice to recommend
No AI company publishes exactly how its assistant chooses which practice to name. What is documented is the input: assistants that search the web draw on pages their crawlers can reach and on search indexes, and Google says its AI features are rooted in its core ranking systems and can draw on Business Profiles. The reasonable bet is to be easy to find, easy to verify and consistent everywhere.
- 3Schema markup for medical practices (MedicalClinic, Physician, FAQ)
Schema markup is code that labels the facts on your pages, such as name, address, hours and providers, so machines read them without guessing. Use the most specific schema.org type that fits, such as MedicalClinic or IndividualPhysician, and mark up only what patients can see. Google says it is not required for its AI features, and FAQ rich results no longer appear in Google Search.
- 4llms.txt: what it is and whether it matters yet
llms.txt is a proposed file, first published in 2024, that gives AI tools a plain markdown summary of a website with links to its key pages. Google says Google Search, including its AI features, ignores it, and the crawler documentation from OpenAI, Anthropic and Perplexity does not say their assistants read it on ordinary websites. It is harmless to add, but it belongs near the bottom of a practice's list.
- 5robots.txt and AI crawlers: should you block GPTBot?
Usually there is no need, and if you do, block only the training crawlers. GPTBot collects content that may be used to train OpenAI's models; OAI-SearchBot is the separate crawler that lets ChatGPT search show your site. Blocking the search crawlers of ChatGPT, Perplexity or Claude can keep your practice out of their answers, which for most practices costs more than it protects.
- 6Sites built in JavaScript that AI crawlers cannot read
Google renders JavaScript, so it can usually see content your site builds in the browser, though rendering can be delayed. OpenAI, Anthropic and Perplexity do not say in their crawler documentation that their crawlers run JavaScript, and Google itself notes that not all bots can. If your provider names, hours or services appear only after scripts run, assume some AI crawlers see a nearly empty page.
- 7Writing answer first content that AI quotes
Answer first means the first sentence under each question gives the answer, and the explanation follows. It is how patients skim on a phone, and it produces short, self contained passages that are easy to quote accurately. Google says you do not need to write in a special way for its AI features, so treat answer first as good patient writing that also happens to travel well, not as a trick.
- 8Your practice as an entity: Google Business Profile, Healthgrades, Doximity and consistency
To Google and to AI assistants, your practice is an entity: one real organization with a name, address, phone, providers and specialties, described by many sources across the web. When your website, Google Business Profile and directory profiles agree, every source supports the same facts. When they conflict, a system has to choose between them, and patients may see the wrong one.
- 9Why original data gets cited and opinions do not
A page with something no other page has, such as your measured wait for a new patient visit or the questions your phones get most, gives search engines and AI assistants a reason to point to you rather than to pages repeating common knowledge. Google asks directly whether content offers original information or analysis. Nobody outside the AI companies can prove what they cite.
- 10Physician reviewed content and E-E-A-T for medical pages
E-E-A-T stands for experience, expertise, authoritativeness and trust, the qualities Google asks its quality raters to assess, with trust the most important. It is not a single ranking factor, and health topics face the strictest standard. For a practice, the strongest signal is simple and honest: real clinicians who write or review the page, named on it.
- 11Plain language and reading level for medical practice websites
Write so a patient understands the page the first time they read it: short sentences, everyday words, each medical term explained once, and the most important point first. That is how the CDC defines plain language. It helps worried patients act, and a sentence that is hard to misread is also hard to misquote, though Google says its AI features do not require any special writing style.
- 12Headings, lists and tables: structure machines can parse
Use real headings that name what each section answers, lists for steps and options, and simple tables for facts that compare, such as hours by location or plans accepted at each office. Structure helps patients skim and helps screen readers and crawlers understand the page. Google says its systems understand multiple topics on a page without special chunking, so structure for the reader, not for a bot.
- 13How to check what AI says about your practice today
Ask the assistants yourself, the way patients would: by specialty, location and need, not only by your practice name. Run the same set of questions in ChatGPT, Gemini, Perplexity and Google's AI Mode, record who gets named and what each says about you, and repeat monthly, because answers vary from one run to the next. Google's Search Console now also reports how often your pages appear in AI Overviews and AI Mode.
Does ChatGPT recommend your practice?
The Surge Score measures five things about your practice website: how often ChatGPT, Gemini and Perplexity recommend you, your PageSpeed, your structured data, your indexing and your reviews, each with its evidence and date. Free.
Frequently asked
Do I need a separate strategy for AI search?
Not a separate one. Google says optimizing for its generative AI features is still SEO and requires no special files or markup. Other assistants rely on their own crawlers and search sources, which read the same public pages. Strong fundamentals cover all of them.
Can I pay to be recommended by an AI assistant?
Google says there is no way to pay for better local ranking, and none of the crawler documentation we reviewed describes paying for placement inside an assistant's answer. Some assistants sell separate ads; OpenAI, for example, documents a crawler that checks pages submitted as ads on ChatGPT.
Which AI assistant matters most for my practice?
No one can say with certainty, and it changes quickly. Google's AI Overviews appear inside the search results most patients already use, so Google remains the place to start. Check the others with the test in our guide on what AI says about your practice.
How long does it take to show up in AI answers?
There is no published timeline. Crawlers have to reach your updated pages, and OpenAI and Perplexity say robots.txt changes take about a day to register. Content and reputation changes take longer to show, so measure over months rather than weeks.
Sources
- AI features and your website, Google Search Central
- Optimizing your website for generative AI features on Google Search, Google Search Central
- Overview of OpenAI crawlers, OpenAI
- Perplexity crawlers, Perplexity
- Does Anthropic crawl data from the web, and how can site owners block the crawler?, Anthropic
- Tips to improve your local ranking on Google, Google Business Profile Help