When someone asks AI for a telehealth clinic, be one of the names it gives.

ChatGPT, Perplexity and Google's AI answers read the web, pick a few sources and name them. We make your brand readable, quotable and consistent enough to be one of them. It's one of the six workstreams in our telehealth SEO service.

Four things an engine needs before it will name you.

This is the checklist our SEO team audits every site against. None of the engines publish exactly how they choose sources, so it's built from their own crawler documentation and from what we've seen get cited.

  1. 1Understand

    A website it can understand.

    Its crawler has to get in, and the page has to make sense without running JavaScript. If your firewall blocks it or your text only appears after a script loads, you don't exist to that assistant.

  2. 2Parse

    A structure it can parse.

    One topic per page, headings that say what's under them, structured data written into the HTML, and internal links that show how your pages relate to each other.

  3. 3Lift

    An answer it can lift.

    The answer in the first two sentences, and questions answered in blocks that stand on their own, so a passage still makes sense when it's quoted out of context.

  4. 4Credit

    A human it can credit.

    A named author, a named clinical reviewer where there is one, and a business whose name, address and licensing read the same everywhere it's listed.

The work, in four parts.

Access01 / 04

Every AI crawler checked

We test each search crawler against your site and fix what blocks it. On Cloudflare that's two separate settings, and fixing one looks exactly like fixing both.

Structure02 / 04

Pages built to be quoted

The answer in the first two sentences, questions answered in blocks that stand on their own, a named author, and structured data written into the page's HTML so crawlers that skip JavaScript still see it.

Entity03 / 04

One version of you

Organization and founder data that ties your site, your socials and your listings to one business, plus the same name, address and phone everywhere you're listed.

Mentions04 / 04

Your name on other sites

Directory profiles, provider listings and roundup articles. When we checked search results for telehealth marketing agencies, three of the top eight were best-of lists, so getting onto the right ones is part of the job.

We found this out the expensive way.

For five months, Cloudflare returned an error to every AI crawler that tried to read scaleclinics.io, while Google's crawler got through fine. Fourteen articles we'd written to be cited by ChatGPT were invisible to it, even when someone pasted the link straight into a chat.

When we opened the firewall, a second setting was still working against us: a robots file Cloudflare serves ahead of your own, telling the well-behaved crawlers to stay away. We fixed both. That sweep is now the first thing we run on every site we take on.

Live on scaleclinics.io today
  • Structured data in the HTML of every page, not injected by script
  • One organization and founder entity tying the site to our profiles
  • A curated llms.txt describing what each page answers
  • AI search crawlers allowed through, checked after every change
  • Service pages that answer the question first, with standalone FAQs
scaleclinics.io/llms.txt

Three numbers that get mixed up.

Retrievals

How often AI crawlers fetch your pages. Easy to count, and all it tells you is that they're reading.

Mentions

How often you're named when the engines are asked the questions your buyers ask. We benchmark this in Ahrefs Brand Radar in month one and track it every month.

Patients

The only one that pays. ChatGPT tags some links it shows with utm_source=chatgpt.com, so those clicks show up. People who read your name and type your address in don't, which is why we add a how-did-you-hear question to your forms.

What changes when the buyer asks a machine.

01

What an AI crawler sees on a telehealth site

A lot of telehealth sites are built on page builders and quiz funnels that draw everything with JavaScript. A person sees a full page. A crawler that doesn't run scripts sees a title and not much else. The intake flow is often worse, because it's hosted by the provider platform behind a form or a login, and no crawler will ever get that far.

So the pages that explain your program, your pricing and your clinicians need to exist as plain, server-rendered pages on your own domain. We check this by fetching your pages the way a crawler does, with scripts off, and reading what actually comes back.

02

Health questions get more careful answers

Assistants hedge on medical questions and usually tell the person to talk to a clinician. For a telehealth brand, the next question is where to find one. We write for that moment, with pages that say plainly who your clinicians are, which states they're licensed in and what happens at a visit, so an engine has something specific to cite.

It also means the compliance rules on your site apply here too. If an engine repeats an overblown claim it found on your page, it's still your claim, and it's now in front of a buyer with your name on it.

03

One set of facts, repeated everywhere

When an engine finds your brand described three different ways, with a different phone number on a directory, an old address or a founder title that changed, it has less reason to trust any of them. Most of the entity work is unglamorous: making the same facts match across your site, your provider's pages, directories and profiles, and marking them up on your own site so a machine can read them.

Publishing more articles won't fix a brand the engines can't identify. Once the facts agree, new content has something to attach to.

04

What we report each month

Each month you get the Brand Radar read against your month one benchmark, a set of checks where we ask the engines the questions your buyers ask and record who gets named, which pages AI crawlers fetched where your host shows it, and any clicks tagged from AI tools that reached your site.

We report those separately and don't roll them into one score. A single number hides which part is moving.

AI search questions

What is AI search visibility?

It's how often AI tools like ChatGPT, Perplexity and Google's AI Overviews name or cite your brand when someone asks a question you should be the answer to. The work also goes by GEO, for generative engine optimization, and AEO, for answer engine optimization.

Is this different from SEO?

It overlaps a lot. Google's AI Overviews run on Google's normal crawler, so good SEO carries straight over. The other assistants use their own crawlers and lean harder on page structure and on what other sites say about you, so we do extra work there. We run it as one of the six workstreams in our telehealth SEO service.

Can you guarantee ChatGPT will recommend us?

No. The engines don't publish how they choose sources, and the same question can get a different answer the next day. We benchmark where you're mentioned now, do the work, and show you the change every month.

How long does it take?

Access and structure fixes count as soon as the crawlers re-read your pages. Getting named takes longer, because it depends on mentions building up on other sites. Plan on months.

Should we block AI crawlers from our site?

Blocking the search crawlers, like OAI-SearchBot, ChatGPT-User, Claude-SearchBot and PerplexityBot, means those assistants can't cite you. Training crawlers like GPTBot and CCBot are a separate decision about whether your pages train future models. Google's AI answers aren't affected either way, since they use Googlebot.

Does llms.txt help?

Google has said it doesn't use it. Some other tools read it, it takes about an hour to write, and it gives you one file that states what each page covers, so we write one and don't count on it.

Find out if AI tools can even read your site.

Book a call and we'll tell you what's in the way, starting with whether the AI crawlers can reach you.

Book an SEO call

No guarantee of mentions. Part of our telehealth SEO service.