An owner with a sick cat asks an assistant first. An answer can only name a practice for what some page says.

Four moments when an owner asks an assistant instead of calling around

The questions cluster around four moments: the emergency at night, the new pet or the new town, the procedure with a price attached, and the animal most practices do not see. A fifth runs through all of them: the anxious dog or the cat that hates the carrier, whose owner types "fear free" or "low stress" and filters on it.

The emergency question is about now. The AI assistant needs your hours in text, what happens after close (an on-call vet, or the emergency hospital you refer to, by name), and the address once. A site that says "emergencies: call us" with no hours cannot tell an owner at midnight whether anyone will pick up, so the assistant cannot either.

The procedure question is about money before it is about you. "How much is a dental cleaning with anesthesia" can only be answered with a figure from pages that publish one. State the range you already quote on the phone and what it includes (the exam, bloodwork, anesthesia, extractions billed separately), and yours can be the page an assistant quotes.

The species question is a question about words. An owner with a bearded dragon types the animal's name, and a page that never names the animal gives an answer nothing to repeat about it. "We treat all animals" contains no animal. "Dogs, cats, rabbits, ferrets, birds and reptiles" contains six, and each one is a question your page can now answer.

What a practice website has to state before an assistant can name it

When an owner's question sends an assistant to the web, it fetches pages the way an AI crawler does and reads the text they serve; OpenAI's documentation names ChatGPT-User as the fetcher that opens a page on a user's behalf. It does not see the waiting room. It sees the words on the page, in the form they are written.

VeterinaryCare is schema.org's type for a vet's office, but it sits under MedicalOrganization, not under LocalBusiness, and schema.org gives opening hours to places, not to organizations. Google's local business documentation asks for each location as a LocalBusiness type and for more than one type as an array, so the practice's node carries both, ["VeterinaryCare", "LocalBusiness"], and the fields Google documents apply: name, address, telephone, opening hours, geo and price range. Filled in and matching the visible page, they make the LocalBusiness schema usable. Left as the page builder's defaults, the markup describes your website and not your practice.

One caution specific to this trade: describe the procedure, the visit and the price, not the outcome. An assistant quotes what a page states, and a claim about results travels with your name into every answer it appears in. Species by name, hours by day and a range with what it includes answer a question. A promise answers none.

What our crawl of 9,483 sites found, and what a practice should take from it

In our audit of 9,483 small business websites (40 trades, 90 metros, late August to early September 2026), a crawler that ran no JavaScript and obeyed robots.txt fetched the homepage, the contact page and the about page of each site. The audit publishes no separate cut for veterinary practices; the figures describe the whole sample, and the method section says how it was drawn.

61.5% had no question-and-answer content of any kind, and 50.7% had a heading phrased as a question and did not answer it in a form a machine can isolate. "Do you see exotics?" as a heading, followed by "call us to find out", is the veterinary version.

43.1% were inconsistent about who and where they are across their own three pages; 36.9% published more than one phone number, and 34.1% had no postal address a machine could parse. A practice with a main line, an after-hours line and a pharmacy line has three numbers; unlabeled, none of them answers the midnight question.

64.7% showed no review signal a crawler could read, and only 8.9% carried rating schema. 17.8% had no schema.org markup at all; of the 82% that did, most described the website rather than the business, and 13.7% had business schema missing the fields that would make it usable.

7.2% sat behind bot protection that served the crawler a challenge page instead of the site, and 3.0% rendered their content only in JavaScript. Our free crawler access check reads your robots.txt the way each of the nineteen AI crawler tokens would.

What to fix first, in order

1. One address and one main phone number, in text, identical on every page and in the markup. If there is an after-hours line, label it and keep it identical too.
2. Hours by day, and the after-hours arrangement in one sentence: who answers, or which emergency hospital you refer to.
3. Species and services by their plain names, one page per service, with the price range you quote by phone and what it includes. State whether you are accepting new patients.
4. VeterinaryCare and LocalBusiness as the node's two types, in JSON-LD on every page, with address, telephone, openingHoursSpecification and areaServed filled in. JSON-LD is the format Google recommends, and the markup should state only what a visitor can read on the page.
5. Reviews on the page as text, with a first name, a month and the platform. Google's review snippet rules give a practice no review stars for reviews of itself, and say a platform's ratings are not to be aggregated into AggregateRating on its own site.
6. The five questions your front desk answers most, as headings, with a real answer under each.
7. Check what a crawler receives: the crawler access check for robots.txt, then the source of your services page with scripts off.

Nothing on this list puts a practice into an answer, and nobody outside the companies that run the assistants can. The list takes away the reasons an assistant could not name you; whether it then does is a measurement, and ours is published.

Sources: our audit of 9,483 small-business sites, with its method · Google, local business structured data · Google, introduction to structured data · Google, review snippet · schema.org, MedicalBusiness · schema.org, LocalBusiness · OpenAI, crawler documentation