Over the last year, almost every SEO conversation I’ve had with business owners has ended with a variation of the same question:
“Okay, but how are you accounting for AI?”
On one level, it’s a fair concern. People are using ChatGPT, Gemini and other tools to research problems, compare options and sense-check decisions. If those tools never mention your brand, it feels like you’re missing out on a new channel.
The problem is that most businesses think of “AI visibility” as some new trick you bolt on to an existing website. In reality, the way you show up in AI tools is exactly the same way you show up in Google: by having a clear, crawlable, well-structured presence that actually answers the questions people ask.
This is the part nobody wants to hear. There’s no hack, no secret upload, no button marked “submit brand to AI”. What I do day-to-day for clients is still SEO. It just happens to be the version that modern search engines and AI models actually understand.
🔑 Key Takeaways
- “AI visibility” isn’t a separate channel. It’s the by-product of doing modern SEO properly.
- Most businesses overestimate AI traffic and underestimate how niche-specific it is.
- If your brand has no clear digital footprint, AI tools have nothing reliable to cite.
- The fastest wins come from fixing crawl issues, clarifying services and adding structured, chunkable content.
- Fancy “AI citation” tools can distract you from the basics that actually move the needle.
- Real AI visibility follows strong architecture, entity clarity and BOFU content – not the other way round.
🧭 Jump to:
Why everyone suddenly wants “AI visibility”
Recently, I was on a strategy call with the owners of a new SEO client. We’d just walked through a full plan: technical fixes, content architecture, local signals, the usual. At the end, they asked for a follow-up call specifically to talk about how we were “accounting for AI”.
When we got back on the phone, their assumption was simple: SEO is one thing, AI is another. They thought there would be a separate checklist for “getting into ChatGPT” or “appearing in AI Overviews”.
Here’s what I actually explained:
When we write a blog for them, every section is designed to be chunkable. We tighten up definitions, break out comparisons into tables and bundle facts into neat, scannable blocks. That helps Google, but it also makes life extremely easy for AI systems that are looking for clean, citeable pieces of information.
If someone asks an AI tool, “This brand offers X – but can they handle my situation Y?”, the answer is only ever going to be “yes” if the website has a page or section that actually explains that scenario clearly. If their competitors haven’t done that, our client becomes the only high-confidence source worth mentioning when people compare options.
None of that is magic. It’s just architecture, clarity and relevance. Which is why I get slightly nervous when “AI visibility” gets treated as a separate, shiny thing to chase.
Common misconceptions I hear every week
There are a few beliefs that crop up so often I can almost predict them at the start of a call.
1. “AI is going to be our biggest channel, right?”
The short answer: not yet, and not for most businesses.
I’ve seen brands dominate references in AI tools for specific topics and still get a tiny fraction of the traffic they see from ranking top in Google for the same queries. AI usage is growing, but the behaviour patterns are messy, and the volumes aren’t evenly distributed.
You can absolutely get fantastic leads through AI tools. I’ve seen it happen. But in 2025, being number one on Google for a high-intent query will still beat “we’re mentioned a lot in ChatGPT” nine times out of ten.
2. “Our customers use every AI tool under the sun”
I work with a lot of owners who have discovered a niche AI tool and assume their customers are using it too. The reality is usually more boring. Outside tech and marketing circles, most people are still using:
- ChatGPT on their phone or laptop
- Gemini baked into Google products
- Maybe one more assistant if their job nudges them towards it
Tools like Perplexity are brilliant for people who live in this world all day. Most customers do not. It’s worth being honest about that when you’re deciding where to put your effort.
3. “If we fix AI, everything else will follow”
Some niches are simply not AI-heavy. If your boiler explodes or your roof starts leaking, you are not spending two hours in ChatGPT comparing five service providers. You are searching “roofer near me”, checking reviews and making a phone call.
On the other hand, if you are in a complex, high-stress space – legal support, finance, health, anything to do with children – people will sit in AI for a long time trying to make sense of what’s happening. Those are the niches where AI visibility genuinely matters.
The difficulty is that a lot of business owners misjudge where they sit on that spectrum.
The real reasons AI doesn’t mention your business
When AI tools don’t mention a brand, owners often assume something is “wrong with the AI”. Most of the time, the issue is much more straightforward: there isn’t enough clear, consistent, structured information available for the model to use.
Here’s what that looks like in practice:
| Reason AI ignores you | What it looks like in the wild | Impact on AI visibility |
|---|---|---|
| No digital footprint | Thin website, hardly any content, dead socials, no blog, no story. | Models have nothing to latch onto, so they default to bigger brands. |
| No clear services or locations | Site doesn’t list all services, areas or use cases in plain language. | AI tools cannot confidently answer “Do they help with X in Y area?”. |
| No pricing or commercial detail | Pages avoid price ranges, package names or concrete offers. | You never get cited in “who’s cheapest / best value” style questions. |
| No brand narrative | About page is a paragraph of fluff, no “who we are / who we’re for”. | AI struggles to separate you from dozens of similar-sounding competitors. |
| No topical depth | No guides, no FAQs, no use-case blogs, no real problem-solving content. | You never show up when users dig into edge cases or specific scenarios. |
| Inconsistent entity signals | Different names, addresses, descriptions and service lists across the web. | Models can’t be sure you are the same business, so they stay conservative. |
| Content that’s impossible to “chunk” | Walls of text, no headings, no tables, no summaries, no structure. | Even if the facts are there, they’re painful to extract and reuse. |
None of these issues are glamorous. They are boring fundamentals. But they are exactly what makes the difference between “never mentioned” and “routinely cited alongside the big names”.
Real fixes I’ve implemented for clients
To make this less abstract, here are a few anonymised patterns from real projects.
1. Crisis-driven legal support brand
This was a national service in a niche where people are in crisis and spend hours talking to AI about their situation. On paper, it should have been a perfect match for AI visibility. In practice, their website had three major issues:
- Over a hundred blog posts that weren’t properly indexed.
- No consistent “Who we are and who we help” block across content.
- Very little depth around specific scenarios people actually search and ask about.
We fixed the technical problems first so their content could be crawled. Then we introduced a standard introduction in every guide that clearly stated what the brand does and who it supports. Finally, we expanded content around the real situations people described in calls and AI prompts.
Once those changes bedded in, AI tools started citing the brand alongside the best-known names in that space. Nothing mystical happened behind the scenes. We just gave the models something coherent and consistent to work with.
2. A marketing and creative agency in a competitive city
Another example sits at the opposite end of the spectrum. This was a creative agency in a brutal niche: lots of competitors, years of old content that had been deleted and no meaningful social presence. We were effectively starting again.
The owners understandably wanted to know how they could “get into AI” quickly. The honest answer was that they couldn’t. Not without doing the groundwork that should have been there already.
The work here looked like this:
- Rebuilding a full set of service and sub-service pages to match what they actually deliver.
- Making those pages more crawlable and easier to understand at a glance.
- Creating BOFU content first: realistic cost guides, process explainers and comparisons.
- Layering in case studies and topical blogs that matched how clients ask questions in real life.
AI mentions did come, but only after Google signals, entity clarity and basic authority improved. It was slow and sometimes frustrating, which is why I’m always wary of anyone promising “fast AI SEO” in competitive spaces.
3. Local service businesses with almost no digital footprint
The fastest wins I’ve seen have come from small local businesses where the bar is incredibly low. Typical issues include:
- A one-page site with almost no copy.
- No clear list of services, locations or prices.
- No About page, no FAQs and no blog content.
For these businesses, just creating a proper site – with individual service pages, honest price ranges, clear location references and a handful of tightly written FAQs – can make a noticeable difference in both Google and AI tools. Suddenly, when someone asks, “Who are the most affordable [service] providers in [town]?”, there’s a page that actually addresses that question.
Why niche matters far more than people think
One of the biggest mistakes I see is assuming AI visibility is equally important for every business. It is not.
Think about two extremes:
- Emergency trades – leaking roofs, broken boilers, power cuts, burst pipes.
- High-stakes, complex services – family law, financial restructuring, health conditions.
In the first group, the buying cycle is short and emotional. People are standing in the rain with water coming through the ceiling. They will search for a roofer with good reviews, check the first page of Google and make a decision quickly. AI might be involved, but it is not the primary channel.
In the second group, people are scared, overwhelmed and unsure who to trust. They might spend hours asking AI variations of the same question, exploring different options and sense-checking what they’ve been told. In those niches, being mentioned and accurately described in AI tools is much more valuable.
Most businesses sit somewhere between those extremes. The art is in being honest about where your customers actually are, rather than chasing AI visibility just because it’s trending on LinkedIn.
The hard truth about AI citation tools
The SEO tool vendors have, predictably, rushed out dashboards that claim to show “AI citations”, “answer engine visibility” and other shiny metrics.
There is a fundamental problem with all of them: we do not have access to the prompts people type into these tools, or the full paths of their conversations. Without that, the data is always going to be partial at best and misleading at worst.
If you try to micromanage performance based on these snapshots, two things usually happen:
- You spend hours chasing minor fluctuations in a system you cannot see directly.
- You neglect the basics – content, structure, links, brand cohesion – that actually influence both Google and AI tools over time.
I am not against experimentation. I use these tools for directional insight. What I am against is building a strategy around metrics that sound clever but don’t map cleanly to how real people use AI in the wild.
What actually works if you want AI mentions
So if there is no secret “submit to AI” form, what should you do?
Here is the pattern I’ve seen work consistently across different clients and niches.
1. Fix crawlability and indexation first.
If large parts of your site are blocked, broken or unindexed, nothing else matters. AI tools lean heavily on what search engines can already find.
2. Build a full, honest service architecture.
Create dedicated pages for each service, each key audience and each location you genuinely serve. Use plain language your customers would use, not internal jargon.
3. Add real pricing and commercial detail.
You don’t have to publish a rate card, but you should offer ranges, package examples and what affects cost. My SEO audit cost guide, local SEO cost guide and AI SEO pricing guide all follow this pattern.
4. Make your content chunkable.
Use clear headings, short sections, summary boxes, comparison tables and FAQs. AI systems are more likely to reuse information that is already neatly packaged.
5. State who you are and who you help, everywhere.
Add a consistent block that explains your brand, your specialism and your ideal client. This helps both search engines and AI models understand your positioning.
6. Build topical depth around real questions.
Write guides that mirror what people actually ask in consultations, emails and AI chats. Cover edge cases, not just surface-level definitions.
7. Align your footprint across the web.
Make sure your name, address, descriptions and service lists line up across your website, Google Business Profile, directories and social channels.
None of this looks like “AI SEO” in a sales deck. It looks like solid, slightly unglamorous work on structure, clarity and entity signals. That is precisely why it works.
Final word: AI visibility comes after real SEO, not before
If you take nothing else away from this, let it be this: you can’t skip the fundamentals and jump straight to being an AI favourite. Every brand that consistently shows up in tools like ChatGPT and Gemini has already done the groundwork.
- A clear, well-structured website.
- Consistent information across every platform.
- Depth of content around real problems, not generic definitions.
- Honest pricing and concrete service information.
- A recognisable narrative about who they help and how.
If those foundations aren’t in place, chasing “AI visibility” is putting the cart before the horse. A model can only cite what it can confidently retrieve – and retrieval depends entirely on how strong, clear and complete your digital footprint is.
Once the structure is right, AI mentions tend to follow the same pattern as SEO: gradually, then suddenly. You build the architecture first; the citations are a side-effect of getting everything else right.
If you want a realistic, unbiased audit of where you currently stand – in both Google and AI tools – our AI SEO Agency service breaks down exactly how we approach AI visibility for real UK businesses. No gimmicks, no shortcuts – just the modern SEO foundations that AI systems actually trust.
The simplest place to start is with a question I ask every new client: “If someone described your situation to an AI tool today, would your website be the obvious answer?”
