Customers are asking ChatGPT, Google's AI Overviews and similar tools to recommend a tradie, accountant or clinic. Businesses those tools can't understand risk being left off the shortlist.

A homeowner with a leaking roof used to type "roofer near me" into Google and scroll through a page of links. Now many type a full question instead. Which roofers near me handle storm damage? Who is licensed? Roughly what will it cost? Then they read an answer written by AI before clicking anything at all.

For service-based businesses, that shift matters. The customer may never see a list of ten websites. They may see two or three business names in a summarised answer, and choose from those.

It is a change Search Studio, an SEO agency in Toowoomba, has been watching play out across trades, health and professional services. The businesses named in those answers are not always the biggest or the best reviewed. They tend to be the ones whose information is clearest, most consistent and easiest for AI tools to trust. That is the territory of AI SEO, also called generative engine optimisation, or GEO.

What AI search actually is

Traditional search returns a ranked list of pages. The customer does the comparing.

AI search does some of the comparing for them. Google's AI Overviews, ChatGPT, Gemini, Perplexity and Microsoft Copilot read information from websites, business listings, reviews and other sources, then write a single answer. Sometimes that answer cites its sources. Sometimes it simply names businesses.

The practical difference is simple. In a list of ten links, a business on the first page has a chance. In a three-sentence answer, a business is either mentioned or it isn't.

Why service businesses feel it most

Service enquiries are naturally question-shaped. People want to know who does the job, whether they cover a particular suburb, how long it takes, what it roughly costs and whether they are licensed or accredited.

These are exactly the questions AI tools are built to answer. A product search might still end on a retailer's site. A service search can end inside the AI answer itself, with the customer picking up the phone without ever visiting a website.

Consider a simplified illustration. Two electricians in the same town do similar work. One has a website with a single generic services page and no mention of the suburbs it covers. The other has a clear page for each service, says plainly where it works, explains its process and has consistent details across its Google Business Profile and directories. When someone asks an AI tool for an electrician who installs EV chargers in that town, the second business gives the tool far more to work with.

Neither business has done anything wrong. One has simply made itself easier to understand.

AI SEO is not a replacement for SEO

There is a temptation to treat GEO as a brand new discipline that makes everything before it obsolete. It doesn't.

AI tools still rely heavily on the open web. Pages that are well built, easy to crawl, genuinely useful and trusted by other sites remain the raw material those tools draw on. A business with weak SEO foundations rarely becomes a favourite of AI search overnight.

What AI SEO adds is a sharper focus on clarity and consistency. It asks whether a machine reading a business's online presence can confidently answer three questions: what does this business do, where does it do it, and why should anyone trust it?

For businesses weighing up where to start, a structured approach to AI search optimisation usually begins with an honest audit of those three questions before anything else is changed.

What makes a business easier for AI to recommend

Clear service and location pages. One page per core service, written in plain language, works better than a single page listing everything. If a business serves several towns or suburbs, it should say so clearly, without stuffing place names into every sentence.

Answers to real customer questions. The questions customers ask on the phone are the questions they now ask AI. Price ranges, timeframes, what happens on the first visit, what a quote includes. Businesses that answer these openly on their own site give AI tools something reliable to quote.

Consistency across the web. Business name, address, phone number, opening hours and services should match everywhere: the website, Google Business Profile, industry directories and social profiles. Conflicting details create doubt, and AI tools tend to favour information they can confirm from more than one place.

Reviews with substance. A pile of five-star ratings with no detail says little. Reviews that mention the actual job, the suburb and the experience give AI tools useful context. Asking happy customers to describe what was done, rather than just leave a rating, is a small change that helps.

Mentions on trusted third-party sites. Coverage in local news, membership of industry associations, sponsorship of community groups and listings with relevant bodies all help establish that a business is real and active. These signals have always mattered for SEO. They matter more when an AI tool is deciding which names to put forward.

Structured data. Schema markup is code that labels key facts on a page, such as the business type, location, services and reviews. Customers never see it. Search engines and AI tools read it, and it removes guesswork.

How to check where a business stands

There is no perfect dashboard for AI visibility yet. Measurement tools are improving, but they are still catching up.

The simplest test costs nothing. Ask the questions a customer would ask, in ChatGPT, Gemini and Google, and note what comes back. Is the business mentioned? Are competitors mentioned instead? Is the information about the business accurate, or out of date?

It is also worth adding "AI tool or chatbot" as an option when asking new customers how they found the business. It won't capture everything, but over a few months it shows whether that channel is growing.

What to avoid

As with any new channel, shortcuts have appeared quickly. Some are a waste of money. Some carry real risk.

Mass-producing thin AI-written pages tends to dilute a website rather than strengthen it. Buying fake reviews breaches platform rules and Australian Consumer Law. And any provider promising guaranteed placement in ChatGPT or AI Overviews is promising something nobody controls.

AI answers change from one query to the next. The goal is not to game a single answer. It is to become the business that keeps turning up because the evidence supports it.

The bottom line

AI search does not change what customers want from a service business. They still want someone competent, local, available and trustworthy. What has changed is who does the first round of shortlisting.

For smaller and regional businesses, that is not all bad news. AI tools are less impressed by advertising budgets than by clear, consistent and credible information. A well-run local business that explains itself properly online can hold its own against much larger competitors.

The businesses that do well in AI search will mostly be the ones that already do the basics well, and make sure the internet knows about it.