In June, Roy Morgan reported that 13.6 million Australians, 58% of everyone aged 14 and over, had used an AI tool like ChatGPT, Gemini, Copilot or Claude in the previous four weeks. ChatGPT alone accounted for 10.5 million of them. (Roy Morgan, 2 June 2026)
That number turns up in my client conversations in a very practical form. Someone asks ChatGPT to recommend a firm like theirs, in their own town, just to see. Their firm doesn’t appear. A competitor half their size does.
The question that follows is always some version of: what have they got that we haven’t?
It’s the question I’m being asked more than any other this year. And more often than you’d expect, part of the answer is sitting in code nobody at the firm has ever looked at or understood.
Schema, in plain English
Schema (also called structured data) is a layer of code on your website that labels the facts about your business, so machines don’t have to guess.
A person reading your About page can work out that Jane is a director, that you’re based in Wollongong and that you specialise in SMSF advice. A machine reading the same page sees words. Schema tells it: this is an organisation, this is its address, this person works here, this is a service it offers.
Google has used schema for years. The search indexes that AI tools draw on use it too. When ChatGPT, Claude or Google’s AI Overviews assemble an answer, they lean on sources they can understand and trust. Clearly labelled, consistent information makes you easier to understand. That makes you easier to cite.
You’ll hear AEO (answer engine optimisation) and GEO (generative engine optimisation) thrown around. Don’t let that muddy the waters. Both describe the same goal: being the business an AI tool is confident enough to name. The foundations are the ones good SEO always needed, with a lot less room for vagueness.
What most professional services sites are missing
The pattern we see most often goes like this. A plugin added a basic Organisation block when the site was built, and nobody has touched it since. It tells the world the firm exists. It says nothing about who runs it, what it does or where it works.
For an accounting, legal, financial planning or consulting firm, these are the gaps that matter most:
- The right business type. Not just “Organisation”, but the specific type that fits (AccountingService, LegalService, FinancialService or ProfessionalService) with your address, phone, service area and links to your Google Business Profile and LinkedIn.
- Your people. Clients in professional services buy judgement, and judgement belongs to people. Yet most firms don’t mark up their directors and principals at all. Person schema connects a name to a role, credentials and a LinkedIn profile.
- Your services. Each core service page labelled as a service your firm provides, so a machine knows SMSF advice is something you do, not something you once wrote a blog about.
- Your answers. FAQ markup on pages where you genuinely answer client questions. Google stopped showing FAQ dropdowns in results for most sites in 2023, so this isn’t about how you look in search. It’s about labelling your answers clearly for anyone (or anything) reading the page.
- Your articles. Blog posts marked as articles with a named author, linked back to that author’s Person details. That’s how your expertise gets attributed to you.
None of this is visible to your clients. All of it is visible to the machines deciding whether to mention you.
Schema is not a magic switch
Here’s the uncomfortable bit. You can add every schema type in the book and still not get named.
Schema labels what’s already on your site. If your service pages are thin, your people are invisible, and nobody outside your network ever mentions you, Schema just labels a thin site very accurately.
There’s also genuine debate about how directly each AI tool reads structured data. Some lean heavily on search indexes that use it. Others work more from the visible text on the page. Nobody outside those companies knows the exact weighting, and anyone promising you a ChatGPT recommendation is guessing.
What does hold up is this. AI tools favour businesses they can describe with confidence. That confidence comes from consistent facts across your website, Google Business Profile, and LinkedIn; service pages that answer real questions; named experts with real credentials; and other credible sites talking about you.
Schema ties those signals together. It doesn’t replace them.
The 15-minute check you can do this afternoon
- Ask the AI tools (5 minutes). Open ChatGPT and Claude and ask the question your best client would ask: “Who’s a good [your service] in [your town]?” Note who gets named and how they’re described. Then ask about your own firm by name. If the description is vague, out of date or wrong, that’s your first clue.
- Test your schema (5 minutes). Paste your homepage and your main service page into Google’s Rich Results Test, then the Schema Markup Validator for the full picture. Look at what’s actually there. A lone Organisation block, or nothing at all, means machines are working out the rest on their own.
- Check your facts match (5 minutes). Compare your business name, address, phone and how you describe your services across your website, Google Business Profile and LinkedIn. Small differences – Pty Ltd on one, an old phone number on another – make it harder for a machine to be sure it’s looking at the same business.
If step two came back thin, you’re in good company. It’s the most common gap we find.
Where to start
The firms getting named aren’t always the best firms. They’re the ones that have made themselves easy to understand.
If an AI tool can’t work out who you are, it will recommend someone it can.
So – when did you last ask ChatGPT about your own business? Try it this afternoon. I’d love to hear what it said.
If you’d like us to take a look, our SEO & AEO Readiness Audit covers technical health, site structure, content and schema, and gives you a prioritised plan for what to fix first. Or book a 15-minute call and we’ll talk through where your site sits right now.
