What Is AI Visibility? How to Check What ChatGPT and Google Say About Your Brand

More of your future clients now ask an AI assistant before they ask a colleague. This is a plain guide to what “AI visibility” means for architects, interior designers and design brands, how to check yours in about five minutes, and what actually changes the answer.

A developer in Dubai looking for an interior architecture practice, or a London homeowner looking for a lighting designer, used to open Google and work down a list of links. Many still do. But a growing share now type the question into ChatGPT, Perplexity or Gemini, or read the AI Overview that sits above Google’s usual results, and they get a short, confident answer with names in it.

If your studio is one of those names, that answer does a lot of your introduction for you. If it isn’t, you don’t lose a ranking position you can see. You simply aren’t in the conversation, and nobody tells you.

That gap is what I mean by AI visibility. I’ve spent a lot of this year testing it for design brands and for my own site, and the good news is that most of what decides it is within your control.

What AI visibility actually means

AI visibility is how often, and how accurately, AI assistants mention your brand when someone asks a question you should be part of the answer to. It has three parts, and it helps to think about them separately.

• Presence. Are you named at all when someone asks for, say, “residential architects in Surrey” or “interior designers in Dubai for a villa refurbishment”?
• Accuracy. When you are named, are the details right? Your location, your specialisms, the kind of projects you take on, who you’ve worked with.
• Framing. Are you described the way you’d describe yourself? “A boutique practice known for hospitality interiors” is a very different introduction from “a design firm based in London”.

Traditional SEO is mostly about the first of these: whether a page ranks. AI answers put more weight on the other two, because the assistant isn’t handing over a list for the client to judge. It’s doing some of the judging itself.

Where the answers come from

It’s easy to imagine these tools “know” about your studio in some deep way. In practice, the answer to a question about suppliers or practices is usually assembled from what the assistant can find and read on the web at that moment, plus what it absorbed during training.

ChatGPT, Perplexity, Gemini and Google’s AI Overviews all run some form of web search for questions like these, then summarise what they find. That means the sources feeding the answer tend to be:

• your own website, if it clearly says what you do, where and for whom;
• directories and “best of” listings, which are often the easiest pages for a machine to turn into a shortlist;
• press coverage, awards and features in the design media;
• review platforms and your Google Business Profile;
• the websites of people you’ve worked with, such as suppliers, collaborators and clients who credit you.

The practical consequence is that an AI assistant tends to repeat whatever is written down most clearly and most consistently about you. Quality of work doesn’t travel on its own. It has to be described in words, somewhere the assistant can read it.

A five-minute check you can do today

You don’t need any special tools to get a first read on this. Open ChatGPT, Perplexity and Gemini (a free account is fine for each), and search the same question on Google to see whether an AI Overview appears. Then ask the kind of questions a real client would ask. For example:

1. The shortlist question. “Which [your discipline] practices would you recommend in [your city] for [your typical project]?”
2. The named question. “What can you tell me about [your studio name]?”
3. The comparison question. “How does [your studio] compare with [a studio you often compete with]?”
4. The specialist question. “Who designs [a specific thing you’re known for] in [your market]?”
5. The buying question, for product brands. “Where can I buy [your product or brand] in [city]?”

For each answer, note three things: whether you appear, whether the details are correct, and which websites the assistant cites or links to. That last point is the most useful. It shows you which sources the tool trusts for your corner of the market, and those are the places worth being described well.

Run the same questions again in a few weeks. The answers vary from one day to the next and between tools, so a single result is a snapshot, not a verdict. Patterns over several checks are what matter.

What I found when I ran these checks in Dubai

On 13 September 2026 I ran a set of these questions for the Dubai market while preparing my Dubai & UAE page. Two results stood out.

Asked for interior architecture practices to shortlist in Dubai, one assistant named thirteen practices. Most of its sources were third-party listing sites rather than the practices’ own websites. A practice with genuinely landmark work in the city wasn’t on the list, and as far as I could tell the reason was simple: very little about that work was written down anywhere the assistant could read it.

A second question, about where to buy two well-known Italian furniture brands in Dubai, returned the brands’ own global pages. The UAE company that holds the exclusive rights to both brands in the region didn’t appear.

Neither result says anything about the quality of those businesses. Both say a great deal about what was, and wasn’t, written down about them.

Why design brands get left out

Design businesses are often very good at showing their work and less good at describing it. That is exactly the wrong way round for an AI assistant, which reads text far more reliably than it interprets a beautiful photograph. The patterns I see most often:

• Project pages that are all image. A gallery with a project name and nothing else tells a machine almost nothing. Location, typology, scale, the brief, the collaborators and the outcome are what let an assistant connect you to a question.
• Vague positioning. “We create timeless spaces” could describe thousands of studios. “Residential architects specialising in listed buildings in south-west London” is something an assistant can match to a question.
• Inconsistent details. A different address on Google, an old studio name on a directory, a service you stopped offering still listed on a profile. Inconsistency makes an assistant less confident, and less confident usually means less likely to name you.
• Absent from the listings that feed the answers. If the shortlists in your market are built from a handful of directories and features, being missing from all of them is a real disadvantage, however strong your own site is.
• Nobody else describes you. Suppliers, collaborators and clients who credit you by name, with a link, are some of the most persuasive sources there are. Many studios never ask.

What to fix first

If you’ve run the check and don’t like what you found, this is the order I’d work in. None of it is exotic, and most of it improves your ordinary search visibility at the same time.

1. Say clearly who you are, what you do, where and for whom. On your homepage and your About page, in plain sentences. This is the text an assistant is most likely to quote.
2. Write up your projects properly. Start with your five most representative projects. For each, add a short paragraph covering location, type, scale, the brief and what changed. Name collaborators where you can.
3. Make your details consistent everywhere. Your website, Google Business Profile, LinkedIn, Instagram bio and any directory listings should agree on your name, location and specialisms.
4. Get into the sources the assistants already cite. Your five-minute check tells you which listings and publications appear for your questions. Those are your targets for entries, features and awards.
5. Ask the people you work with to credit you. A line and a link on a supplier’s or client’s project page is worth more than you’d think.
6. Answer the questions clients actually ask. A short FAQ on the relevant service or project pages, in your own words, gives an assistant something direct to draw on.

What llms.txt does, and what it doesn’t

You may have come across advice to add an “llms.txt” file to your website. It’s a simple text file, proposed as a way to give AI tools a tidy summary of a site and its most important pages. My site has one, and it takes minutes to set up with most SEO plugins, so there’s little reason not to.

But it isn’t a switch that makes AI assistants recommend you. Google has said publicly that it doesn’t use llms.txt in search, and there’s no clear evidence that any of the major assistants rely on it when deciding which businesses to name. Treat it as good housekeeping, not a strategy. The work that moves the answer is the work in the list above.

How this fits with ordinary SEO

AI visibility isn’t a replacement for search engine optimisation. It’s built on the same foundations: a website that loads well, that search engines can crawl, with clear pages that each do one job. Google has said that its AI features draw on the same core systems as its regular results, and that the same good practice applies.

What changes is the emphasis. With a list of links, a client could click through and judge your portfolio for themselves. With an AI answer, the assistant makes a first judgement on their behalf, based on what it can read. That puts more weight on clear description, on consistency, and on what other people say about you.

For design brands, that’s a real opportunity. Most of your competitors haven’t looked at this yet, and the fixes are mostly writing and tidying rather than technical overhaul.

Want me to check yours?

If you’d rather not run the questions yourself, I offer a free AI visibility check for architects, interior designers and design brands. I run a set of client-style questions across the main assistants and send you a short summary of where you appear, what’s said about you and the first things I’d fix, within five working days. If you want help making the changes, that’s what my AI SEO service is for.

Related: Free AI visibility check · AI SEO · Marketing for architects · Marketing for interior designers · How AI is transforming social media