Something has quietly shifted in how people find products and services. Not everyone has clocked it yet, but if you pay attention to how your own friends and colleagues make decisions these days, you will notice it. Let’s have a look: Someone needs an invoicing tool. Instead of typing it into Google, they open ChatGPT and just ask. Someone wants to know which CRM is worth the money for a ten person team. Perplexity has an answer ready before they even finish typing.
And those AI tools are recommending specific brands. Not just categories. Actual named companies. The question you should be sitting with right now is whether yours is one of them.
Why does ChatGPT recommend certain brands at all?
A lot of people assume there is some kind of directory to be listed in, or a form to fill out. There is not.
Models like ChatGPT are trained on a huge amount of text from across the internet. Websites, books, forums, news articles, product reviews, documentation, social media threads. During training, the model builds up an understanding of what exists in the world, what different brands do, and how people talk about them. When someone asks for a recommendation, the model draws on all of that.
The brands that come up are the ones the internet talks about, links to, reviews, and discusses. Not the ones with the flashiest website or biggest social following. The ones with a real, substantive footprint across the web.
No amount of money will buy you a mention in ChatGPT. There is no ad product, no sponsored slot. The only way is to become the kind of brand the internet genuinely references.
Your website content is the starting point
If the content on your website is mostly there to make your company look good rather than to actually help someone, that is the first thing worth addressing.
AI tools respond to content that is specific, detailed, and written for a real person trying to solve a real problem. Generic claims, buzzwords, and vague value propositions do not register. What does register is content that gets into the detail of something useful.
Guides that go all the way
There is a real difference between a guide that introduces a topic and one that actually takes someone through it. The internet is full of the first kind. What gets referenced, shared, and linked to is the second. Write guides thoroughly enough that someone could follow them without needing to go anywhere else. That level of usefulness is what earns the kind of online attention that feeds into AI training data over time.
Comparison content that does not flinch
People ask AI tools comparison questions constantly. “Should I go with X or Y?” “Is X worth it compared to Z?” If you produce comparison content in your category, including content where your own product is one of the things being compared, and you do it honestly rather than as a thinly veiled sales pitch, you become part of the conversation around that category in a meaningful way.
Data and research other people want to cite
Original research has a compounding quality to it. You publish a survey, someone writes about it and links to you, someone else cites them, and six months later your brand is woven into discussions you had nothing to do with. Even a modest study of a few hundred people in your industry, with genuinely interesting findings, can take on a life of its own if the topic is relevant and the findings are honest.
FAQ content written in real human language
Read your support inbox. Look at what gets asked in Reddit threads about your category. Those are the questions your FAQ section should be answering, written in the exact language people use. Not the polished corporate version of those questions. The actual ones.
This matters because the way people phrase questions to AI tools sounds a lot like how they would ask a knowledgeable friend. Informal, specific, sometimes a bit rambling. If your content already sounds like that, it aligns better with how people query these tools.
Write for how people talk to AI, not how they type into Google
On Google, people compress their thoughts into fragments. “Best project management software small team.” On ChatGPT, they write the full thing. “We are a small creative agency with six people and we have been using spreadsheets to manage projects but it is getting chaotic. We do not have a big budget and we need something that is not too complicated to set up. What would you suggest?”
Same question. Completely different shape. And if your content speaks to that full context rather than just the compressed keyword version, it becomes far more useful in an AI conversation.
Write specifically for your different customer types
A freelance photographer has different needs from a marketing manager at a mid sized company. A bootstrapped founder thinks about software differently from a procurement manager at a large organisation. If your content treats all of these people as one, it will feel generic and miss the specific queries each of them types into AI tools.
Build content that speaks to your actual customer segments. Not just because it helps with SEO but because it matches the context people provide when they ask AI tools for help.
Answer the questions people ask before they know you exist
A good portion of AI queries come from people who are just starting to figure out a problem. They do not know which brands operate in your space yet. They are asking things like “what kind of software do I even need for this?” If your content meets people at that early stage and helps them understand the landscape, your brand becomes part of their education. People tend to trust, and eventually recommend, the sources that taught them something.
Stay in the know
Keep on top of how the latest marketing, SEO, branding and design tips can be used to
grow your business – delivered straight to your inbox.
Sort out your knowledge graph and structured data
AI tools do not only learn from articles and blog posts. They also draw on structured databases such as Wikipedia, Wikidata, and Google’s Knowledge Graph, which are organised, machine readable collections of facts about the world. If your brand is represented inaccurately in those systems, AI tools are working with a wrong picture of who you are.
Google Knowledge Panel
Type your brand name into Google and look at the information box that appears on the right side of the results. Check whether it exists, and if it does, whether everything in it is correct. Google gives you a way to suggest edits. Get them fixed.
Wikidata
Most people outside the tech world have never engaged with Wikidata, but it feeds into a lot of AI systems. If your brand has an entry, check it for accuracy. If it does not and your company is established enough to warrant one, create one properly following their guidelines rather than treating it as a marketing exercise.
Schema markup on your own site
Schema markup is code that sits on your website and tells machines exactly what your content is about. There is a schema for your organisation, your products, your FAQ content, and your reviews. It takes a developer to implement but it makes a real difference to how AI systems read and understand your site. If you have a developer or an agency managing your site, have a conversation about it.
Reviews matter more than most brands realise
Platforms like Trustpilot, G2, and Capterra carry real weight in AI training data. They are credible, well-indexed, and they represent the voices of actual customers rather than the brand itself. AI tools treat that kind of third-party validation as a meaningful signal.
What matters is not just having reviews but having reviews with substance. A page of five star ratings with no text does very little. Reviews that describe a specific situation, a specific problem, and a specific outcome are the ones that carry weight.
Getting reviews that actually say something
When you follow up with a happy customer, give them a prompt rather than just asking generically. Something like “if you have a moment, it would be great to hear about what you were trying to solve when you came to us and how things have gone since.” That framing tends to bring out the kind of specific, detailed responses that actually mean something to someone reading them.
And do not just collect reviews, respond to them. Especially the negative ones. A company that handles criticism with grace comes across as far more trustworthy than one with a perfect score and nothing to say. People notice how you behave when things do not go perfectly. Often, that tells them more about you than the glowing feedback ever could.
What GEO actually means and why it matters
You will hear the term Generative Engine Optimisation more and more. GEO is the practice of making your content work well in an AI-driven discovery environment rather than just a traditional search one. It builds on SEO, but the emphasis is a bit different.
What this looks like in practice
Write so that each section of your content makes a single clear point. AI tools often pull a specific answer from a longer piece. If your writing buries the point three paragraphs in, the model will struggle to extract a clean answer.
Back your claims up with specifics. Concrete numbers, named examples, real outcomes. AI tools favour content that makes verifiable statements over content that makes big claims without evidence.
Be clear about what category your brand belongs to. If your positioning is deliberately vague or you have coined a new phrase for what you do, you are creating friction for AI systems trying to decide when to surface you. Clarity about your category, paired with clarity about your specific angle, is a much stronger position.
Go and check what AI tools actually say about you
This step takes five minutes and is genuinely worth doing. Open ChatGPT, or Claude, or Gemini, and start asking questions about your brand and your category.
Worth trying right now
Ask “what do you know about [your brand name]?” and read the response carefully. Is it accurate? Does it even know you exist?
Ask “what are the best options for [your product category] in the UK?” and see where you land. Are you in the list?
Ask “I am a [your typical customer] looking for [what you offer]. What would you recommend?” and pay attention to whether your brand comes up.
Ask “how does [your brand] compare to [your main competitor]?” and see how that plays out.
Do this across a few different AI tools because they do not all say the same things. Each one gives you a slightly different window into how your brand is understood.
This takes time, but it does pay off
The brands showing up in AI recommendations today got there because they spent months building a real, substantive presence across the web. There is no shortcut around that. But the brands that start this work now will be in a very different position in twelve months compared to the ones that wait.
A sensible place to begin
Go through your existing website content and ask honestly whether it actually helps your customers or whether it is mostly positioning. That gap is your first priority.
Get someone to look at your schema markup, your Knowledge Panel, and your Wikidata entry. These are foundational and most brands have not touched them properly.
Pick one review platform that matters in your industry and put a real system in place for gathering reviews from happy customers. Not a one-off push but an ongoing habit.
Commit to publishing one genuinely useful long form piece of content a month. Not a press release dressed up as a blog post. Something your ideal customer would actually read and find helpful.
Do those four things consistently and you will be building the kind of presence that AI tools reflect.
Contact us
AI visibility is not a distant priority. It is already shaping which brands get found, considered, and chosen. The businesses appearing in ChatGPT recommendations today are not there by accident; they built a real, substantive presence across the web, and the AI reflected it back.
The good news is that this is not out of reach. It does not require an enormous budget or a complete rebrand. It requires consistency, clarity, and a genuine commitment to being useful to the people you are trying to serve.
Whether you are starting from scratch or you already have the foundations in place and want to sharpen them, the next step is simply understanding where you stand right now.
If you would like to talk through that, we are here for it. We work with businesses at every stage of this journey and would love to have that conversation with you.
