Example Dental Clinic

Across 48 customer-style answers, Example Dental Clinic was named in 5, which is 10 percent (the range, allowing for how much answers vary, is about 5% to 22%). That is the Developing band. ChatGPT and Perplexity named the clinic most often (2 of 12 each) and Claude did not name it in any of its 12 answers.

Asked about the clinic directly, every assistant found it and described it accurately. When a customer asks for a dentist in Brisbane, the assistants recommend Riverbend Example Dental and Northgate Sample Dentists instead, and describe them with specific services, hours and reviews. The clinic is described, where it is described at all, in one generic sentence.

The biggest gaps are on the clinic's own website and profiles: the structured data does not state services, hours or location, and the health directory the assistants cite most lists an older address. The most useful first step is to correct that directory listing and complete the Google Business Profile.

Named most often by ChatGPT and Perplexity (2 of 12 each), once by Gemini (1 of 12) and not at all by Claude (0 of 12). With a dozen answers per assistant these counts are estimates, so the differences between assistants are indicative rather than exact.

The pattern by question: the clinic was named for "best dentist in Brisbane" and "who are the top dentist businesses near Brisbane", and for "I need a dentist in Brisbane, who would you recommend" by Perplexity only. It was never named for "what should I check before choosing one, and who would you suggest", the question where the assistants give advice first and a shortlist second.

Your own website was cited in 3 of 48 customer-style answers. In two of those the assistant cited the site and still recommended other clinics, which usually means the page it read did not give it a reason that matched the question.

The two questions that ask about the clinic by name (Q4 and Q5) were answered by every assistant, 24 answers in all. Accuracy: 22 correct, 2 wrong. The two wrong answers gave the older address from the health directory listing. Otherwise the descriptions are correct: a family dental clinic in Brisbane, general and cosmetic dentistry, online booking, a 4.8 rating from 61 reviews on Google.

On the day of the audit the two main customer questions were asked once each, by hand, in Google AI Mode. This is a spot check of two answers on one day. It is not repeated like the API runs and it is not part of the visibility figure.

• "best dentist in Brisbane": not named. Google recommended Riverbend Example Dental, Northgate Sample Dentists and Harbourline Demo Dental, and said of the first: "A long-running family practice with extended hours, online booking and consistently high ratings." It worked from Google Business Profiles and one review site.
• "I need a dentist in Brisbane, who would you recommend and why": not named. Google grouped clinics by need (families, cosmetic, emergency) and named Riverbend Example Dental, Southside Sample Smiles and Northgate Sample Dentists.

Google's AI leans on Google Business Profiles, where the clinic's rating is strong but its services list is short.

The clinics named most across the 48 customer-style answers:

• Riverbend Example Dental: named in 31 answers. Described with specific services (family, cosmetic, emergency), extended opening hours and online booking.
• Northgate Sample Dentists: named in 24 answers. Described by location and transport access, with many reviews cited.
• Harbourline Demo Dental: named in 17 answers. Described as family-focused, with children's services named.
• Southside Sample Smiles: named in 9 answers. Described for cosmetic work and payment plans.
• Parkway Example Family Dental: named in 7 answers.

What the assistants say about these clinics that they do not say about you: specific services by name, opening hours including evenings and Saturdays, the suburb and nearby transport, how many reviews they have and what the reviews praise. Your clinic is usually described in one generic sentence, and in two answers with an old address.

The sources cited most across the 48 customer-style answers, by the number of answers citing them:

• Health directory: example-health-directory.example (29). Lists the clinic, with Suite 2 where the website says Suite 4, and a short description with no services.
• Review site: sample-reviews.example (22). The clinic has a profile with 9 reviews, the most recent about a year ago; the clinics recommended ahead of you have 40 to 120.
• Local news roundup: localnews.example (14), a "best dentists in Brisbane" article updated this year. The clinic is not in it.
• The clinics' own websites: riverbend-example-dental.example (18), northgate-sample-dentists.example (12), harbourline-demo-dental.example (8).
• Your own site: exampledental.example (3).

Entity consistency: the website, Google Business Profile and Facebook page agree on the name, phone and Suite 4 address. The health directory lists Suite 2, and two assistants quoted that address when asked about the clinic directly. Bing Places has no listing.

From the homepage check on the day of the audit:

• Structured data: a basic Organization block is present. There is no LocalBusiness or Dentist type, and no services, opening hours, address or links to profiles.
• AI crawler access: robots.txt allows every crawler checked, including the search crawlers for ChatGPT, Claude, Perplexity, Google and Bing.
• llms.txt: none.
• Sitemap: present.
• Homepage: loads over https, one H1, title and description present, readable without JavaScript, phone number on the page.

Nothing technical is stopping the assistants from reading the site. The gap is in what the site and the profiles say, not whether they can be read.

This audit is your baseline: 5 of 48 customer-style answers named Example Dental Clinic (10%, range about 5% to 22%), and your own website was cited in 3 of 48. To see whether the changes make a difference, run the same six questions, three times each, in the same four assistants about 8 weeks after you have made them, and compare the ranges.

• Watch the share of answers that name you, not a single answer. Answers vary from run to run.
• If the new range overlaps this one, treat the difference as within normal variation.
• Check whether the directory you corrected now shows the right address.
• Keep a dated note of each change, so you can see what came before any movement.
• Repeat the two Google AI Mode questions by hand on the same day.

This is an illustrative report. Example Dental Clinic, the competitors, the sources, the quotes and the figures are invented to show the format and depth of a real report. Nothing in it was measured, and none of the names belong to a real business. A real report follows exactly this structure with your business, the assistants' actual answers and real sources.

In a real audit we ask six questions, each three times in each of four assistants, with the search tool on and the location set to your area. ChatGPT, Gemini, Claude and Perplexity are asked through their search-enabled APIs, and the report states the model used for each and the date. These answers approximate what customers see in the apps but are not identical to them, and answers vary from run to run. Empty, cut-off or declined answers are left out of all counts and the report says how many.

The questions, with {category} filled as "dentist" and {city} as "Brisbane":

• Q1: best dentist in Brisbane
• Q2: who are the top dentist businesses near Brisbane, with a short reason for each
• Q3: I need a dentist in Brisbane, who would you recommend and why
• Q4: what do you know about Example Dental Clinic in Brisbane
• Q5: is Example Dental Clinic (exampledental.example) a good dentist in Brisbane
• Q6: dentist in Brisbane: what should I check before choosing one, and who would you suggest

Visibility is the share of customer-style answers (the questions a customer would ask when looking for this kind of business) in which your business was named, across all assistants and repeats. It is shown as a percentage with a 95% range, because answers vary from run to run and a count out of a few dozen answers is only an estimate. The range reflects that variation; it does not cover the choice of questions. Bands: Low under 10%, Developing 10 to 29%, Moderate 30 to 59%, Strong 60% and over. Position when listed, how often your own website was cited, and how accurate the answers about your business were are reported separately and are not part of the percentage. This measures what the assistants said on the day we checked. It is not a ranking.

How many runs named the business, out of the usable runs, for each question and assistant:

• Q1 (best dentist in Brisbane): ChatGPT 1/3, Gemini 0/3, Claude 0/3, Perplexity 1/3
• Q2 (top dentist businesses near Brisbane): ChatGPT 1/3, Gemini 1/3, Claude 0/3, Perplexity 0/3
• Q3 (I need a dentist in Brisbane): ChatGPT 0/3, Gemini 0/3, Claude 0/3, Perplexity 1/3
• Q4 (what do you know about Example Dental Clinic): ChatGPT 3/3, Gemini 3/3, Claude 3/3, Perplexity 3/3
• Q5 (is Example Dental Clinic a good dentist): ChatGPT 3/3, Gemini 3/3, Claude 3/3, Perplexity 3/3
• Q6 (what should I check, and who would you suggest): ChatGPT 0/3, Gemini 0/3, Claude 0/3, Perplexity 0/3

Google AI Mode is checked by hand for Q1 and Q3, once each, and reported in its own section.

In a real report this section also lists anything that could not be measured: a profile site that blocked reading, an assistant that could not be run, a check that covers the homepage only. The full text of every answer is kept and is available on request.