AI Visibility: What It Is, Key Metrics and How to Score It

AI visibility is how often, how prominently and how accurately your brand appears in answers generated by AI platforms such as ChatGPT, Google AI Overviews, AI Mode, Gemini, Perplexity and Claude. It covers whether the AI mentions you, links to you as a source, recommends you over competitors and describes you correctly. Also called AI search visibility, it is measured across a set of prompts your buyers actually ask.

This guide defines AI visibility, breaks down the seven metrics behind it, shows exactly how an AI visibility score is calculated with a worked example, explains why your results differ so much between platforms using our own research, and lists the practical steps that improve it.

What is AI visibility?

AI visibility is a measure of your brand's presence inside AI-generated answers, rather than in a list of search results. When someone asks an AI assistant a question, the system writes one answer from several sources. Your brand is either part of that answer, cited as a source, or absent.

That makes it different from traditional SEO visibility. In classic search, you can rank fifth and still get seen. In an AI answer, there is often no fifth place: the response names a handful of options, links a few sources, and stops.

For a short definition you can share internally, see Lantern's glossary entry on AI search visibility.

Why AI visibility matters

AI visibility matters because more discovery now happens inside AI answers, and fewer people click through to the sources behind them. If your brand is not in the answer, many buyers never see you at all.

The usage is large. Google said in May 2026 that AI Mode had passed one billion monthly users one year after launch, with queries more than doubling every quarter.

The click behavior has changed too. A Pew Research Center study of 900 US adults' browsing in March 2025 found users clicked a traditional result in 8% of visits when an AI summary appeared, versus 15% when it did not. They clicked a link inside the AI summary in just 1% of visits.

Traffic that does come through can be valuable. Google's own AI features guidance says clicks from results pages with AI Overviews are "higher quality," meaning users are more likely to spend more time on the site. But you can only earn those clicks if the answer includes you.

AI visibility metrics: what to measure

AI visibility is not one number. It is built from seven metrics, each answering a different question about how AI represents your brand.

Mentions and citations

A mention is when the AI names your brand in the answer. A citation is when it links to a source, ideally your own page. The two are different, and both matter: a mention builds awareness, while a citation is what earns a click. Track them separately so you can see whether AI knows you but sends readers elsewhere.

Share of voice

Share of voice compares your mentions with competitors' mentions across the same prompts. If three brands are mentioned 180 times in total and you account for 54, your AI share of voice is 30%. See the glossary definition of AI share of voice for variations.

Position and sentiment

Position measures where you appear when you are mentioned: first recommendation, one of several, or a passing reference. Sentiment captures whether the AI describes you positively, neutrally or with caveats. A brand mentioned often but always last, or always with a warning, has weaker visibility than its mention rate suggests.

Accuracy

Accuracy asks whether the AI gets your facts right: what you sell, who it is for, what it costs and which category you belong in. Inaccurate answers can cost deals even when your mention rate looks healthy. Our guide to detecting and correcting AI hallucinations about your brand covers how to fix them.

Referral traffic

Referral traffic is the visible downstream result: sessions in your analytics that come from ChatGPT, Perplexity, Gemini and other AI platforms. It understates true influence, because many people read an answer and never click. Our analysis of AI referral traffic sources and GA4 tracking shows how to separate it from other referrals.

The six stages of AI visibility

The most useful way to read these metrics is as stages, from AI knowing your brand to a buyer visiting your site. Lantern's research on the AI Recognition Gap measures six of them separately.

1. Recognized. AI can explain your brand when the question names it.
2. Cited. AI links to your website or another relevant source about you.
3. Associated. AI connects you with the right category, audience or use case.
4. Discovered. AI includes you when the question does not name you.
5. Recommended. AI presents you as a suitable option.
6. Visited. A person clicks from the AI answer to your site.

The gap between stages can be large. In that study of four emerging brands tracked from April to September 2026, ChatGPT correctly identified a brand in 95.0% of answers that named it. Yet only one of the four brands appeared in a live unbranded category test, and that brand showed up in one of five audience-specific ChatGPT answers (20%). Being known is not the same as being recommended.

How is an AI visibility score calculated?

An AI visibility score is a composite number, usually 0 to 100, that combines several metrics (typically mention rate, citation rate, position and sentiment) across a fixed set of prompts and platforms. There is no industry standard formula, so every tool weights the inputs differently.

That is why two tools can give the same brand different scores. Before trusting any score, ask which prompts it uses, how many times each is run, which platforms it covers and how each metric is weighted.

There is academic precedent for weighting by prominence. The GEO research paper by Aggarwal et al. measured visibility with a "position-adjusted word count," which gives less weight to sources cited later in an answer because earlier sentences are more likely to be read.

Worked example: calculating a visibility score

Here is a transparent, illustrative formula you can reproduce in a spreadsheet. It is not Lantern's or any vendor's proprietary method.

Step 1: Build the prompt set. Choose 40 unbranded prompts buyers ask in your category. Run each three times on ChatGPT and three times on Google AI, giving 120 answers per platform.

Step 2: Score every answer. For each answer, record four values:

• Mentioned: 1 if your brand is named, otherwise 0.
• Cited: 1 if the answer links to your domain, otherwise 0.
• Position weight: 1.0 if you are named first, 0.7 second, 0.5 third, 0.3 lower, 0 if absent.
• Sentiment: 1 positive, 0.5 neutral, 0 negative or absent.

Step 3: Average each metric across all answers, then weight them. This example uses 40% mention, 25% citation, 20% position and 15% sentiment:

1Visibility score = 100 x (0.40 x mention rate2                        + 0.25 x citation rate3                        + 0.20 x average position weight4                        + 0.15 x average sentiment)

Step 4: Apply it per platform. Suppose the results look like this:

For ChatGPT: 100 x (0.40 x 0.450 + 0.25 x 0.150 + 0.20 x 0.279 + 0.15 x 0.383) = 33.1. For Google AI, the same formula gives 17.7.

Step 5: Add share of voice for context. If two competitors were mentioned 81 and 45 times in the same ChatGPT answers, this brand's ChatGPT share of voice is 54 out of 180, or 30%.

Notice what a single blended score would hide. Averaging the two platforms gives 25.4, which describes neither reality: the brand is reasonably visible in ChatGPT and weak in Google AI. Report each platform separately.

Use enough prompts: Small prompt sets produce noisy scores. In the Recognition Gap study, expanding a tracked prompt set from 10 to 100 questions cut variation in reported brand coverage by about 73%. Run each prompt more than once and track trends, not single snapshots.

What is a good AI visibility score?

A good AI visibility score is one that beats your direct competitors on the same prompts and platforms, and keeps rising. Because scoring methods differ, an absolute number like 60 out of 100 means little on its own.

Benchmark against two or three competitors, compare platform by platform, and judge progress over 60 to 90 days. A score built from branded prompts will always look higher than one built from unbranded category prompts, so check which kind you are looking at.

Why AI visibility varies by platform

AI visibility varies by platform because each AI system uses different retrieval, different sources and different rules for when to cite. The same brand can be well known to one engine and nearly invisible in another.

Our research on emerging brands shows how large the differences are:

Sources: The AI Recognition Gap (linked above) and Where New Brands Earn AI Authority, Lantern, September 2026.

Three patterns stand out:

• Recognition differs. ChatGPT recognized named brands far more reliably than Google AI did.
• Owned pages play a small, platform-specific role. In the Beyond the Homepage report (September 2026), for the brand with the broadest owned footprint, ChatGPT cited editorial and resource pages most (38.1% of owned citations), while Google AI cited product and service pages most (53.6%).
• Platform changes look like brand problems. Google AI recognition fell at roughly the same time for several unrelated brands, which points to a platform change rather than anything those brands did.

The lesson is to measure each platform on its own, and to look for patterns across brands before reacting to a drop.

How to check your AI visibility

You can check your AI visibility manually by asking AI platforms the questions your buyers ask and recording whether, where and how your brand appears. Automated tools do the same thing at scale.

A quick manual check looks like this:

1. Write 20 to 30 real buyer questions, mixing branded, unbranded and comparison prompts.
2. Run each on ChatGPT, Google AI Overviews or AI Mode, Perplexity and Gemini.
3. Record mention, citation, position, sentiment, accuracy and competitors named for each answer.
4. Repeat on a schedule and compare trends.

Our step-by-step guide to measuring AI search visibility covers baselines, GA4 setup, monitoring and competitor benchmarking in detail. If you want to compare monitoring platforms, the AI search visibility tools buyer's guide lays out evaluation criteria.

The fastest way to get a baseline is Lantern's free AI Visibility Report, which shows how ChatGPT, Perplexity, Gemini and Claude surface your brand, including visibility scores, citations and share of voice. No account is required.

How to improve AI visibility

You improve AI visibility by giving AI systems more credible evidence about your brand, in the places they already look, structured so they can extract it. Third-party evidence matters most, because it supplies the vast majority of AI citations.

1. Build an evidence network. Earn accurate coverage on the publications, review sites, directories, comparison pages and communities already cited for your category.
2. Expand beyond the homepage. Create focused pages for products, audiences, use cases, pricing and proof, each answering one clear question.
3. Fix inaccurate facts. Keep product, pricing and category information consistent everywhere it appears, and correct third-party errors at the source.
4. Cover the full question. AI systems split prompts into sub-questions through query fan-out, so pages need to answer follow-ups such as price, comparisons and fit.
5. Write extractable passages. Lead each section with a direct answer, keep one idea per paragraph and use tables for comparisons.
6. Keep crawlers in. Allow the search crawlers that power AI answers, such as Googlebot and OpenAI's OAI-SearchBot. OpenAI's ChatGPT search help page says a site must allow OAI-SearchBot to be eligible for inclusion. Lantern's free AI Crawler Checker shows what your robots.txt allows.
7. Measure by stage and platform. Track recognition, citation, discovery, recommendation and visits separately for each engine.

For a deeper playbook on earning citations, see the AI search citation playbook.

Common mistakes

• Relying on one blended score. Averages hide a strong result on one platform and a weak one on another.
• Testing only branded prompts. Recognition looks great; discovery is where most brands struggle.
• Judging from one answer. AI responses vary between runs. Patterns across many prompts and repeats are the signal.
• Publishing more pages instead of better evidence. The Recognition Gap report notes that, in a wider study, web mentions had a much stronger relationship with AI visibility than the number of pages on a site.

Lantern tracks mentions, citations, sentiment, share of voice and recommendation position across ChatGPT, Google AI Overviews, Gemini, Perplexity and Claude, and connects that data to Search Console, analytics and AI referral traffic. Its agents then turn visibility gaps into review-ready work, such as briefs, content updates and outreach, with your team approving what gets published.

AI visibility decides whether buyers hear your name before they ever visit your site. Measure it by platform and by stage, strengthen the evidence AI relies on, and track the trend over time.

Frequently asked questions

What is the difference between AI visibility and AI search visibility?

They mean the same thing. Both describe how often and how accurately a brand appears in AI-generated answers. "AI search visibility" emphasizes search-style experiences like AI Overviews and AI Mode, while "AI visibility" is often used more broadly to include chat assistants.

How do I check my AI visibility for free?

Ask ChatGPT, Gemini, Perplexity and Google AI Mode the questions your buyers ask and record whether you are mentioned, cited and described correctly. For a faster baseline, free AI visibility reports run a set of prompts for you and return scores, citations and share of voice.

How often should I measure AI visibility?

Track priority prompts on a regular schedule, such as weekly, and review trends monthly. AI answers change between runs, so a single check is not reliable. Look at movement over 60 to 90 days before drawing conclusions.

Is AI visibility the same as share of voice?

No. Share of voice is one component of AI visibility. It shows your mentions relative to competitors, while AI visibility also covers citations, position, sentiment, accuracy and referral traffic.

Why does my brand appear in ChatGPT but not in Google AI Overviews?

Each platform retrieves and cites sources differently. In our research on emerging brands, ChatGPT recognized named brands in 95.0% of answers versus 56.8% for Google AI, and the two preferred different page types. Check which sources each platform cites for your category and close the gaps platform by platform.

How long does it take to improve AI visibility?

It depends on what is missing. Technical fixes such as unblocking crawlers can matter as soon as pages are recrawled, while earning third-party coverage and building authority takes longer. Measure over at least 60 to 90 days with a stable prompt set.