Sampled visibility snapshot · AI visibility & AEO monitoring
The state of AI-visibility answers, July 2026
We build an AEO scanner, so we can point it at our own category. On 2026-07-26 we probed live assistant-search lanes across five buyer intents a person actually types when they want to know whether AI assistants mention their brand — and recorded every domain that got cited.
The result is not the one the category's own marketing predicts. Nobody owns this answer. Five queries produced 57 citations spread across 46 different domains, and 83% of those domains were cited exactly once. AttractOS — the tool publishing this page — was cited zero times.
AttractOS competes with several products named on this page. That is a real conflict of interest and you should read the numbers with it in mind. Two things make it checkable rather than something you have to take on faith: the ranking is produced by the same probe code that runs every other benchmark on this site, with no per-category tuning; and it is reported as it came back, which is why the headline finding is that our biggest-name competitor was barely cited and we were not cited at all. Every citation is listed below so you can re-run any query yourself and disagree with us.
Probe date: 2026-07-26. Sample: 5 buyer-intent queries × 2 live assistant search lanes (a third was attempted and returned nothing — see below). This is a single, dated sample per lane — not a rank tracker. Assistant answers vary between runs and shift as models and their indexes change; treat this as a snapshot of one moment, not a repeatable ranking.
| Lane | Model ID | Source |
|---|---|---|
| OpenAI web-search API sample | openai/gpt-4o-mini-search-preview | via OpenRouter — no response this run |
| Perplexity Sonar API sample | perplexity/sonar | via OpenRouter |
| Gemini grounded API sample | gemini-2.5-flash | via Gemini API |
On the OpenAI lane: it produced no answers at all on this run. Every call returned model_not_found: OpenAI has deprecated the gpt-4o-mini-search-preview model this lane calls, and the request now fails upstream. We are reporting that rather than quietly presenting a two-lane run as a three-lane one. Everything below therefore aggregates the Perplexity Sonar and Gemini grounded lanes only — the same two lanes our other benchmarks rank on — and the true state of the OpenAI answer for these queries is unknown, not zero. The lane has since been rebuilt on openai/gpt-5.4-mini:online — an OpenAI model answering over web results retrieved by OpenRouter — and answers again. This page is left as the two-lane run it actually was rather than re-scored against a later probe. We resolve every cited link to its registrable domain (eTLD+1) before counting. Full method: /methodology.
Intents sampled
Five buyer-shaped questions, written the way someone with this problem actually asks it:
- "best ai visibility monitoring tools"
- "how do i track whether chatgpt recommends my brand"
- "tools to monitor brand mentions in ai search results"
- "how to track citations in perplexity and gemini for my website"
- "cheapest way to check if my website shows up in ai answers"
The headline: the category has no incumbent in the answer
| Measure | Result |
|---|---|
| Buyer intents probed | 5 |
| Live lanes (of 3 attempted) | 2 |
| Total citations recorded | 57 |
| Distinct domains cited | 46 |
| Domains cited exactly once | 38 (83%) |
| Domains cited on more than one intent | 7 |
| Domains cited by both lanes | 7 |
| Domains cited on all five intents | 1 |
For comparison, in our small-business CRM and AI-marketing-tools benchmarks a handful of names recur across most intents. Here, seven domains out of forty-six manage to appear twice. An 83% cited-once rate is what a category looks like when the assistants have no settled answer and are reaching for whatever page happens to match the phrasing of the question.
Top cited domains
Aggregated across the Perplexity Sonar and Gemini grounded lanes over all five intents. Citations counts each lane×intent citation, so a domain cited by both lanes on one intent counts twice; Intents is the number of distinct queries it appeared in. Every domain below the cut appears exactly once — the full list is in the per-intent evidence further down.
| Domain | Citations | Intents | What it is |
|---|---|---|---|
| otterly.ai | 5 | 5 | Purpose-built AI-visibility monitoring product |
| frizerly.com | 2 | 2 | Automated SEO-content SaaS; AEO is one feature |
| pixelmojo.io | 2 | 2 | Agency that also sells its own visibility tool |
| rankability.com | 2 | 2 | SEO suite for agencies, AI-visibility module |
| semrush.com | 2 | 2 | Incumbent SEO suite, AI-visibility toolkit |
| seranking.com | 2 | 2 | Incumbent SEO suite, AI-visibility tracker |
| siftly.ai | 2 | 2 | Purpose-built AI-visibility monitoring product |
| meltwater.com | 2 | 1 | Enterprise media monitoring, added AI tracking |
The "what it is" column comes from fetching each of these eight sites on 2026-07-26 and reading what they sell, not from the probe.
Reading 1 — otterly.ai is the only domain the assistants agree on
It is the sole domain cited on all five intents, and one of only seven cited by both lanes. Nothing else in the sample is close. If you were to name a single "who currently wins the AI answer for AI-visibility tools" from this run, it is Otterly — not on the strength of any one query, but on being the only name that survives every rephrasing of the question.
Reading 2 — the category's best-known name is nearly absent from its own category's answer
Profound (tryprofound.com) is the name most associated with enterprise AI-search monitoring, and the one we get asked about most. In this sample it was cited once: a single Gemini citation on the plain "best ai visibility monitoring tools" query, and nowhere else. It does not appear on the ChatGPT-tracking question, the brand-mentions question, the Perplexity/Gemini-citations question, or the price-shaped question.
The honest caveat matters here: this is one dated sample on five queries, and a low citation count is not a statement about product quality, company size, or revenue — Profound is a substantially bigger business than the citation count suggests, and our comparison page concedes the capabilities where it beats us. What the number says is narrower and still useful: brand prominence in this category is not currently converting into citation share in assistant answers.
Reading 3 — the engines are citing sellers, not referees
This is the finding we did not expect, and it is the one worth carrying away. We fetched all eight top-cited domains to check what each one actually sells. Zero of the eight are neutral publishers. Two are purpose-built AI-visibility products (otterly.ai, siftly.ai). Three are established search suites that now ship an AI-visibility module (semrush.com, seranking.com, rankability.com). One is an enterprise media-monitoring platform that added AI tracking (meltwater.com). One is an automated SEO-content SaaS whose own marketing is programmatically generated (frizerly.com). One is an agency that sells its own visibility audit tool alongside its services (pixelmojo.io).
In our AI-marketing-tools benchmark the single most-cited domain was an independent review site that out-cited every vendor. Here there is no equivalent. Extend the check past the top eight and the same shape holds: several of the cited "best AI visibility tools" roundups are published by companies selling a competing scanner, and they place their own product at or near the top of their own list — growbydata.com, omniseo.com, scopesite.co.uk and pixelmojo.io each rank themselves first or second in a roundup they wrote.
So when an assistant answers "which AI visibility tool should I use," it is very often relaying a ranking written by one of the contestants. That is worth knowing before you act on such an answer — including this page, which is also written by a contestant, and which is why we put the disclosure at the top and the full evidence at the bottom.
The practical implication for anyone selling here is the one our field survey argues from first principles: the route into these answers is not publishing a better page about yourself. It is being named on the third-party write-ups the assistants already reach for — and in this category, those write-ups are unusually scarce.
Reading 4 — the two lanes barely agree with each other
Perplexity Sonar produced 30 citations across 27 distinct domains; Gemini grounded produced 27 across 26. The overlap is seven domains. Two engines asked the same five questions on the same day returned almost entirely different sets of sources. Gemini also declined to cite anything at all on two of the five intents. Any tool — ours included — that reports a single "AI visibility score" is compressing a genuinely large per-engine disagreement into one number.
Reading 5 — we are not in our own category's answer
attractos.com was cited zero times across five intents and two lanes. We are publishing that because a benchmark that quietly omits its author's absence is an advertisement. It is also consistent with Reading 3: we have not been listed on the third-party roundups these answers are built from, and no amount of publishing on our own domain has changed that. We are working the same list this page implies, and we will re-run this probe and report whether it moved.
The pricing shape behind the cheap-first-check question
The fifth intent — "cheapest way to check if my website shows up in ai answers" — returned no tool vendor at all in the Perplexity lane's top results and filled with agency explainers. That is worth pairing with a second, non-probe check: on the same date we read the live public pricing pages of the three most prominent priced tools in this category.
| Tool | Entry plan (as printed 2026-07-26) | Shape |
|---|---|---|
| otterly.ai | Lite $29/mo · Standard $189/mo · Premium $489/mo | Subscription |
| brandgeo.co | Starter $79/mo · Business $149/mo · Agency $349/mo | Subscription |
| tryprofound.com | Starter $99/mo · Growth $399/mo (billed yearly) · Enterprise not publicly listed | Subscription |
Nine published tiers across the three, and not one of them is a one-time purchase. The cheapest way into any of them is $29/mo, or $348 a year, to answer a question most buyers ask once. We have an obvious interest in pointing this out — our own audit is a $79 one-time purchase — so treat it as a checkable fact rather than an argument: the prices above are printed on those vendors' own pricing pages and will change, and you should verify them there before buying anything.
Scan your own AI-visibility tool against these same intents — free
Run a free sampled visibility scan on your domain across the 5 buyer intents in this benchmark (add up to 5 of your own on the next screen). No account, no card — your report opens in the browser.
- ›best ai visibility monitoring tools
- ›how do i track whether chatgpt recommends my brand
- ›tools to monitor brand mentions in ai search results
- ›how to track citations in perplexity and gemini for my website
- ›cheapest way to check if my website shows up in ai answers
Per-intent raw evidence (Perplexity Sonar + Gemini grounded)
Every citation behind the counts above, in the order the lane returned them. The OpenAI web-search lane returned model_not_found on all five intents and contributed nothing — see the methodology block.
"best ai visibility monitoring tools"
- Perplexity: rankability.com, brainz.digital
- Gemini: wpengine.com, position.digital, frase.io, tryprofound.com, seranking.com, onelittleweb.com, llmclicks.ai, otterly.ai
- Note: the only intent in the run where Profound was cited, and the only one where a plain "best X" phrasing pulled a hosting platform (wpengine.com) into the answer.
"how do i track whether chatgpt recommends my brand"
- Perplexity: omniseo.com, gracker.ai, dageno.ai, otterly.ai, trysight.ai, visiblie.com, meltwater.com
- Gemini: keyword.com, rankability.com, grro.io, frizerly.com, ai-semantica.com, workduo.ai, siftly.ai, visualping.io, ahrefs.com, semrush.com, growbydata.com, meltwater.com
- Note: the densest intent in the sample (19 citations) and the only domain both lanes independently reached for was meltwater.com — an enterprise media-monitoring incumbent rather than an AEO tool.
"tools to monitor brand mentions in ai search results"
- Perplexity: builtin.com, otterly.ai, useomnia.com, wix.com, siftly.ai, beamtrace.com, gumloop.com, xfunnel.ai
- Gemini: no citations returned.
"how to track citations in perplexity and gemini for my website"
- Perplexity: pixelmojo.io, moburst.com, indexly.ai, otterly.ai, frizerly.com, dev.to
- Gemini: no citations returned.
- Note: the two intents Gemini declined on are the two phrased as "monitor/track this specific thing" rather than "best tools for X" — worth a larger sample, but the pattern is that the grounded lane engages with shopping-shaped queries and abstains from mechanism-shaped ones.
"cheapest way to check if my website shows up in ai answers"
- Perplexity: sitespeak.ai, semrush.com, seranking.com, pixelmojo.io, amivisibleonai.com, scopesite.co.uk, tilio.co.uk
- Gemini: artificialintelligencesummit.com, eseospace.com, openassistantgpt.io, squarespace.com, otterly.ai, freshmovemedia.com, bubblegumsearch.com
- Note: the price-shaped query is the one where the subscription platforms disappear almost entirely and the answers fill with agency explainers and site builders. No product in this sample owns the cheap-first-check question — which is the question a large share of this category's actual buyers are asking.
What we did not measure
- Whether these tools are any good. Citation share is not product quality. A page cited five times may be a thin listicle; a product cited zero times may be excellent. We measured what gets cited, nothing else.
- The OpenAI answer. One of the three lanes was unavailable, so a third of the intended sample is missing rather than negative.
- Anything over time. One date, one run per lane. Assistant answers move; the only honest way to know whether these positions hold is to probe again, which we will.
- Everyone in the category. This lists who got cited on five queries. Real products absent from the table are absent from these answers, which is not the same as not existing.
Related
- The ranked view of this same run — same evidence, "who wins the AI answer" format
- Best AI visibility monitoring tools — our honest survey of the field, competitors named
- AttractOS vs the field — head-to-heads with live-checked pricing and the rows we concede
- The Lab — the standing experiment where we probe our own domain weekly and log every result, win or lose
- All benchmarks
Probed 2026-07-26. AttractOS sells a product in the category measured on this page; the conflict-of-interest note above is part of the report, not a footnote to it. Single dated sample, two live lanes — see the methodology block for what that does and doesn't support.