Back to Insights
Who Wrote the "Best AI Visibility Tools" List You Are Reading? A 2026 Audit of 4 Roundups
Chinmay Dhamapurkar
Chinmay Dhamapurkar
,
,
Data Scientist
Data Scientist
6 min read
6 min read
If you are a marketing director or agency lead shortlisting AI visibility vendors this quarter, you have almost certainly read a “best AI visibility tools” roundup in the last month. We pulled four of the most visible roundups in the category. In every single one, the publisher ranked itself first.
TL;DR
Four vendor-published roundups, four #1 finishes for the publisher. Profound, Evertune, Frase, and Kime each publish a “best AI visibility tools” list. Each list puts the publisher at the top.
None of the four discloses the conflict. Not in a banner, not in a footer, not in the byline. Two are written by the publisher’s own marketing staff. One is written by the CEO.
The rankings are not reproducible. Two roundups state no evaluation method at all. The other two list criteria without weighting, testing scope, or which competitors were actually used.
The engines are reading these pages. In our own August measurement of 123 AI-visibility buying queries, vendor-authored roundups and review-aggregator pages took a large share of every citation the engines produced.
A neutral roundup is possible. It takes six disclosures. None of the four make more than one.
What we found
We audited four roundups that surface repeatedly for AI-visibility buying queries. We looked at four things only: who publishes it, where the publisher ranks itself, whether the conflict is disclosed, and what method the ranking claims.
Publisher | Tools ranked | Publisher’s own rank | Author | Conflict disclosed? | Stated method |
|---|---|---|---|---|---|
Profound | 18 | 1 | Nick Lafferty, Founding Marketing Engineer | No | None stated |
Evertune | 10 | 1 | Not bylined | No | None stated |
Frase | 10 | 1 | Georgina D’Souza, Marketing Manager | No | Six criteria, no weighting or testing scope |
Kime | 11 | 1 | Vasilij Brandt, Founder and CEO | No | Seven criteria; hands-on “where we have access”, scope unstated |
Four out of four. The odds of that happening by chance, if the rankings were independent of who paid for the page, are about one in a few thousand depending on how you count the eligible field. They are not independent, and nobody involved thinks they are.
What is broken here
None of these pages is lying. Profound’s list is a genuinely useful map of 18 tools, and Kime’s seven criteria are reasonable criteria. The failure is structural: a ranked comparison published by a ranked competitor, with no disclosure and no reproducible method, is a positioning document formatted as a buyer’s guide. A marketer reading it in a shortlist meeting has no way to tell which of the two it is.
The part that should worry you more
A biased roundup you can spot is a minor problem. A biased roundup that the answer engines have absorbed as ground truth is a different one.
At the end of August we ran our own category as a measurement program. 123 distinct buyer queries about AI visibility tools, run out to 250 responses, producing 1,960 citations across 525 distinct pages. We did it to measure ourselves, which is a separate story. What the run also produced was a map of exactly which pages the engines lean on when a buyer asks this question.
Review aggregators and vendor-authored roundups own it. G2 review pages took an 11.4% share of all citations (95% CI 7.9–14.9). Frase’s domain took 4.0% (2.9–5.0), and its own ten-tool roundup was among the ten most-cited individual pages in the entire run. Capterra’s “how to choose” guide, a 42dm top-ten platforms page, Semrush’s measurement post, and Zapier’s “best AI visibility tool” page all landed in the cited set. Kime’s domain took 1.1% (0.6–1.6), on a thin sample we would not lean on hard.
So when a CMO asks ChatGPT which AI visibility tool to buy, a meaningful slice of the answer is assembled from pages written by the vendors being ranked. The conflict does not stay on the vendor’s blog. It gets laundered through the answer engine and arrives at the buyer with no byline attached at all.
What a neutral roundup would have to disclose
We are not neutral either. IQRush sells in this category, and this post is on our blog, which is exactly why we are not publishing our own ranked list of competitors. If we did, we would have to meet the following bar, and so should anyone else.
Publisher relationship, at the top. One line, above the first ranking: whether the publisher sells a competing product, and where it appears in the list.
Author and their employer. Not just a name. A marketing manager at the #1-ranked vendor is a relevant fact about the ranking.
What was actually tested. Which products the author used, for how long, on what account tier. “Hands-on where we have access” is not a scope statement.
The criteria and their weights. Six criteria with no weighting can produce any ordering the author wants. Publish the weights or drop the ordering.
What would change the ranking. A falsifiable condition. If vendor X shipped feature Y, would it move? If nothing would move it, it is not a ranking.
Date and re-review cadence. This category ships monthly. A ranking with no review date is a snapshot presented as a standing judgment.
That is the whole bar. No page we audited clears more than one of the six.
Marketer-Side Impact
Run the cost forward. A shortlist assembled from vendor-authored roundups produces a three-vendor demo cycle, roughly 20 hours of internal time across marketing and analytics, plus procurement and security review. Call it $12,000 to $18,000 of loaded cost before anyone signs anything. If the shortlist was shaped by four pages that each ranked their author first, you paid that to evaluate a field somebody else picked. The larger cost lands later: an annual contract, typically $30,000 to $120,000 in this category, against a tool whose measurement layer nobody in the room ever audited, because the roundup graded on engine coverage and pricing tiers rather than on whether the numbers can tell signal from noise. Then the second run does not reproduce the first, and you are explaining that to a board.
Three questions to take into your next shortlist meeting
“Who published the list this shortlist came from, and where did they rank themselves?” Ask it out loud in the room. It changes the conversation in about ten seconds.
“What would have to be true for this ranking to change?” Put it to the vendor whose page you read. A ranking nobody can move is marketing copy.
“Which of these vendors would still be on the shortlist if we built it from the IAB’s decision-grade criteria instead?” Sample size, query volume, prompt-type coverage, testing cadence, reproducibility, documented method. That is a shortlist somebody can defend.
Frequently asked questions
Are these vendors doing something wrong? Nothing unusual, and nothing hidden in a legal sense. Publishing a comparison that flatters your own product is standard content marketing across every software category. The specific problem here is that AI visibility is a measurement category, and a measurement vendor’s credibility rests on disclosure. Publishing an undisclosed self-ranking undercuts the thing they are selling.
Should I just use G2 or Capterra instead? They are better, but they are not neutral infrastructure. Both operate commercial relationships with the vendors listed, and both weight review volume, which favors whoever ran the biggest review campaign. Use them as one input. Do not use review counts as a proxy for measurement quality, because the two are unrelated.
Is IQRush on any of these lists? No. We did not appear in any of the four roundups, and we placed zero citations in our own August run of the category. We are writing this from outside the list, which you should factor in when reading it.
How do I build a shortlist that is not shaped by vendor content? Start from criteria, not from a list of names. Write down the six IAB decision-grade criteria, score every candidate against them in a live demo, and let the shortlist fall out of the scoring. It takes about an hour longer and produces a defensible answer.
What about the roundups written by consultants and agencies rather than vendors? Better on the conflict question, usually weaker on the method question. Most agency roundups still rank on engine coverage, pricing, and integrations, which are common to every vendor in the category and tell you nothing about measurement reliability. Apply the same six disclosures.
Chinmay Dhamapurkar works on IQRush’s measurement pipeline. IQRush builds measurement infrastructure for probabilistic AI search. Falsifiable version: we will re-audit these four pages in March 2027. If two or more have added a publisher-relationship disclosure above the first ranking, the category moved and we will say so here.
Published October 21, 2026. Roundups audited as of August 31, 2026.
Sources: Profound roundup · Evertune roundup · Frase roundup · Kime roundup
If you are a marketing director or agency lead shortlisting AI visibility vendors this quarter, you have almost certainly read a “best AI visibility tools” roundup in the last month. We pulled four of the most visible roundups in the category. In every single one, the publisher ranked itself first.
TL;DR
Four vendor-published roundups, four #1 finishes for the publisher. Profound, Evertune, Frase, and Kime each publish a “best AI visibility tools” list. Each list puts the publisher at the top.
None of the four discloses the conflict. Not in a banner, not in a footer, not in the byline. Two are written by the publisher’s own marketing staff. One is written by the CEO.
The rankings are not reproducible. Two roundups state no evaluation method at all. The other two list criteria without weighting, testing scope, or which competitors were actually used.
The engines are reading these pages. In our own August measurement of 123 AI-visibility buying queries, vendor-authored roundups and review-aggregator pages took a large share of every citation the engines produced.
A neutral roundup is possible. It takes six disclosures. None of the four make more than one.
What we found
We audited four roundups that surface repeatedly for AI-visibility buying queries. We looked at four things only: who publishes it, where the publisher ranks itself, whether the conflict is disclosed, and what method the ranking claims.
Publisher | Tools ranked | Publisher’s own rank | Author | Conflict disclosed? | Stated method |
|---|---|---|---|---|---|
Profound | 18 | 1 | Nick Lafferty, Founding Marketing Engineer | No | None stated |
Evertune | 10 | 1 | Not bylined | No | None stated |
Frase | 10 | 1 | Georgina D’Souza, Marketing Manager | No | Six criteria, no weighting or testing scope |
Kime | 11 | 1 | Vasilij Brandt, Founder and CEO | No | Seven criteria; hands-on “where we have access”, scope unstated |
Four out of four. The odds of that happening by chance, if the rankings were independent of who paid for the page, are about one in a few thousand depending on how you count the eligible field. They are not independent, and nobody involved thinks they are.
What is broken here
None of these pages is lying. Profound’s list is a genuinely useful map of 18 tools, and Kime’s seven criteria are reasonable criteria. The failure is structural: a ranked comparison published by a ranked competitor, with no disclosure and no reproducible method, is a positioning document formatted as a buyer’s guide. A marketer reading it in a shortlist meeting has no way to tell which of the two it is.
The part that should worry you more
A biased roundup you can spot is a minor problem. A biased roundup that the answer engines have absorbed as ground truth is a different one.
At the end of August we ran our own category as a measurement program. 123 distinct buyer queries about AI visibility tools, run out to 250 responses, producing 1,960 citations across 525 distinct pages. We did it to measure ourselves, which is a separate story. What the run also produced was a map of exactly which pages the engines lean on when a buyer asks this question.
Review aggregators and vendor-authored roundups own it. G2 review pages took an 11.4% share of all citations (95% CI 7.9–14.9). Frase’s domain took 4.0% (2.9–5.0), and its own ten-tool roundup was among the ten most-cited individual pages in the entire run. Capterra’s “how to choose” guide, a 42dm top-ten platforms page, Semrush’s measurement post, and Zapier’s “best AI visibility tool” page all landed in the cited set. Kime’s domain took 1.1% (0.6–1.6), on a thin sample we would not lean on hard.
So when a CMO asks ChatGPT which AI visibility tool to buy, a meaningful slice of the answer is assembled from pages written by the vendors being ranked. The conflict does not stay on the vendor’s blog. It gets laundered through the answer engine and arrives at the buyer with no byline attached at all.
What a neutral roundup would have to disclose
We are not neutral either. IQRush sells in this category, and this post is on our blog, which is exactly why we are not publishing our own ranked list of competitors. If we did, we would have to meet the following bar, and so should anyone else.
Publisher relationship, at the top. One line, above the first ranking: whether the publisher sells a competing product, and where it appears in the list.
Author and their employer. Not just a name. A marketing manager at the #1-ranked vendor is a relevant fact about the ranking.
What was actually tested. Which products the author used, for how long, on what account tier. “Hands-on where we have access” is not a scope statement.
The criteria and their weights. Six criteria with no weighting can produce any ordering the author wants. Publish the weights or drop the ordering.
What would change the ranking. A falsifiable condition. If vendor X shipped feature Y, would it move? If nothing would move it, it is not a ranking.
Date and re-review cadence. This category ships monthly. A ranking with no review date is a snapshot presented as a standing judgment.
That is the whole bar. No page we audited clears more than one of the six.
Marketer-Side Impact
Run the cost forward. A shortlist assembled from vendor-authored roundups produces a three-vendor demo cycle, roughly 20 hours of internal time across marketing and analytics, plus procurement and security review. Call it $12,000 to $18,000 of loaded cost before anyone signs anything. If the shortlist was shaped by four pages that each ranked their author first, you paid that to evaluate a field somebody else picked. The larger cost lands later: an annual contract, typically $30,000 to $120,000 in this category, against a tool whose measurement layer nobody in the room ever audited, because the roundup graded on engine coverage and pricing tiers rather than on whether the numbers can tell signal from noise. Then the second run does not reproduce the first, and you are explaining that to a board.
Three questions to take into your next shortlist meeting
“Who published the list this shortlist came from, and where did they rank themselves?” Ask it out loud in the room. It changes the conversation in about ten seconds.
“What would have to be true for this ranking to change?” Put it to the vendor whose page you read. A ranking nobody can move is marketing copy.
“Which of these vendors would still be on the shortlist if we built it from the IAB’s decision-grade criteria instead?” Sample size, query volume, prompt-type coverage, testing cadence, reproducibility, documented method. That is a shortlist somebody can defend.
Frequently asked questions
Are these vendors doing something wrong? Nothing unusual, and nothing hidden in a legal sense. Publishing a comparison that flatters your own product is standard content marketing across every software category. The specific problem here is that AI visibility is a measurement category, and a measurement vendor’s credibility rests on disclosure. Publishing an undisclosed self-ranking undercuts the thing they are selling.
Should I just use G2 or Capterra instead? They are better, but they are not neutral infrastructure. Both operate commercial relationships with the vendors listed, and both weight review volume, which favors whoever ran the biggest review campaign. Use them as one input. Do not use review counts as a proxy for measurement quality, because the two are unrelated.
Is IQRush on any of these lists? No. We did not appear in any of the four roundups, and we placed zero citations in our own August run of the category. We are writing this from outside the list, which you should factor in when reading it.
How do I build a shortlist that is not shaped by vendor content? Start from criteria, not from a list of names. Write down the six IAB decision-grade criteria, score every candidate against them in a live demo, and let the shortlist fall out of the scoring. It takes about an hour longer and produces a defensible answer.
What about the roundups written by consultants and agencies rather than vendors? Better on the conflict question, usually weaker on the method question. Most agency roundups still rank on engine coverage, pricing, and integrations, which are common to every vendor in the category and tell you nothing about measurement reliability. Apply the same six disclosures.
Chinmay Dhamapurkar works on IQRush’s measurement pipeline. IQRush builds measurement infrastructure for probabilistic AI search. Falsifiable version: we will re-audit these four pages in March 2027. If two or more have added a publisher-relationship disclosure above the first ranking, the category moved and we will say so here.
Published October 21, 2026. Roundups audited as of August 31, 2026.
Sources: Profound roundup · Evertune roundup · Frase roundup · Kime roundup
Frequently asked questions
Are vendors who rank themselves first in their own roundups doing something wrong?
Nothing unusual, and nothing hidden in a legal sense. Publishing a comparison that flatters your own product is standard content marketing across every software category. The specific problem here is that AI visibility is a measurement category, and a measurement vendor's credibility rests on disclosure. Publishing an undisclosed self-ranking undercuts the thing they are selling.
Should I just use G2 or Capterra to shortlist AI visibility tools instead?
They are better, but they are not neutral infrastructure. Both operate commercial relationships with the vendors listed, and both weight review volume, which favors whoever ran the biggest review campaign. Use them as one input. Do not use review counts as a proxy for measurement quality, because the two are unrelated.
Is IQRush on any of the best AI visibility tools lists?
No. We did not appear in any of the four roundups audited, and we placed zero citations in our own August run of the category. We are writing this from outside the list, which you should factor in when reading it.
How do I build an AI visibility vendor shortlist that is not shaped by vendor content?
Start from criteria, not from a list of names. Write down the six IAB decision-grade criteria (sample size, query volume, prompt-type coverage, testing cadence, reproducibility, documented method), score every candidate against them in a live demo, and let the shortlist fall out of the scoring. It takes about an hour longer and produces a defensible answer.
What about AI visibility roundups written by consultants and agencies rather than vendors?
Better on the conflict question, usually weaker on the method question. Most agency roundups still rank on engine coverage, pricing, and integrations, which are common to every vendor in the category and tell you nothing about measurement reliability. Apply the same six disclosures.