Why small businesses may have more to gain from AI search than they think
There is a slightly strange assumption built into the conversation around AI search: that the businesses most likely to benefit are the ones that already dominate the internet. The household names, the enormous brands, the companies with endless content budgets and years of accumulated authority.
It makes intuitive sense. If an AI is asked to recommend a business, surely it will reach for the businesses it already knows, except that AI search is not quite working that way.
When someone asks ChatGPT for the best family lawyer in Melbourne, the question is not simply who is the biggest? It is who appears to be the most relevant, credible, and useful answer to this particular question?
That creates an interesting opportunity for smaller businesses. All you need to be understandable, relevant and provable, not already famous.
To summarise:
AI systems need information they can find, interpret and use with some degree of confidence. For a small business, that means the foundations still matter: your website needs to be technically accessible, your services and expertise need to be clearly explained, and your business information should be consistent across the places where people and search systems encounter it.
The bigger opportunity, however, sits beyond technical SEO. AI search draws on a much wider information ecosystem, so your reputation is no longer contained neatly within your own website. Industry publications, reviews, directories, podcasts, YouTube, professional associations and genuine discussions elsewhere on the web can all contribute to the picture an AI forms of your business.
You are not trying to become famous. You are trying to become the obvious answer to a particular question.
For a specialist business, that can be a much more achievable proposition.
The problem with being a perfectly good business nobody can find
Small businesses often have an odd relationship with visibility. They may have been operating for ten years, know their customers intimately and possess considerably more expertise than their competitors, yet their website says almost none of this.
Instead, there is a homepage announcing that they are "committed to delivering innovative solutions", a services page full of broad descriptions, an About page written in the third person and perhaps a handful of blog posts that haven't been updated since 2022.
The business itself is perfectly real. The digital version of it is remarkably vague.
This matters because an AI system trying to answer a recommendation query has to construct an understanding of your business from the information available to it. It needs to establish what you do, where you operate, who you serve, what you are particularly good at and whether there is enough evidence to support including you in an answer.
If that information is scattered, contradictory or buried beneath corporate language, the system has less to work with.
Your competitor may simply be better documented, and that is one of the more useful ways to think about GEO.
Just be relevant
Traditional search has encouraged businesses to think in increasingly broad terms. Rank for lawyer, get more traffic, get more backlinks, move up the results. AI search makes the question more conversational and, in many cases, considerably more specific.
A person might ask, "Who are the best boutique hotels in Daylesford for a romantic weekend?" or "Which Melbourne SEO consultants work with small professional services firms?" or "What should I look for in a family lawyer after separation?"
These are requests for an answer, nothing abstract about it. And a small business can be exceptionally relevant to an answer without being particularly famous.
This is where the long tail of search becomes particularly interesting in an AI environment.
You don't need to win the whole category. You need to own the questions that make sense for your business.
Make yourself easy to understand
The technical side of GEO is less glamorous than the headlines suggest, but it is where the work begins.
Your website needs to be accessible to search engines and AI systems, which means reviewing things such as robots.txt, XML sitemaps, indexability, internal linking, canonicalisation, structured data and the way important pages are rendered.
There is also growing discussion around llms.txt, an emerging convention intended to provide AI systems with a concise overview of a website and its important resources. It is worth understanding and, in some cases, implementing, but it should be treated as one piece of the technical picture rather than a shortcut to AI visibility.
The more fundamental question is whether your website makes sense. Can someone land on the homepage and understand what you do? Are your services described in language your customers actually use? Is your location clear?Are the people behind the business identifiable? Do your About, service and contact pages reinforce the same story? Does your structured data reflect what is actually on the page?
If the answer to these questions is no, there is little value in obsessing over the latest AI optimisation tactic. You have a clarity problem first.
Give the machine something worth saying
This is where good editorial thinking and good GEO strategy have considerably more in common than the phrase "optimised for AI" might suggest.
If someone asks how much SEO costs, answer the question. If they want to know how long it takes, give them a realistic timeframe. If they want to understand what a particular service involves, explain it without making them excavate the answer from five paragraphs of brand positioning.
Answer-first writing is useful because it works for people. It is also useful because it creates self-contained passages that an AI system can interpret and potentially draw upon.
The same principle applies to specificity.
"Years of experience" is a fairly empty claim.
"More than 15 years advising Victorian businesses on employment law" gives a reader something they can actually understand.
"Exceptional accommodation" tells us very little.
"A 24-room boutique hotel in the Macedon Ranges, 15 minutes from Daylesford" gives us a location, a category and a concrete proposition.
Specific information is easier to remember, easier to verify, and easier to use.
That is good writing. It also happens to be good GEO.
Your website is only one witness
This is probably the most important shift for businesses thinking about AI search.
Traditional SEO has often treated the website as the centre of the universe. Build the page, optimise the page, get links to the page and watch the rankings move.
GEO asks a slightly broader question: what does the rest of the internet say about you?
Your website might describe you as a leading Melbourne architect. But does an industry association list you as an architect? Has a publication quoted you? Do clients review you? Has your work appeared anywhere credible? Does your Google Business Profile confirm your location? Has anyone interviewed you about the thing you claim to know about?
The more these sources agree, the easier it becomes to establish a coherent entity.
You are effectively giving the AI several witnesses who tell the same story.
That is much more persuasive than asking your own website to make every claim on your behalf.
Don't turn your business into a content factory
This is where the current AI-content conversation can become counterproductive. Once businesses hear that AI search needs more content, the instinct is often to produce more of it: another blog, another FAQ, another generic guide, another 1,500 words about "the benefits of choosing a trusted provider".
That is unlikely to make you more interesting to anyone, including an AI. A better question is: what do you know that your customers actually want to know?
The answer might become a service page, a case study, a short video, an interview, a detailed FAQ, a LinkedIn post, a podcast appearance or an article in an industry publication. The format matters less than the usefulness of the information.
One genuinely insightful article can do more for your authority than ten pieces of content produced simply to satisfy a publishing calendar.
And then there are the places you don't own
AI systems can draw on a wide range of information sources, and different platforms have different retrieval systems and source preferences. That makes third-party visibility increasingly interesting.
Reddit can provide real-world discussion. YouTube provides video, transcripts and descriptions. Industry publications provide editorial authority. Professional directories can reinforce your category and location. Reviews provide customer perspective. Podcasts establish expertise in a conversational format. None of these should become a checklist of places to spam your business.
The point is to participate where it makes sense.
If your customers genuinely use Reddit, contribute something useful. If your industry is discussed through podcasts, offer an informed perspective. If there is a respected professional association in your field, make sure your profile is accurate and complete.
What should a small business actually do?
Start with the fundamentals, then work outwards.
First, make sure your website is technically sound and that your core business information is accurate and consistent. Then review your highest-value pages and ask whether they genuinely answer the questions a potential customer would have before contacting you.
Next, identify the expertise that exists inside the business but hasn't made it onto the internet. Turn that into useful, specific content rather than generic SEO articles.
Finally, look at the rest of your digital presence. Search your own business name. Look at your profiles, reviews, mentions and directory listings. Ask whether an AI system encountering all of this information would come away with a clear understanding of who you are. Then test it.
Take the questions your customers actually ask and put them into ChatGPT, Perplexity and Google's AI search experiences. Record which businesses appear, which sources are cited and what the systems say about you. That is your baseline.
From there, GEO becomes something considerably more useful than a buzzword. It becomes an ongoing visibility practice.
Want to know whether AI search can find you?
BE SEEN can audit how your business currently appears across Google, ChatGPT and other AI search platforms, then identify the practical changes that could make your expertise easier to discover, understand, and recommend.
Reach out to get started.
In conversation
Do I really have a chance if I'm a small business?
Yes, particularly when the question is specific. AI search has created more opportunities for businesses to be relevant to narrow, local and specialist queries, where being the biggest brand isn't necessarily the same thing as being the best answer.
So I don't need thousands of backlinks?
Backlinks remain useful, particularly when they come from relevant and authoritative websites, but they are only one part of the picture. For a small business, the combination of strong first-party content, credible mentions, accurate business information and genuine expertise can create a much more useful digital footprint than a large collection of unrelated links.
What makes something worth citing?
Usually, it is information that is relevant to the question, clearly expressed, specific enough to be useful and supported by a credible source. The easier you make it for an AI system to understand exactly what a passage is saying, the more useful that passage becomes as a potential source.
Should I be creating an llms.txt file?
It is worth understanding, but don't mistake it for the main event. Technical accessibility, useful content, structured information and a credible wider digital presence are considerably more important foundations for AI visibility.
Do I need to be on Reddit and YouTube?
Only if those channels make sense for your audience and expertise. The goal is not to collect platforms; it is to make your expertise discoverable in the places where relevant conversations are already happening.
How long does this take?
Some technical changes can be reflected relatively quickly, while content, authority and third-party signals generally take longer to compound. There is no honest universal timeline for appearing in AI answers, which is why establishing a baseline and monitoring a consistent set of prompts is more useful than promising a particular result by a particular date.
Can an SEO consultant guarantee that ChatGPT will recommend me?
No. AI systems change, their retrieval processes differ, and individual answers can vary. A good GEO strategy can improve the quality and quantity of signals surrounding your business, but nobody can responsibly guarantee a particular recommendation from an AI platform.