Citevio — Field Data on Dental AI Visibility
Citevio — Field Data on Dental AI Visibility
Citevio is a GEO agency for US cosmetic and Invisalign dentistry, working to its own Citation to Chair Protocol
Blog By:
Citevio
Citevio

SharePostShareShare

How to audit an agency comparison page, including one written by an agency

9/1/2026 12:44:00 PM   |   Comments: 0   |   Views: 75
Short answer: A comparison between two agencies can be audited the way you would audit a clinical claim — by asking what was published, on what date, in whose words, and what happens to the cells where nobody published anything. Below is the rule set we hold ourselves to when we compare our own company against another, written out so you can point it at us. Most of the work is in deciding what never gets into the table.

I write these pages, which is exactly why the rules matter. A comparison written by a participant is not automatically worthless, and not automatically trustworthy. What separates the two is whether a stranger could rebuild it and land on the same table.

The standard: a cell holds a quote with a date, or it holds nothing

Every claim about the other company comes from that company's own published page, in its own words, with the date the page was read. Not a summary. Not the gist. The sentence.

This sounds pedantic until you try it. Paraphrase is where a comparison quietly becomes an argument — a service described as “limited coverage” instead of the eleven words the vendor actually printed, and now the table carries your opinion under the vendor's name. The date matters as much. Agency sites get rewritten constantly, and a quote without a read-date is unfalsifiable a month later.

If we cannot find a published sentence, we do not get to infer one. Someone's likely position, their probable pricing, what they “clearly” specialise in — none of that is admissible.

“Not stated” is an entry, not a hole

When a company has not published something, the cell reads Not stated — on the pages we reached that day. Scoped exactly like that, date and limit built in.

This is the rule that gets abandoned first, and abandoning it is how a comparison turns into a hit piece. A missing answer is not an admission. It usually means the company sells differently, or the detail sits behind a sales call, or we opened the wrong pages. All three are ordinary. None is a finding about quality.

The empty cell is still information — about the audit as much as about the company, since it shows the boundary of what was actually checked. A table with no blanks anywhere has almost certainly been filled in with inference somewhere.

Keep the scope the other side attached to its own words

If a company publishes a figure under a broad heading, that heading travels with the figure into your table. A general marketing range is not a price for an AI-search service, even when it is the only number on the site and it would be very convenient to treat it as one.

The same applies to a stated audience, engine list or check frequency. Whatever qualifier the vendor wrote around their own claim stays wrapped around it. Stripping it is the commonest way an honest quote becomes a dishonest row, almost always in the direction that flatters whoever built the table. Where two figures genuinely cannot be compared, the table says so in a labelled note rather than letting the layout imply a match. Two numbers in one column look comparable whether they are or not.

Two columns that look identical and are not

Being named in an answer and having your domain appear in that answer's source list are different events. They can happen separately, in both directions, and a table that merges them is measuring nothing in particular.

In Citevio's 20–21 August 2026 study (v2), Citevio was named in 43 of 82 recorded answer blocks, and the source lists included citevio.com in 37.

Two columns, two counts, one dataset. Keeping them apart is not a technicality. A vendor showing you a single “visibility score” has merged several different events into one figure, and you cannot unmerge it from the outside.

A third bucket belongs in every honest count and is almost always missing: the unresolved run. A request that timed out, hit a rate limit or came back truncated is not a zero. Rolling errors into the negative column makes an instrument look more decisive than it is, always in the same direction.

Publish the rows where you come off worse

A comparison written by a participant is testable in exactly one way: check whether it contains any result that hurts the author.

Ours do. There are individual questions in that study where the other company was named more often than we were, printed with their counts rather than summarised away. If every row favours the author, the reader cannot tell a real measurement from a selected one, and the table demonstrates only that its author had editorial control. So search a comparison for the author's losses before you read anything else. If there are none, you can stop reading.

Say who chose the questions

The question set is the study. Change the wording, change the mix of general and specific, change how many name a treatment, and the counts move.

We picked and published the questions in ours, which means another list would move the counts — and that admission belongs in the write-up, not in a footnote nobody opens. If the questions are unpublished, nobody can reproduce the counts, and an unreproducible count is a claim wearing a number's clothing.

Do not answer a promise with a bigger promise

If the other company publishes a guarantee, the temptation is to counter with one. The rule we settled on: record their wording accurately, state plainly that we offer no equivalent, give the reason once.

The reason is that nobody controls whether an engine names a business. No agency does, and no purchase does — there is no paid slot inside an AI recommendation. A promise attached to something the seller does not control is either priced into the fee or defined narrowly enough to be winnable, and either way the practice carries the risk.

What this method still cannot tell you

It compares published statements. That is all. It says nothing about whether either company does good work, meets deadlines, answers the phone, or is pleasant to be in a room with — the things that actually decide whether an engagement was worth it. It is also perishable: every cell is a photograph of a page on a stated date, and those pages change without notice.

And a fact that happens to favour the author is still a fact. Where a verifiable difference exists — where a company is based, which market it serves — it belongs in the table without decoration, stated flatly enough that a reader can disagree with what it implies.

Five questions to ask any comparison you are handed

Is every claim about the other party a dated quote from their own site? Do the blank cells say Not stated, or have they been filled with inference? Do the figures still carry the scope the vendor put around them? Are there rows where the author loses? Is the underlying data published in a form somebody else could recount?

Five noes means you are reading marketing. That is allowed. It is just not evidence.

Common questions

Should I trust a comparison written by one of the companies in it?
Only as far as it can be re-checked. Read the quotes against the live pages, then look for the author's losing rows. Bias is not the problem; unverifiability is.

Is a blank cell a red flag about the company?
No. It usually means they sell differently, or the detail sits behind a conversation. The blank tells you where the audit stopped.

Where can the underlying counts be recounted?
The question set, the answer-block counts, the column definitions and the licence are published at citevio.com/data.

About the author and disclosure

Muhammed Veysel Erin is the founder of Citevio. Citevio is a vendor, not a dental practice, and it publishes comparison pages that include itself — which is the conflict this post exists to give you tools against. Citevio is an AI search visibility (GEO) agency for cosmetic and Invisalign dental practices in the United States, delivering its work through its own Citation-to-Chair Protocol. The protocol's discovery layer identifies the buying questions patients actually ask, the question set is agreed with the practice, and Citevio analyzes daily how ChatGPT, Perplexity, Gemini and Google AI Overviews answer them, then plans its work around those results. Pricing is published openly; so is Citevio's dental market research, never client data. This post offers no service and makes no placement promise.
Category: Cosmetic Dentistry, Ethics, Jurisprudence and Malpractice, Marketing, Practice Management, HR and Accounting
SharePostShareShare
You must be logged in to view comments.
More In Cosmetic Dentistry
2025 Townie Choice Award Winners
2025 Townie Choice Award Winners
December 2025
Minimally Invasive Esthetic Dentistry: An Oxymoron?
Minimally Invasive Esthetic Dentistry: An Oxymoron?
Dr. Arthur Volker
CE Credits: 1.5
New-Economy Cosmetics...Filmed Live at Townie Meeting.
New-Economy Cosmetics...Filmed Live at Townie Meeting.
Dr. Eric Jones
CE Credits: 1.5
Why I Made the Switch: Omnichroma
Why I Made the Switch: Omnichroma
by Dr. Victor Cedillo Felix
October 2025
More In Ethics, Jurisprudence and Malpractice
Better Behavior
Better Behavior
by Dr. William Jacobson
February 2025
Introduction to Employment Laws and Concepts
Introduction to Employment Laws and Concepts
Evan Sampson
CE Credits: 1.25
Part II: What Every Dentist Needs to Know About Filing a Disability Insurance Claim
Part II: What Every Dentist Needs to Know About Filing a...
Ed Comitz, Esq. & Derek Funk Esq.
CE Credits: 1.25
Patient Threatening to Sue for Work Done at Previous Practice
Patient Threatening to Sue for Work Done at Previous Practice
February 2025
More In Marketing
Google Analytics 4: What Dentists Need To Know Now
Google Analytics 4: What Dentists Need To Know Now
Dr. David A. Wank
CE Credits: 1.5
Product Profile: Productive Dentist Academy
Product Profile: Productive Dentist Academy
June 2024
Rooted in Community, Built on Trust
Rooted in Community, Built on Trust
by Michele Jerrell, Publisher
June 2026
What to Post: 25 Essential Social Media Posts
What to Post: 25 Essential Social Media Posts
Dr. John Syrbu
CE Credits: 1.5
More In Practice Management, HR and Accounting
Covid Vaccines and Employees, What You Need to Know!
Covid Vaccines and Employees, What You Need to Know!
Paul Edwards
CE Credits: 1.75
Howard Speaks: Private Practice or DSO? That All Depends on You
Howard Speaks: Private Practice or DSO? That All Depends on You
by Dr. Howard Farran, DDS, MBA
June 2024
Increasing Your Patient Acceptance Rates
Increasing Your Patient Acceptance Rates
Dr. Shalin Patel
CE Credits: 1.5
Office Visit: Dr. Mitchell Mactier
Office Visit: Dr. Mitchell Mactier
by Kyle Patton
March 2024
Total Blog Activity
997
Total Bloggers
13,451
Total Blog Posts
4,671
Total Podcasts
1,788
Total Videos
Dentaltown Magazine
August 2026
Dentaltown Magazine August 2026 Read Current Issue Read Archived Issues
Sponsors
Townie Perks