More people now ask ChatGPT, Gemini or Google’s AI Overviews for a recommendation before they ever click a website. They type something like “who does affordable web design near Bognor Regis” and read the answer the assistant gives them. If your business is named in that answer, you get the enquiry. If you are not, you never existed as far as that person is concerned.
We covered the shift itself in an earlier piece on how AI search is changing the way people find your business. This post is the practical follow-up. Here is what actually makes an AI system pick your name over a competitor’s, and what a small business can do about it this month.
Why AI recommends some businesses and skips others
An AI assistant is not ranking ten blue links. It is writing a short answer and deciding which sources to pull from. It leans towards content that is easy to lift out of a page, easy to verify against other sources, and clearly written by someone who knows the subject.
That last part matters more than people expect. These systems are trying to avoid being wrong. So they favour businesses that give clear, specific, checkable information over ones that talk in vague marketing language. “We deliver quality solutions” tells the machine nothing. “We build WordPress sites for small businesses in West Sussex, with projects usually starting around a few thousand pounds” gives it something concrete to repeat.
Getting ranked was the old goal. Getting quoted is the new one. A page can be cited by an AI answer even when it sits on page two of Google, because the AI is choosing sources by clarity and trust rather than by position alone.
Answer the questions people actually ask
The single most useful thing you can do is write plain answers to the real questions your customers ask. Not keyword phrases. Actual questions, in the words a person would use.
Think about what someone types or says out loud. “How much does a small business website cost?” “Do I need to pay monthly for a WordPress site?” “How long does an SEO campaign take to show results?” Each of those deserves a proper answer on your site, and the answer should come first, before the waffle.
A good pattern looks like this. Ask the question as a heading. Give the direct answer in the first sentence or two, in about 40 to 60 words. Then expand underneath for the reader who wants detail. AI systems tend to grab that opening block, so if your answer is buried three paragraphs down, it gets missed.
- Lead with the answer, then explain. Never make the reader dig for it.
- Use the customer’s language, not your internal jargon.
- Give real numbers, timescales and specifics wherever you honestly can.
- Keep each answer able to stand on its own, so it makes sense lifted out of the page.
An FAQ section on your key service pages does a lot of heavy lifting here. So does a well written blog that treats each post as the definitive answer to one question. This is where consistent publishing pays off, and it ties directly into wider search engine optimisation work rather than being a separate job.
Structure your pages so a machine can follow them
AI systems read structure. Clear headings, short paragraphs, proper lists and tables all help the machine understand what your page is about and which bit answers which question.
A wall of text is hard to parse. A page with descriptive H2 and H3 headings that match how people phrase things is easy. If you sell to different sectors or cover different services, give each its own clearly labelled section rather than mashing everything into one long ramble.
Tables earn their place too. If someone asks “what is the difference between a starter website and a full custom build”, a simple comparison table is far easier for an AI to quote than the same information written as flowing prose. The same goes for step by step processes, which work better as numbered lists.
Schema markup gives the AI facts it can trust
Schema is code you add behind the scenes that spells out your details in a format machines read directly. Your business name, location, opening hours, services, reviews, the author of an article, the questions in an FAQ. It removes the guesswork.
For a local business this is one of the higher value jobs on the list, because it feeds the exact facts an assistant needs to recommend you confidently. We explain the whole thing in plain English in our guide to what schema markup is and why your site needs it. The short version: FAQ schema, Organisation or LocalBusiness schema, and Review schema are the ones most worth having in place.
You do not need to hand code it. Most WordPress SEO plugins can add the main types for you, and it is a normal part of how we set up a site. The point is that it exists and describes you accurately.
Reviews and consistent listings are your credibility signals
AI assistants cross check. When they weigh up whether to recommend you, they look at whether your name shows up consistently across the web and whether real people say good things about you.
Two practical jobs come out of this. First, your Google reviews. A steady flow of genuine, recent reviews is one of the clearest trust signals a machine can read, and it helps your traditional local rankings at the same time. If you are not actively asking, start. Our guide on how to get more Google reviews walks through a simple system for it.
Second, consistency. Your business name, address and phone number should match everywhere they appear, from your website footer to your Google Business Profile to any directory you are listed in. Conflicting details make an AI less sure it has the right business, and uncertainty is the enemy of a recommendation.
Get genuinely referenced on sites the AI already reads
Here is the part most small businesses miss. AI systems often cite you through other people’s websites, not your own. A mention in a local news piece, an industry roundup, a supplier’s partner page or a relevant forum thread can carry more weight than another page on your own blog.
So the aim is to be genuinely referenced by others. Not spammy link building. Real mentions earned by being useful, quotable or newsworthy. A few ways small businesses do this:
- Get involved with local press and community sites when you have something worth saying.
- Contribute a genuinely helpful answer or guest piece to an industry publication.
- Publish original observations or small bits of data from your own work, which others then cite.
- Keep your profiles current on the directories and review platforms that matter in your field.
This is slower than the on-page work, but it is what closes the gap on bigger competitors over time. It is the difference between an AI knowing you exist and an AI being confident enough to name you.
How to check whether any of this is working
You cannot improve what you never look at. Once a month, take your ten or fifteen most important questions and ask them across ChatGPT, Gemini and a normal Google search that triggers an AI Overview. Note whether you get mentioned, who does if you do not, and which page of theirs got pulled in.
Do that for a few months and a pattern appears. You start to see which competitors the AI trusts, what kind of content it favours, and where your own gaps are. That tells you what to write next far better than guessing.
Where to start
If this feels like a lot, it does not all have to happen at once. Pick the questions your best customers actually ask, answer them clearly on your site, make sure your schema and reviews are in order, then work on being mentioned elsewhere over time. Good old fashioned SEO is still the foundation under all of it. The AI layer sits on top and rewards clarity, honesty and specifics.
That happens to be how we like to build and market sites anyway. If you want a hand getting your business into these AI answers, or you are not sure where your gaps are, feel free to get in touch and we will take a look.
Frequently asked questions
Yes. AI systems choose sources by clarity, structure and trust rather than pure size, so a small business with clear answers and consistent details can be cited alongside much larger competitors. Being genuinely useful and easy to read matters more than having a huge budget. The businesses that miss out are usually the ones with vague, hard to parse pages.
Traditional SEO is about ranking your pages in the list of results. AI search is about getting your business quoted inside the answer the assistant writes. The foundations overlap heavily, but AI search puts extra weight on plain answers, clear structure, schema and being referenced on other trusted sites. Good SEO is still the base you build on.
It is not strictly required, but it helps a lot. Schema spells out your facts, like your location, services and reviews, in a format machines read directly, which removes guesswork when an AI decides whether to recommend you. For a local business, LocalBusiness, FAQ and Review schema are the most useful to have in place. Most WordPress SEO plugins can add them for you.
Reviews act as a credibility check. A steady stream of recent, genuine reviews signals to an AI that real people trust you, and it strengthens your local search rankings at the same time. Consistency helps too, so make sure your business details match across your site, Google Business Profile and any directories. Actively asking happy customers for a review is the simplest way to build this up.
Take your ten to fifteen most important customer questions and ask them across ChatGPT, Gemini and a Google search that shows an AI Overview. Note whether you are mentioned, who is if you are not, and which page got pulled in. Repeat this monthly and you will spot patterns in what the AI trusts, which tells you what to fix or write next.
