AI search for hearing care practices

Is your practice one of the names?

Somebody’s daughter asks ChatGPT where her dad can get a hearing test near home. She gets three names and calls one of them. Nobody looked at a website, and nobody scrolled past an ad.

Whether your practice is in that answer depends on what the assistant could read about you. Most practice websites were never built to be read by one.

Take the self-check

What every site we build ships with

The AI crawlers, named and invited.
Every site we build names and invites GPTBot, ClaudeBot, PerplexityBot, Google-Extended and OAI-SearchBot. Not a blanket “everyone welcome” line that an assistant has to interpret. Each one, by name.
A fact sheet written for machines.
Every site ships an llms.txt with your practice facts: who you are, every office, the hours, the services, and a link straight to the right page for each.
Clinic markup on every location.
Location pages carry medical-clinic, FAQ and speakable schema, so an assistant reads “this is a hearing clinic, here is its address, these are its common questions” instead of guessing from the layout.
A check, every month.
A monthly report on whether AI assistants name your practice, and what we’re doing about it when they don’t.

Take a site apart

Three files a patient never sees, and an assistant reads first. Shown on a sample practice.

01 · robots.txt

The doorman

The first file any crawler asks for. This one greets the AI crawlers by name, so nobody has to interpret a blanket rule.

riverbendhearing.example/robots.txt01 · The doorman

1User-agent: *
2Allow: /
3 
4User-agent: GPTBot
5Allow: /
6 
7User-agent: OAI-SearchBot
8Allow: /
9 
10User-agent: ClaudeBot
11Allow: /
12 
13User-agent: PerplexityBot
14Allow: /
15 
16User-agent: Google-Extended
17Allow: /
18 
19Sitemap: https://riverbendhearing.example/sitemap.xml

02 · llms.txt

The fact sheet

Written for an assistant, not a person. Offices, hours and services, each with a link to the exact page, so the answer it gives is the right one.

riverbendhearing.example/llms.txt02 · The fact sheet

1# Riverbend Hearing Center
2> Hearing tests, hearing aids and repairs in Riverbend, Texas. Two offices.
3 
4## Offices
5 
6- [Main Street](https://riverbendhearing.example/locations/main-street/): 120 Main St. Mon–Fri 8–5. (940) 555-0100
7- [Lakeview](https://riverbendhearing.example/locations/lakeview/): 48 Shore Dr. Tue–Sat 9–5. (940) 555-0101
8 
9## Services
10 
11- [Hearing evaluations](https://riverbendhearing.example/services/hearing-test/)
12- [Hearing aids](https://riverbendhearing.example/services/hearing-aids/)
13- [Repairs and cleanings](https://riverbendhearing.example/services/repairs/)

03 · locations/main-street/

The labels

Invisible markup on the location page: a medical clinic with its address and hours, its common questions, and the lines to read aloud.

riverbendhearing.example/locations/main-street/03 · The labels

1{
2  "@type": "MedicalClinic",
3  "name": "Riverbend Hearing Center, Main Street",
4  "address": "120 Main St, Riverbend, TX",
5  "telephone": "(940) 555-0100",
6  "openingHours": "Mo-Fr 08:00-17:00",
7  "availableService": "Hearing evaluation"
8}
9{ "@type": "FAQPage",
10  "mainEntity": [ "Do you take walk-ins?", … ] }
11{ "@type": "WebPage",
12  "speakable": {
13    "cssSelector": ["h1", ".clinic-summary"]
14  } }

What we found when we looked

In our October 2026 check, no other hearing-care agency’s site or sampled client site named AI crawlers in robots.txt.

We read the robots.txt file on 29 sites on October 17, 2026: hearing-care agencies’ own sites and a sample of the client sites they publish. A file that doesn’t mention an AI crawler isn’t necessarily blocking it. It just isn’t saying anything, and that’s a choice an assistant has to interpret.

It’s one check, on one day. We’ll run it again, and if the answer changes we’ll change this page.

The two-minute self-check

Eight questions. Most take one search or one address typed into a browser. Nothing is sent anywhere.

Ask ChatGPT, “Who’s a good hearing provider near [your town]?” Is your practice named?
Ask Perplexity the same question. Are you named there too?
Search the same question on Google and read the AI answer at the top. Are you in it?
Open yourpractice.com/robots.txt. Does it name GPTBot, ClaudeBot and PerplexityBot, and allow them?
Open yourpractice.com/llms.txt. Is there one, and does it list your offices, hours and services?
Does every office have its own page, with its own address, phone number and hours?
Paste a location page into validator.schema.org. Does it find a MedicalClinic, or a similar clinic type?
Does anybody send you a monthly answer to “did AI assistants name us this month?”

Your score

0 / 8

Start with the first three. If you aren’t named, the rest of the list is where the answer usually is.

Have us run it with you

Your answers stay in this browser. Nothing is sent.

What this can’t do

Nobody can make an assistant recommend you, and anybody who promises it is guessing. What these files do is make sure that when an assistant goes looking, it finds clear facts instead of a guess.

Reviews, your listings and how long you’ve been in the community still matter more than any file. That’s why the monthly report checks those too.

How the monthly report works

Questions owners ask

What is robots.txt?

A short text file at the root of every website that tells automated visitors which parts of the site they may read. Most practice sites only address search engines in it. Ours name each AI crawler and invite it in.

What is llms.txt?

A plain-text page written for AI assistants: who the practice is, where each office is, its hours and services, with a direct link to the right page for each. It saves an assistant from having to work those facts out from the design.

What is schema?

Invisible labels in a page’s code, in a vocabulary every search engine and assistant shares. It says “this is a medical clinic, this is its address and phone number, these are its questions and answers, and this is the paragraph to read aloud.”

Can you guarantee ChatGPT will recommend my practice?

No. Nobody can. We make sure your site gives assistants clear, correct facts, and we check every month whether they name you, so you know where you stand instead of guessing.

If I let AI crawlers in, can they copy my site?

They read the same public pages any visitor or search engine already reads. Some of these crawlers gather material for training AI models and others fetch a page to answer a question in the moment. We allow both by default, because a practice wants to be known. If you would rather your pages not be used for training, we can turn the training crawlers off and leave the answering ones on.

Do I need a new website to do this?

Every site we build ships with all of it. If you are not ready to change sites, the self-check above will tell you what your current site is missing, and you can take that list to whoever runs it.

Want us to run the check for you?

Bring your site to a strategy call. We’ll ask the assistants about your town on the call, read your files with you, and tell you what we’d fix first, whether or not you end up working with us.

Book a strategy call