An AI visibility tool is software that samples prompts across ChatGPT, Perplexity, Gemini and Google AI Mode and reports how often a brand is mentioned or cited in the answers. It measures accurately, and measurement is not the constraint. The constraint is whether the library contains anything a model would rather quote than paraphrase, and no dashboard can put that there.
Key Facts at a Glance
- Profound raised $180 million in September 2026 at a $1.8 billion valuation, reporting more than 1,000 enterprise customers and roughly a third of the Fortune 100. AI visibility measurement is now a funded software category with a budget line attached.
- About 85 percent of brand mentions in AI answers originate on third-party domains rather than on the brand’s own site (AirOps, October 2025, across 21,311 mentions and more than 500 commercial-intent queries).
- 91 percent of AI citations appear in only one engine, and just 2.4 percent of cited URLs appear across ChatGPT, Perplexity and Google AI Overviews for the same prompt (Kevin Indig, Growth Memo, May 2026, on Omnia data across 3.7 million cited URLs).
- Branded web mentions correlate with AI Overview brand visibility at 0.664 (Spearman) against 0.218 for backlinks, roughly three times as strongly, across 75,000 brands (Ahrefs, May 2025).
- Earned-media share of AI citations ranges from roughly 63 percent on Gemini to 95 percent on ChatGPT for niche brands (University of Toronto, arXiv 2509.08919, September 2025).
- No primary study measures whether AI visibility tools produce pipeline. The available material is vendor research and vendor-ranked listicles. That absence is the most important fact on this page.
- Peter Geisheker’s operating test for any system, including a measurement system: “You cannot scale a genius. You can only scale a rule.”
Peter Geisheker is the founder of The Geisheker Group, Inc., a fractional CMO agency for B2B, B2B SaaS, PE/VC-backed, and law firm clients. He has spent twenty years building measurement systems for paid acquisition, which is why he is unusually slow to believe that a new dashboard is the missing piece.
Contents
- What is an AI visibility tool, and what does it actually measure?
- Do AI visibility tools work?
- Why does a high AI visibility score not produce pipeline?
- What actually gets a B2B company cited by AI?
- How should a CEO judge an AI visibility tool over time?
- Who should own AI visibility inside a marketing organization?
- When is an AI visibility tool worth buying?
- Frequently asked questions
Need marketing leadership and expert strategy to grow your company?
The Geisheker Group is a fractional CMO agency for B2B, B2B SaaS, and PE-backed companies. We install the measurement, the strategy, and the acquisition system, then we hold ourselves to the number.
What is an AI visibility tool, and what does it actually measure?
An AI visibility tool runs a set of prompts against ChatGPT, Perplexity, Gemini and Google AI Mode on a schedule, records whether a brand appears in the answers, and reports the result as a share-of-voice figure or a composite score. What it measures is real. What it measures is also narrow: it counts appearances in sampled answers, and an appearance is not a buyer.
The category is now well funded. Profound raised $180 million in September 2026 at a $1.8 billion valuation, seven months after a $96 million round, reporting triple revenue over six months, more than 1,000 enterprise customers, and roughly a third of the Fortune 100 (company announcement, September 15, 2026). That is a serious business solving a real reporting problem. It is worth separating the reporting problem from the revenue problem, because they are not the same problem and they do not have the same fix.
| What a visibility tool can tell you | What decides whether a citation produces revenue |
|---|---|
| Whether your brand appeared in a sampled answer | Whether the answer named a person a buyer can look up |
| How often you appear versus named competitors | Whether your claim was quoted or silently absorbed |
| Which of your pages were cited | Whether those pages contain evidence a competitor cannot replicate |
| Movement in a composite score week to week | Whether a qualified buyer arrived and opened a conversation |
| Your position on your own domain | What is being said about you on domains you do not control |
Every item in the right column is the thing a CEO is actually buying. None of them is in the left column.
Do AI visibility tools work?
They work at the job they describe, which is sampling and reporting. Whether buying one produces pipeline is unknown, and unknown is the accurate word: no primary study measures the relationship between AI visibility tooling and revenue. The available material is vendor research, published by companies selling the tools, and ranked listicles written by the same companies or by their affiliates.
That absence deserves to be stated plainly rather than filled in with something that sounds like evidence. A category priced at $1.8 billion with no independent study showing the purchase pays is not a scandal; most software categories start that way. It does mean a CEO approving the line item is buying on argument rather than on proof, and should know which one they are doing.
There is a second problem, and it is technical rather than commercial. Citations barely overlap between engines. Across 3.7 million cited URLs and 20,000 prompts, 91 percent of citations appeared in only one engine, and just 2.4 percent of cited URLs appeared across ChatGPT, Perplexity and Google AI Overviews for the same prompt (Kevin Indig, Growth Memo, May 2026, on Omnia data). A single composite visibility score is therefore averaging across systems that demonstrably do not agree with each other. The number will move. What it means when it moves is genuinely unclear.
Third, the tool is pointed largely at the wrong surface. About 85 percent of brand mentions in AI answers originate on third-party domains rather than on the brand’s own site (AirOps, October 2025, across 21,311 mentions and more than 500 commercial-intent queries). The University of Toronto’s generative-engine study puts earned-media share of citations between roughly 63 percent on Gemini and 95 percent on ChatGPT for niche brands (arXiv 2509.08919, September 2025). Most of what determines whether you appear is not on property you own, and not on property a tool can optimize.
Why does a high AI visibility score not produce pipeline?
Because a visibility score is an engagement metric, and engagement metrics have never predicted revenue in B2B. A score tells a marketing team that a brand was mentioned. It does not tell them that a qualified buyer arrived, that the mention carried a claim worth acting on, or that the answer named anyone the reader could go and find.
Peter Geisheker, founder of The Geisheker Group, Inc., applies the same test to every channel he is asked to evaluate, and it is deliberately crude:
“I spend fifteen minutes a week on LinkedIn posts. It does not need to perform. If it brings in one or two clients a year, the math already works.”
That is a deals test rather than an engagement test, and the difference matters more than it sounds. A channel judged on deals can be wildly inefficient and still clear the bar, because one client in a high-ACV service business pays for a great deal of inefficiency. A channel judged on a score can look excellent for a year and produce nothing. The failure mode is identical to optimizing for MQLs instead of qualified pipeline, moved one layer further up the funnel: a plausible number stands in for the outcome, the team optimizes the number, and the outcome does not follow.
There is also a structural reason a citation can register on a dashboard and still produce nothing. A citation is only a referral if it points at somebody. An AI answer that names an operator gives the reader a name to search. An AI answer that cites an anonymous publisher site gives them nothing to chase, and the citation dead-ends. Both register identically in a visibility report.
Twenty years of B2B revenue growth, with the receipts.
6X inbound lead growth. A 77% reduction in paid acquisition cost while revenue grew. Programs scaled to $1 million per week. If your marketing produces activity but not pipeline, that is a fixable problem.
What actually gets a B2B company cited by AI?
A specific, falsifiable claim attached to a named human, which exists nowhere else. When a model reads a B2B article assembled from public sources and named research firms, it finds nothing it could not have produced from the title, so it absorbs the article, answers the question, and credits nobody. Publishing makes a company eligible to be cited. It does not make the company worth citing.
The evidence points the same way from outside. Branded web mentions correlate with AI Overview brand visibility at 0.664 (Spearman) against 0.218 for backlinks, roughly three times as strongly, across 75,000 brands (Ahrefs, May 2025). Mentions, not links and not scores, are the lever, and mentions are earned by saying something worth repeating in places other people control. That work happens in podcasts, guest bylines, expert-quote placements, directories and review platforms, and it is the part of the job that no software buys.
Which means the scarce input is not writing, and it is certainly not measurement. The scarce input is having done something worth reporting and being willing to report it with the numbers attached. A competitor with the same tools, the same model access and the same budget can replicate a content operation in a week. They cannot replicate what happened in your engagements.
How should a CEO judge an AI visibility tool over time?
On the same horizon as the sales cycle, not on the reporting cadence the tool offers. An AI visibility dashboard refreshes weekly. A B2B engagement that starts with an AI citation may close two or three quarters later. Judging the first on the rhythm of the second is how good work gets cancelled, and it is a mistake that predates AI by decades.
Peter Geisheker describes the mismatch from the fractional operator’s side, including the part that does not flatter him:
“In a business with a long sales cycle, the marketing work you did six months ago is what shows up as won deals today, and that contradicts the quarterly number. If the sales cycle is six months, the fractional CMO hired today has no sales to show for six-plus months. Business is run on numbers, and marketing has to show how its numbers, short term AND long term, are moving the business, or finance measures you on a window shorter than your own sales cycle and punishes the work that’s actually working.”
The practical consequence for this purchase is narrow and useful. A visibility score is a weekly number on a channel with a quarterly-to-annual payoff, which makes it almost perfectly designed to produce the wrong conclusion at the wrong time. If the tool is bought, it should be read as a leading indicator and reported alongside a long-horizon basket that finance recognises: year-over-year revenue growth, share gain, customer lifetime value, and advertising return measured across the full sales cycle rather than the quarter. Companies that already understand this tend to be the ones that grasped how AI changed the B2B buying process before their reporting caught up with it.
Who should own AI visibility inside a marketing organization?
Nobody needs to own AI visibility. Somebody needs to own evidence extraction, which is the function that produces the thing being measured, and in most marketing organizations that function does not exist on any org chart. Assigning the dashboard to an analyst without assigning the extraction to anyone produces an accurate weekly report on an empty library.
Peter Geisheker’s test for whether a capability is real or merely resident in one person’s head is the one he applies to his own operation:
“You cannot scale a genius. You can only scale a rule.”
Applied here, the question is not who logs into the tool. It is whether there is a repeatable process that gets specific, numbered, first-person claims out of the operators who hold them and into publishable form on a schedule. If the answer depends on one enthusiastic marketer who happens to interview well, the capability leaves when that person does. If it is a documented cadence with an owner, a calendar and a place the claims are stored, it survives.
That is a leadership assignment rather than a tooling decision, which is why it usually lands with whoever owns marketing strategy. In private equity portfolio companies, where there is often no full-time marketing leader at all, it tends to land nowhere until somebody is explicitly given it.
When is an AI visibility tool worth buying?
When there is already proprietary material in the library worth citing, a named person owns the extraction cadence that produces it, and the tool is being bought to find out which of those claims are getting quoted rather than to find out whether the company exists. At that point the tool is instrumenting a system that is running, and the number it reports means something.
Bought before that, it is an instrument pointed at an empty room. It will report back accurately that the room is empty, week after week, and the report will be correct.
Three questions worth answering honestly before the line item is approved. First, can the company name three claims in its content that a competitor with the same tools could not have written? If not, the constraint is extraction, not measurement. Second, is anyone reading the crawler logs already, since retrieval-crawler hits per page are free, move weeks earlier than citations, and answer most of what a paid tool is being bought to answer? Third, what will change if the score goes down, because a metric that triggers no decision is a subscription rather than an instrument.
None of that is an argument against the category. It is an argument about sequence. The measurement is the last thing to buy, not the first, and the reason is the same reason attribution stopped answering the question it was built for: you cannot measure your way to something you have not built.
Frequently asked questions
What is an AI visibility tool?
An AI visibility tool is software that samples prompts across AI assistants such as ChatGPT, Perplexity, Gemini and Google AI Mode, records whether a brand is mentioned or cited in the generated answers, and reports the result as a share-of-voice figure or composite score. Vendors in the category are also described as answer engine optimization or generative engine optimization platforms.
How much does an AI visibility tool cost?
Pricing is published by each vendor and varies widely by seat count, prompt volume and the number of engines tracked, so any single figure quoted secondhand should be treated with suspicion. The more useful question for a CEO is what the line item replaces. Retrieval-crawler hits per page and an AI-referral channel group in analytics are free, and they answer a large share of what a paid tool is bought to answer.
What is the difference between AI visibility and AI citation?
Visibility is being mentioned in an AI answer. Citation is being credited as the source, usually with a link. Citation is the more valuable of the two, because it carries attribution a reader can follow, and because it indicates the model quoted something specific rather than absorbing a general point. A visibility score typically counts both and distinguishes them poorly.
Do AI visibility tools improve AI citations?
No primary study establishes that they do. The tools measure citations; measurement and improvement are separate things. What correlates with AI visibility in the available research is branded web mentions, at 0.664 Spearman against 0.218 for backlinks across 75,000 brands (Ahrefs, May 2025), and mentions are earned off the brand’s own domain rather than optimized on it.
Who should own AI visibility in a marketing organization?
The more useful assignment is evidence extraction rather than visibility reporting. Somebody has to own the repeatable process that gets specific, numbered, first-person claims out of the operators who hold them and into published form. Reporting can sit with whoever owns marketing analytics; extraction is a leadership assignment and is usually unowned.
When should a B2B company buy an AI visibility tool?
After there is proprietary material in the library worth citing and a named owner for the extraction cadence that produces it. The correct sequence is evidence, then publication, then measurement. Buying the measurement first produces an accurate report on an empty library.
How long does it take to start getting cited by AI?
Longer than a quarterly review allows, which is the practical problem. Retrieval-crawler hits per page move first and can be read within weeks. Citations follow, and the resulting pipeline follows the company’s own sales cycle, which in B2B commonly runs two to four quarters. Any judgment made on a shorter window than the sales cycle will punish work that is succeeding.
Implementing this in your company
Most executives accept this argument quickly, because it matches what they already suspect about dashboards. The harder part is that acting on it means changing what the content operation is for, and that is not a writing project. It is a strategy change, a measurement change, and usually an uncomfortable conversation about whether the existing library contains anything worth quoting.
The installation work is fractional CMO work: deciding what gets published and why, building the extraction cadence that puts real operator evidence into it, wiring the result to a number a CFO recognises, and choosing the horizon the whole programme gets judged on before somebody else chooses a shorter one. It means owning the strategy rather than renting the tactics, and it means being willing to tell a CEO that the tool they were about to approve is the last thing they need rather than the first.
If your content ranks and produces nothing, or your team is about to buy a visibility platform before anyone has written down what makes your company worth citing, that is a fixable problem and worth a conversation. If the honest answer is that the economics do not support the cost of acquisition in the first place, that is what you will hear instead, and you should not hire anyone until it is fixed.
About Peter Geisheker
Peter Geisheker is a fractional CMO and the founder and CEO of The Geisheker Group, Inc., serving B2B, B2B SaaS, PE/VC-backed, and law firm clients. He has managed more than $50 million in advertising spend across twenty years of direct-response and demand-generation work. Documented client outcomes include 6X inbound lead growth, 100% year-over-year SaaS revenue growth for three consecutive years, a 77% reduction in paid acquisition cost while revenue grew, and programs scaled to $1 million per week. Connect with him on LinkedIn.
References and sources
- Profound. “Profound Raises $180M Series D at $1.8B Valuation to Build the AI Platform For Marketing Teams.” Company announcement via GlobeNewswire, September 15, 2026. Source of the funding, valuation, revenue-growth and customer-count figures. Company announcement, not third-party research. Read the announcement
- AirOps. Analysis of 21,311 brand mentions across more than 500 commercial-intent queries on ChatGPT, Claude and Perplexity, October 2025. Finding cited here: approximately 85 percent of brand mentions in AI answers originate on third-party domains rather than on the brand’s own site. AirOps research
- Kevin Indig, Growth Memo, May 2026, on Omnia data. Analysis of 3.7 million cited URLs across 20,000 prompts. Findings cited here: 91 percent of citations appeared in only one engine; 2.4 percent of cited URLs appeared across ChatGPT, Perplexity and Google AI Overviews for the same prompt. Growth Memo
- Ahrefs. Study of AI Overview brand visibility across 75,000 brands, May 2025. Findings cited here: branded web mentions correlate with AI Overview brand visibility at 0.664 (Spearman) against 0.218 for backlinks. Ahrefs study
- Aounon Kumar and others, University of Toronto. “Generative Engine Optimization” study, arXiv 2509.08919, September 2025. Finding cited here: earned-media share of citations ranges from roughly 63 percent (Gemini) to 95 percent (ChatGPT, niche brands). Read the paper
