The short answerAI visibility is how often an AI engine names your brand when people ask about your category, usually reported as the share of tracked answers that mention you. Read it per engine: on the same 144 buyer questions over 28 days, ChatGPT named LoudFace in 36.5% of answers, Google AI Overviews in 15.2% and Perplexity in 13.7%.
The metrics that make up AI visibility, and how each one is calculated:
| Metric | What it measures | How to calculate it | Read per engine? | Common trap |
|---|---|---|---|---|
| Visibility (mention rate) | How often answers name your brand | Answers that name you ÷ all tracked answers × 100 | Yes | Tools divide by different totals, so scores from two tools don't match |
| Citation rate | How often answers link to your pages as a source | Answers that cite your domain ÷ all tracked answers × 100 | Yes | A cited page does not mean a named brand |
| Share of voice | Your mentions against every brand's mentions | Your brand mentions ÷ all brand mentions × 100 | Yes | It comes out far lower than visibility and moves when you add competitors |
| Average position | Where you sit in the list when you are named | Average of your rank across answers that name you | Yes | A brand named once, first, looks better than one named often in third place |
| Sentiment | How favourably answers describe you | Each tool's own scale (0 to 100, or share of positive mentions) | Yes | Scales differ by tool and can't be compared |
Formulas follow the published definitions in Peec AI's and Profound's documentation. Our engine numbers come from LoudFace's own tracking in Peec AI, 27 September to 24 October 2026 (method at the end).
What is AI visibility?
AI visibility is the presence of your brand inside the answers that ChatGPT, Perplexity, Google AI Overviews, AI Mode, Gemini and Copilot write. When a buyer asks one of them which tools or agencies to consider, the engine writes a short answer and names a few brands. You are in that answer, or you are not.
The measurement vendors define it in almost the same words. Peec AI's documentation says "The Visibility Score measures the percentage of AI responses that mention your brand." Semrush says "AI Visibility measures how often and how prominently a brand appears in AI-generated answers." The idea is settled. The arithmetic behind each tool's number is not.
There are two kinds of presence, and you need both. Conductor's documentation draws the line clearly. A mention "is when Conductor identifies one of your brands as appearing in the text of an AI search result." A citation is "a link to your website among the sources an AI search engine used to create an AI search result." A buyer reads the mention. The citation sits in a sources panel most people never open.
So a brand can sit in one of four states on any question:
- Named and cited. The answer recommends you and links to your page.
- Named without a citation. The answer recommends you on the strength of other people's pages.
- Cited without a name. The engine used your page as a source and recommended someone else.
- Absent. Neither.
In our experience, the third state is the one teams watch least.
How is AI visibility measured?
With a fixed set of questions, asked again and again. Every serious tracker works the same way at the core:
- Pick the questions. The prompts a buyer in your category actually types: "best X for Y", "alternatives to Z", "how much does X cost".
- Run them on each engine, repeatedly. Peec AI runs prompts daily "since AI responses naturally vary day to day."
- Read every answer. Record which brands it names, in what order, and which pages it cites.
- Count. Divide, per engine and per question, then trend it over weeks.
The definitions differ, and that is why two tools can give one brand two different scores:
| Tool | What its visibility number is based on |
|---|---|
| Peec AI | All tracked responses |
| Profound | Only responses that mention at least one brand |
| Semrush | Not a raw percentage: a 0 to 100 score against the median mentions of your top competitors |
| Ahrefs Brand Radar | Questions drawn from People Also Ask and Ahrefs' keyword data, updated and tested monthly on a 90-day reporting window |
| HubSpot AI Search Grader | A one-off score out of 100, from what the models say about you based on their training data |
Profound's definition is "The percentage of responses that mention your brand, out of all responses that mention at least one brand." Answers that name nobody drop out of the total, so the same raw data reads higher there than in Peec AI. Semrush's score "reflects how often your brand is mentioned in AI-generated answers compared to the median number of mentions for your top industry competitors." Ahrefs says plainly that its metrics "show potential visibility, not actual audience reach."
None of these is wrong. They answer different questions. What goes wrong is a team comparing a Semrush score from one quarter with a Peec AI percentage from the next and calling it progress. Pick one tool and one definition, write both down, and keep them.
Share of voice causes the most confusion. Peec AI defines it as "the percentage of your brand mentions in AI responses compared with all brands mentioned, including yours." Its own worked example: a brand named in 4 of 10 chats has 40% visibility, but 4 mentions against a competitor's 12 gives 25% share of voice. Conductor warns that "your share of voice will tend to be much lower than your percentage of prompts with mentions." Our engine readings show the gap: 36.5% visibility on ChatGPT, 10.5% share of voice. Both numbers describe the same brand and the same answers. If someone reports one of them to you, ask which.
The question set moves the number too. Six of the 150 questions we track contain our brand name. With them in, every engine's reading rises by 1.2 to 2.0 points. Without them, you see how often engines name you when the buyer didn't already ask about you, which is the reading that matters.
Why one AI visibility score hides what is happening
Because each engine reads different sources and names different brands. Add them into one figure and you lose the engine that is losing.
Take one brand, LoudFace, on one set of 144 non-branded buyer questions, over the same 28 days:
| 27 Sep to 24 Oct 2026, 144 questions | ChatGPT | Google AI Overviews | Perplexity |
|---|---|---|---|
| Visibility (share of answers naming LoudFace) | 36.5% | 15.2% | 13.7% |
| Share of voice | 10.5% | 5.5% | 6.3% |
| Average position when named | 3.4 | 2.5 | 2.5 |
ChatGPT named us about 2.7 times as often as Perplexity on identical questions. Perplexity and AI Overviews named us less often but placed us higher when they did.
The week-to-week picture is sharper. On our 39 main buyer questions, in the week to 24 October 2026, Google AI Overviews named LoudFace in 45 of 260 answers (17.3%), up from 26 of 269 (9.7%) the week before. ChatGPT went the other way: 63 of 273 answers (23.1%), down from 78 of 273 (28.6%). Perplexity moved from 10 of 273 (3.7%) to 12 of 273 (4.4%). One engine rose 7.6 points while another fell 5.5. A single combined score would have shown a quiet week.
Independent research points the same way:
- The engines cite different websites. Ahrefs looked at the top 50 most-mentioned sites in AI Overviews, ChatGPT and Perplexity and found that "only 7 websites are in the top 50 results for all of them."
- They agree more on brands than on sources. Semrush found that "61 of the top 100 brands were common in both ChatGPT and Google AI Mode," and its 2026 index found that AI Overviews and ChatGPT had "the lowest source overlap of any two platforms."
- Semrush itself refuses to blend them. Its AI Visibility Index methodology reads: "Results are shown separately by platform, with no combined visibility score."
- Even Google's two AI surfaces differ. Google's documentation says "AI Mode and AI Overviews may use different models and techniques, so the set of responses and links they show will vary."
In our own study of 987 same-day answer sets on B2B agency questions, 500 had all three engines naming an agency, and in those 500, 71% of the agencies named appeared in only one engine's answer. We wrote that up in do ChatGPT, Perplexity and Google AI Overviews recommend the same agencies?
A single question shows it. On 18 October 2026 we asked each engine "Our brand is invisible in Perplexity and ChatGPT. Who can help a B2B SaaS fix that?" ChatGPT read 16 sources and named four agencies. Perplexity read 10 and named one. Google AI Overviews read 14 and named two others. Each engine came back with its own shortlist.
Cited is not the same as named
This is the gap that makes AI visibility harder than SEO. In Google search, if your page ranks, your brand is on the screen. In an AI answer, your page can be the engine's main source while the answer recommends someone else.
Peec AI's documentation puts it in one line: "You can be visible as a source without being mentioned as a brand." Their reports split the two on purpose, and so should yours.
One ChatGPT answer we tracked on 28 September 2026 cited an agency list six times, in first position, and named eight agencies. The list's publisher was not among them. It happens often enough to track as its own state: a brand can be the top-cited source and still not be named.
The reverse rule holds too. In our tracked week of 17 to 23 September 2026, 383 of the 387 ChatGPT answers that named LoudFace had read a loudface.co page. ChatGPT recommended the brand it had just read about. The breakdown for each engine is in AI engines named us almost only when they had read our page.
Two practical consequences:
- Track named and cited separately. A citation count going up while the named rate stays flat means your pages are useful and your brand is not attached to the answer.
- Fix what the cited page says about you. If your own page is the source and you are still not named, the page is not making the case for you in a form the engine can lift. A short, direct answer near the top that names what you do and a result you delivered helps.
How much do AI answers change from one ask to the next?
A lot, so only a percentage tracked across many runs gives you a usable number.
SparkToro and Gumshoe had 600 volunteers run 12 prompts 2,961 times across ChatGPT, Claude and Google AI Overviews and AI Mode. Their finding: "there's a <1 in 100 chance that ChatGPT or Google's AI, if asked 100X, will give you the same list of brands in any two responses." On order, "it's more like 1 in 1,000 runs before you'd see two lists in the same order."
Presence holds up much better than the lists do. In the same study, one brand appeared in 85 of 95 Google AI responses to a single prompt. Rand Fishkin's conclusion is the right one: "visibility % across dozens to hundreds of prompts run multiple times is a reasonable metric." He calls tracking a ranking position in AI tools foolhardy.
Google's AI Overviews churn too. Ahrefs tracked more than 43,000 keywords and found "AI Overviews have a 70% chance of changing from one observation to the next," with "nearly half (45.5%) of the cited sources are entirely new."
Our own reading agrees. With identical prompt text, ChatGPT named a different set of agencies on 474 of 488 consecutive-day pairs where both days named one. The breakdown is in does ChatGPT give the same answer to everyone?
So when someone shows you a screenshot of ChatGPT recommending them, treat it as one roll of the dice. Ask for the rate across many runs, per engine.
AI visibility vs SEO: what changes
The two are tracked with different units and different tools:
| SEO | AI visibility | |
|---|---|---|
| What you win | A position on a results page | A mention inside a written answer |
| Unit you track | Keywords | Questions (prompts), run repeatedly |
| Headline number | Rank, clicks, impressions | Share of answers that name you, per engine |
| Stability | Positions move over days and weeks | Lists change on most repeat asks |
| Where the data lives | Search Console | An AI visibility tracker; Search Console does not separate AI features |
That last row matters if you report to a board. Google says sites appearing in AI features "are included in the overall search traffic in Search Console," inside the normal Web search type, and that "all links in the AI Overview are assigned that same position." You cannot pull AI Overview clicks out of Search Console on their own.
Clicks are also shrinking where AI answers appear. Pew Research Center's browsing data from 900 U.S. adults found that "Users who encountered an AI summary clicked on a traditional search result link in 8% of all visits," against 15% when no summary appeared. Ahrefs measured that "the presence of an AI Overview now correlates with a 58% lower average clickthrough rate for the top-ranking page." Seer Interactive's data shows a partial recovery since: organic click-through on queries with an AI Overview fell to 1.3% in December 2025 and came back to 2.4% in February 2026. Seer also found that being cited "delivers +120% more organic clicks per impression versus when you are not cited."
Google's own position is that you don't need anything special. Its documentation says "There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary," and that a page must be "indexed and eligible to be shown in Google Search with a snippet." Good SEO gets you into the pool. What gets you named is a different job, and our AEO vs GEO vs SEO breakdown covers where the two split.
Why AI visibility matters for a B2B SaaS company
Because buyers already take software recommendations from AI chatbots.
Forrester reports that "89% of business buyers report using AI in their buying process" in its 2024 survey, and that the figure grew "to 94%" in its 2025 survey. G2's 2026 Buyer Behavior Report found that "Eighty-two percent of buyers sourced software recommendations from an AI chatbot in the last 24 months." G2 also found that review sites (38%) narrowly overtook AI chatbots (37%) as the top source shaping a buyer's shortlist, "for the first time."
AI chat has not replaced everything. Forrester says answer engines "often deliver incomplete or unreliable information, creating mistrust," and that buyers check what they hear with people. But the shortlist increasingly forms before anyone visits your site. If the engine doesn't name you, the buyer may never type your URL.
That is why AI visibility belongs next to pipeline in your reporting, not in place of it. A visibility gain that never shows up in demos booked is a vanity number. Ahrefs makes the same point about its own share-of-voice metric: it "doesn't tell you whether mentions are positive, whether AI actually recommends you, whether you get traffic, or whether that traffic converts."
What drives AI visibility?
Three things, in the order we would fix them.
1. Being readable by the engines. Each engine has its own crawler or search partner, and a blocked one keeps you out entirely. OpenAI says "Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers." Google needs the page "indexed and eligible to be shown in Google Search with a snippet." Check your robots.txt and firewall before anything else.
2. Being written about elsewhere. Ahrefs studied 75,000 brands and found "Brand web mentions show the strongest correlation (0.664) with AI Overview brand visibility," far stronger than backlinks (0.218). Its follow-up found "YouTube mentions show the strongest correlation with AI visibility (~0.737)." These are correlations, not proof of cause. They fit what we see: engines name the brands the pages they read are talking about.
3. Being the page the engine reads for the question. If your page is in the sources and states plainly what you do, for whom, with a named result, the engine has something to lift. If it isn't, you depend on someone else's list. We go deeper on this in how to get your B2B SaaS recommended in ChatGPT.
How do you check your AI visibility?
Use a fixed question set and read it every week.
- Write 30 to 50 questions your buyers ask. Category ("best X for Y"), comparison ("X vs Y", "alternatives to X") and problem ("why doesn't ChatGPT mention us"). Keep branded questions in a separate group.
- Track them daily on each engine. ChatGPT, Perplexity and Google AI Overviews at a minimum. Add AI Mode, Gemini or Copilot if your buyers use them.
- Report per engine. Give each engine its own row for every metric. Never one blended figure.
- Split named from cited. The gap between them is your fastest win.
- Add three to five competitors. Share of voice only means something against a fixed set.
- Read the answers, not just the scores. The words an engine uses about you are the words buyers will repeat.
Tools for this include Peec AI (which we use for tracking), Semrush's AI Visibility Toolkit, Ahrefs Brand Radar and Profound. Free checkers such as HubSpot's AI Search Grader give a quick snapshot of what models think of you; we compared them in free AEO graders tested. Our AI visibility audit guide covers what an audit checks and what it costs.
How we measured our numbers
Our engine figures come from LoudFace's own project in Peec AI, read on 25 October 2026. Peec AI asks each tracked question once a day per engine. The 28-day table covers 27 September to 24 October 2026 and excludes the 6 of our 150 questions that contain our brand name, leaving 144. Visibility is the share of answers that name LoudFace, as Peec AI defines it. The weekly figures cover our 39 main buyer questions, a fixed set, comparing 18 to 24 October 2026 with the 7 days before. The single-question example comes from the three answers Peec AI recorded on 18 October 2026.
These are agency-buying questions in one category. Your engines, your questions and your competitors will give different numbers. The pattern (big gaps between engines, cited without named, lists that change on each ask) is what we expect you to find.
LoudFace's free AI visibility audit checks where your brand stands across ChatGPT, Claude, Gemini, and Perplexity versus your top competitors. For the work itself, our generative engine optimization service is built to get your brand cited and recommended in AI answers.
