People increasingly use AI assistants to research products, compare providers and answer questions that might once have been typed into a search engine. For marketers, that creates a new question: What does an AI assistant say about your brand when someone asks a relevant question?
This is what we mean by AI brand visibility: whether and how your brand appears in AI-generated answers across platforms such as ChatGPT, Gemini and Perplexity. You may also see this described simply as AI visibility.
However, AI visibility can also mean something different in web analytics: visibility into AI-related website activity, from visits referred by AI assistants to chatbot requests and agent activity. We explore that broader measurement challenge in our article on the AI visibility gap.
Here, we’re focusing on AI brand visibility: what it measures, why marketers are paying attention to it and what the results can actually tell you.
What is AI brand visibility?
AI brand visibility describes whether and how a brand appears in AI-generated responses to relevant questions.
Imagine someone asks an AI assistant: “What are the best privacy-focused web analytics platforms for a European organisation?”
The response might mention Matomo as one of the available platforms, recommend it for organisations with particular privacy or data ownership requirements, or cite a page on matomo.org as a source. These are different forms of visibility. Adobe uses a similar distinction between brand mentions, citations and recommendations.
- Brand mentions show whether your brand appears in the answers you are monitoring.
- Recommendations show whether the AI system presents your brand as a suitable option in a particular context.
- Citations show whether your website or content is referenced as a source.
- Brand portrayal shows how the answer describes your organisation, products or services.
A brand might be cited as the source of a statistic without being recommended, while another could be recommended based mainly on information from third-party sources. AI brand visibility therefore goes beyond whether your domain appears in a traditional list of search results. It looks at how your brand and content form part of the answer itself.
AI brand visibility and brand awareness aren’t the same
AI brand visibility is often treated as if it were a measure of brand awareness. The two are related, but they capture different things.
If your brand appears in 30 out of 100 monitored AI responses, you know it appeared in 30% of that sample. You don’t know how many people saw those responses, remembered your brand or changed their perception of it.
Brand awareness measures what people recognise or recall. AI brand visibility measures how your brand appears across the AI-generated answers being monitored. A higher visibility score shows that your brand appears more often within the monitored sample, but it doesn’t show how aware of your brand the wider audience actually is.
Why does AI brand visibility matter?
Consider how someone might research a new analytics platform. They may start by asking an AI assistant what platforms meet their requirements. Next, they might ask for a comparison between two or three products, query a specific privacy feature and ask which option would suit an organisation operating in Europe.
By the time that person reaches a vendor’s website, the AI assistant may already have introduced several brands, explained their perceived strengths and weaknesses and narrowed the options worth investigating.
That makes the answers part of the discovery and consideration process. A brand that appears consistently for relevant questions has an opportunity to become part of that research. A brand that rarely appears may be absent from a conversation that takes place before the website visit.
An answer might describe your product accurately, recommend it for a relevant use case and cite your own documentation. It might instead use outdated information, associate the brand with the wrong type of customer or recommend competitors while leaving your organisation out altogether.
For marketers, the useful questions go beyond whether the brand appears at all:
- Do we appear for topics that matter to our prospective customers?
- How are we described when we appear?
- Are we recommended for the use cases we want to be associated with?
- Which competitors appear for questions where we don’t?
- Which websites and pages are being cited in those answers?
Looking at these patterns can show where your brand is already part of AI-led research, how it’s represented and where competitors have a stronger presence.
How is AI brand visibility measured?
AI-generated answers don’t work like a traditional search results page, where you can check a keyword and record a single ranking position.
The answer can change depending on the platform, how the question is phrased, the user’s location and language, or the model version. To measure AI brand visibility consistently, you need a defined set of prompts and platforms, then track how your brand appears across those responses over time.
Step 1: Choose the questions you want to monitor
The prompts should reflect questions relevant to your market and customer research.
For example, a web analytics provider might monitor prompts around:
- What are the best web analytics platforms for marketing teams?
- What are the best alternatives to [analytics product]?
- Which web analytics platforms are best for privacy-conscious organisations?
- What are the best web analytics platforms for ecommerce?
- Which analytics platforms are best suited to [industry or organisation type]?
- Which web analytics tools offer heatmaps, attribution and conversion tracking?
Branded prompts can also provide useful information, such as questions comparing your brand with a competitor. However, unbranded prompts are particularly useful when you want to understand whether a brand appears before someone has already decided to investigate it.
The platforms matter too. An organisation could have strong visibility in one AI service and limited visibility in another, so a single platform doesn’t necessarily represent the wider picture.
Step 2: Measure how your brand appears
Once you’ve chosen the prompts and platforms to monitor, look at how your brand appears across the resulting answers.
The IAB’s 2026 framework “Measuring Visibility in the AI Era” groups these measurements into four areas: presence, prominence, portrayal and persuasion. Individual tools use different metrics and calculation methods within these areas, so it’s worth checking exactly how each figure is defined.
Presence
Start with whether your brand or content appears at all.
- Mention rate measures how often your brand appears across the responses you’re monitoring.
- Citation rate measures how often your website or content is referenced as a source.
- Share of voice compares your visibility with other brands across the same monitored set.
Prominence
Next, look at where and how substantially your brand appears. A brief mention in a long list gives your brand a different level of visibility from appearing near the top of an answer or receiving a more detailed explanation.
Portrayal
Review how the AI system describes your brand. This can include sentiment, framing and factual accuracy. Check whether the information is current and whether the description reflects what your organisation actually offers.
Persuasion
Finally, look at whether the response actively recommends your brand. The IAB describes recommendation strength as one measure within this area, helping distinguish a passing mention from a stronger endorsement.
These categories provide a useful structure, but the underlying calculations can vary considerably between tools. Check the methodology before comparing figures from different providers.
An AI brand visibility score needs context
AI brand visibility tools don’t all use the same methodology. The IAB highlights this inconsistency as a challenge in AI brand visibility measurement and distinguishes between directional measurements, which can help identify patterns, and more rigorous measurements intended to support significant business decisions.
Before comparing scores or reacting to a change, ask:
- Which platforms were measured?
- Which prompts were included?
- How many responses were analysed?
- Which countries and languages does the data represent?
- What counts as a mention or citation?
- Has the methodology changed since the previous measurement?
A score can summarise the results. Understanding the underlying sample tells you what that score actually represents.
What can AI brand visibility tell marketers?
Compare the questions where your brand appears with those where it doesn’t. Competitors may show up more often for use cases you consider a strength, or AI answers may rely on outdated or third-party information about your product.
That gives you specific areas to investigate across your positioning, content and external coverage.
Your brand is missing from relevant questions
If competitors keep appearing for questions that closely match your product category, start by looking at what gives them that visibility:
- How are they described?
- Which capabilities or use cases are they associated with?
- Which sources are cited alongside those answers?
You may find that competitors have clearer content around a particular need, stronger coverage in third-party sources or more consistent language across their own website. That gives you something specific to work with. Review whether your own content answers the same customer question clearly and whether the positioning you want to own is visible beyond your website as well.
Your brand appears, but the description is inaccurate
An AI answer might mention your brand while getting an important detail wrong. Perhaps it refers to an old product offering, misses a capability you now provide or describes your positioning in a way that no longer reflects the product.
Where citations are available, trace the information back to its source. Check whether the same outdated or ambiguous wording still appears on your own website, in documentation or on prominent third-party pages. If several AI answers repeat the same error, that can help you identify which public information needs attention rather than treating each inaccurate answer as an isolated problem.
Your website is cited, but your brand isn’t recommended
This can be a particularly interesting pattern. Your content is strong enough to help support the answer, but the recommendation goes to another brand.
Look at what the AI system is taking from your content and what it says when explaining the recommendation. Your article might explain a problem extremely well while making only a weak connection to how your product addresses it. Or a competitor may have clearer evidence for the particular requirement in the prompt.
The aim isn’t to make every cited article more product-led. Instead, check whether your content connects your expertise to relevant product capabilities where that connection is genuinely useful to the reader.
Competitors are more visible for particular topics
An overall visibility score can hide where the meaningful differences are. You might perform strongly for privacy-related questions, for example, but appear much less often when people ask about ecommerce analytics or a particular industry use case.
Break the results down by topic and compare them with the areas that matter commercially. If competitors consistently dominate a topic you consider a strength, review how they’re being positioned and which sources support that association. You can then decide whether the gap comes from your content, your wider positioning, third-party coverage or simply the way your current monitoring set is structured.
This is usually more actionable than trying to improve one overall visibility number.
Your visibility changes over time
A rise or fall in visibility deserves investigation, but make sure the comparison is valid first. A new AI platform, a different set of prompts, another language or market, or a larger sample can all change the result even when your actual presence hasn’t shifted in the same way.
Keep the core measurement set stable when you want to track progress. If you do change it, document what changed so the next comparison still makes sense.
Once the methodology is consistent, look beyond the headline number:
- Which topics gained visibility?
- Did recommendations increase, or only mentions?
- Are new sources being cited?
A change becomes much more useful when you can see what actually moved underneath it.
What AI brand visibility doesn’t tell you
AI brand visibility adds another source of evidence to marketing measurement by showing how your brand appears in monitored AI-generated answers. It doesn’t reveal the full role AI may have played in a customer journey.
For example, knowing that your brand appears frequently in monitored ChatGPT responses doesn’t tell you whether a particular person saw those responses and later became a customer.
Likewise, web analytics might show that someone arrived through a link from an AI assistant and completed a conversion. That provides evidence about the website visit and subsequent actions, but it doesn’t reveal every AI interaction that may have happened before that click.
And there is another type of AI activity to consider. Chatbots and agents can access website content without sending a human visitor at all. Matomo, for example, separates identifiable human referrals from supported chatbot and agent activity in its AI reporting.
To understand AI-related discovery and website activity, it helps to separate the signals you can observe:
- AI brand visibility shows how your brand appears in the AI-generated answers being monitored, including mentions, recommendations, citations and portrayal.
- AI referral reports show identifiable visits sent to your website from AI assistants such as ChatGPT, Gemini or Perplexity.
- Chatbot and agent reports show supported automated activity on your website, such as content requests or agent interactions.
- Visitor behaviour and conversion reports show what recorded visitors do after they reach your website, including the pages they view and whether they complete the actions you care about.
These signals can complement each other, but each answers a different question. Keeping them separate helps avoid drawing conclusions that the data doesn’t support.
Looking beyond AI brand visibility
AI brand visibility shows how your brand appears in monitored AI answers. To understand what happens next, combine that view with website data on AI referrals, visitor behaviour and conversions, while keeping automated chatbot and agent activity separate. The goal is to close useful measurement gaps without collecting more data than you need.
Matomo helps you analyse AI-related website activity alongside the rest of your analytics, while keeping control over how your data is collected and processed.
