Brand visibility in AI: what it is and how to measure it
What does brand visibility mean in AI?
Which accountancy firm in Poznań provides services to limited liability companies?
Model response- 1Office A
- 2Your brandreference
- 3Office B
- 4Office C
Your company’s name does not appear in the question. What matters is whether the model mentions it, where in the text, and alongside which other names.
What does brand visibility in AI consist of?
| Layer | What does this mean? | What it says about the brand |
|---|---|---|
| mention | The model names the brand in response to a question about the category | Are you even being considered? |
| Item | The brand’s position on the list of companies in response | Will the customer reach you before they stop reading? |
| Citation | The model cites your website as a source | Is your content suitable for quotation? |
| Recommendation | The model presents the brand as a choice, rather than just one of many | Does the model trust the brand enough to recommend it? |
| The tone of the description | The positive, neutral or negative way in which the model talks about the brand | Does this model describe you in the way you want to be perceived? |
How does brand visibility in AI differ from a ranking on Google?
- A page that search engine bots can crawl and index without any problems
- Content in response to questions asked by customers
- Structural data describing the company and its range of products and services
- Appearances in other people’s lists and category rankings
- Reviews on the websites used by the models
- A consistent description of the company across multiple independent sources
Where do AI chatbots get the brands they use in their replies from?
How can brand visibility be measured in AI?
- Ask about the category, not the brand. A question containing the company name tests brand recognition, not what the searching customer actually sees.
- Have a look at a few models. ChatGPT, Claude, Gemini and Perplexity draw on different sources and mention different brands.
- Stick to a set of questions. Changing the wording alters the result, so you should only compare the same questions when they are asked again.
- Measure at a steady pace. A single result is just a snapshot of a moment. It is only the trend over the coming months that is significant.
- Save, who is being named in your place. And where does the model get its data from? This is what it’s based on list of activities.
Read more: by hand or with a tool?
The manual test takes about half an hour and is enough to get a one-off overview of the situation. You’ll find step-by-step instructions in the article how to find out what AI says about a company, and a shorter version in the post How to check whether AI recommends your brand.
Manual testing is no longer sufficient when you want to compare results over time, measure up against the competition, or show the result to the person responsible for the budget. People unconsciously alter the wording, and as a result the two measurements are no longer comparable. That is why at Aelo we work on frozen query set. We have compiled an overview of the available solutions in this article AI visibility tools, and the method for calculating each metric in methodology.
How can you improve your brand’s visibility in AI?
More
The model matches the brand to the customer’s enquiry. If a customer asks about a boiler repair and the website states „comprehensive heating system services”, the model will not link the two. They help structured data in the format Schema.org Organisation and checking that the robots.txt file does not block web crawlers such as OpenAI’s robots.More
It is these sources that determine the recommendation. The model prefers to list a company that has been verified in multiple places rather than one that is only mentioned on its own website. A single entry in a category ranking can make more of a difference than several new blog posts. We describe this mechanism in the post How AI chatbots choose the brands they recommend.More
If a company describes itself differently in different places, the model may fail to combine these descriptions into a single brand, or may confuse it with another. This also applies to product names: a recognisable product is often associated with a category rather than with a manufacturer. We discuss this in the entry brand aliases.How does brand visibility in AI change in practice?
| Credits | 24 June | 27 July | 10 September | 14 October |
|---|---|---|---|---|
| Questions mentioning the brand | 0 out of 18 | 17 out of 18 | 18 out of 18 | 18 out of 18 |
| Website citation index | 6% | 50% | 94% | 100% |
| Models who are familiar with the brand | 3 out of 4 | 3 out of 4 | 4 out of 4 | 4 out of 4 |
| Models who recommend it | 0 out of 4 | 0 out of 4 | 1 of 4 | 1 of 4 |
Source: Aelo surveys across four models, using the same set of 18 questions, June–October 2026.
What sources is this text based on?
- Aggarwal et al. (Princeton University, IIT Delhi), GEO: Generative Engine Optimisation: a study of the impact of content form on the presence of generative models in responses (arXiv 2311.09735, KDD 2024).
- Google Search Central: AI features and your website: How page content is fed into AI Overviews and AI Mode (updated December 2025).
- Google Search Central: structured data for organisations: how to describe a company in structured data.
- OpenAI: bots and their user agents: OAI-SearchBot, GPTBot, ChatGPT-User and managing them in robots.txt.
- SparkToro and Gumshoe.ai: the recurrence of brand lists in AI responses: 2,961 responses from ChatGPT, Claude and Google AI (January 2026).
- Perplexity Help Centre: how Perplexity works: real-time search and numbered citations in replies.
- Aelo reports for Wapiennik „Węże” dated 24 June, 27 July, 10 September and 14 October 2026 (published with the client’s consent).
Questions about brand visibility in AI
Is brand visibility in AI the same as GEO?
No. GEO refers to activities that improve a brand’s presence in models’ responses, whilst visibility is the metric used to measure the effectiveness of these activities. Read more in the article What is GEO?.
Is a high ranking on Google enough?
No. They do help, but the recommendations in the models’ responses are largely determined by sources outside your website: reviews, rankings and articles.
Why does visibility on AI matter if I’m getting traffic from Google?
Because some decisions are made based on the result itself, without even clicking. In a 2025 Pew Research study, Google users who were shown an AI summary clicked on the result in 8% of visits, compared with 15% of visits where no summary was shown. We discuss the data in the post Visibility in AI and SEO.
Can I check my brand’s visibility on AI myself, without using a tool?
Yes. Ask a few AI chatbots the same questions about their category, without mentioning the company name, and make a note of which brands are mentioned. You’ll find instructions in the article how to find out what AI says about a company.
Why does the model mention the competition rather than my company?
Because, when it comes to the competition, there are more independent endorsements. This is not an assessment of service quality. See the article for details Why does AI recommend the competition?.
How quickly can you improve your brand’s visibility on AI?
Changes to your own website usually become visible after a few weeks, whilst off-site signals take months to take effect. For the lime manufacturer, mentions appeared after five weeks, and the first recommendation came after 11 weeks. We explain this in detail in the article Why aren’t the changes visible straight away?.
Is it possible to guarantee that the model will endorse my brand?
No. Nobody controls what the model generates. You can increase the likelihood of a mention or a recommendation, and measure the change between successive measurements.
Which models are worth looking at?
At least ChatGPT, Claude, Gemini and Perplexity, as everyone uses different sources and mentions different brands. We explain how we measure them at Aelo in methodology.
How can you measure brand visibility on ChatGPT?
Just as with other models: ask ChatGPT a fixed set of questions about your category, without mentioning any brand names, each in a new conversation or in a temporary chat, and note down whether your company appears and, if so, in which position. However, ChatGPT alone isn’t enough, as Claude, Gemini and Perplexity mention other brands. You’ll find the instructions in the article how to find out what AI says about a company.
How can you distinguish between questions about a brand and questions about a category?
Keep them in separate sets and count them separately. Questions without a company name show whether the model mentions you to someone searching within your category, whilst a question containing a brand name merely checks whether the model recognises the brand. The brand may be well known to the model but not appear in any category-related responses. We describe this distinction in the entry assessment and recommendation.
