How to make sure ChatGPT and AI Overviews mention you

Table of contents

• What is GEO?
• Why does GEO matter in 2026? Four numbers you need to hear
• The Australian context: we’re further along than you think
• GEO vs SEO: what actually changed?
• What does the research say about GEO?
• What does Google now say about AI search? (May 2026)
• What actually moves the needle in GEO?
• The schema you should actually add, with code
• The 2-hour Melbourne SMB minimum viable GEO plan
• What won’t save you?
• FAQ

Someone in Hawthorn, Melbourne asks ChatGPT “best accountant near me for a small trades business.” ChatGPT doesn’t hand them ten blue links. It hands them a paragraph answer with three named firms and links to the sources it pulled from.

If your accounting practice isn’t one of the three, you don’t exist in that conversation. Not later on page two, not “fighting for the click.” You’re just not there.

That’s the thing most small business owners I talk to haven’t quite clocked yet. It’s not that AI search is coming. It’s already here, your competitors are showing up in it, and the standard SEO playbook alone won’t get you into the answer. That’s what GEO is about.

This is the long version of the conversation I have with every new client who asks me whether they need to “do something about AI.” Yes. But probably not as much as the people selling AI SEO packages want you to think. Here’s the honest version.

What is GEO?

Generative Engine Optimisation (GEO) is the practice of structuring your website content so that AI tools like ChatGPT, Perplexity, Google AI Overviews, and Gemini can find it, understand it, and cite it in their answers. The term comes from a 2023 Princeton paper by Aggarwal, Murahari and colleagues that introduced GEO as a discipline and measured which techniques actually improve citation rates in generative search (arXiv:2311.09735, published in KDD 2024).

The short version: AI engines don’t rank your site like Google does. They read across hundreds of sources, pick the ones that give them a clean, credible, specific answer, and then paraphrase or quote those sources in their response. Getting cited is the new getting ranked. GEO is how you make that more likely.

It isn’t a replacement for SEO. It’s an extension of it. You still want to rank in traditional search. You just also want to be the source the AI tool pulls from when someone asks a question in natural language instead of typing keywords into a box.

Why does GEO matter in 2026? Four numbers you need to hear

Four numbers tell the story.

1. ChatGPT crossed 900 million weekly active users in February 2026. That’s more than double the 400 million from a year earlier (Search Engine Land, Feb 2026). OpenAI also reports 50 million paying consumer subscribers and 9 million paying business users. This isn’t a niche tool anymore. It’s closer to “a quarter of the internet uses ChatGPT every week.”

2. Google AI Overviews now show on 16-48% of searches, depending on the study. Semrush’s research tracked Google AI Overviews appearing on around 16% of US queries in late 2025, with informational queries hit hardest at 39.4% (Semrush AI Overviews study, 2025). BrightEdge’s tracking of specific industries showed AIO surging from 31% of queries in Feb 2025 to 48% by Feb 2026, with education queries jumping from 18% to 83% (Search Engine Land summary, 2026). Use the range. One number alone hides the story.

3. Click-through rates drop 47-61% when an AI Overview appears. Pew Research found users click a traditional search result on just 8% of visits where an AI summary appears, versus 15% where none does — roughly half as often (Pew Research, July 2025). Seer Interactive measured a 61% drop in organic CTR on SERPs with AIO present (Seer, Sept 2025). Ahrefs saw similar patterns in December 2025 (Ahrefs). The upside: brands cited inside the AI Overview itself get 35% more organic clicks than they would without it.

Three separate studies, three different methodologies, roughly the same picture. Being on the page but not cited is worth about half the traffic it used to be.

4. One in four consumers now start their product research with AI, not Google. Adobe’s 2026 AI and Digital Trends report found 25% of customers now cite AI-powered platforms like ChatGPT as a top research tool, with 49% open to AI for product recommendations and 44% for customer service (Adobe, 2026). Gartner forecasts a 25% drop in search engine volume by 2026 due to AI chatbots and virtual agents (Gartner, Feb 2024).

Put them together and the picture is obvious. A growing share of the people who would have clicked on your site from Google are now getting their answer from an AI tool instead. Some still click through. Many don’t. The only way to be in the conversation is to be one of the sources the AI pulls from.

The Australian context: we’re further along than you think

Here’s the bit that surprises my Melbourne clients. Australia isn’t behind on AI search. We’re ahead.

DataReportal’s Digital 2026 Australia report clocks Australian internet users at 26.2 million — 97.1% penetration, effectively the whole adult population online. And a Google/Ipsos survey found 49% of Australians used generative AI in the past year, up from 38% in 2023, with 74% of those using it at work.

Nearly half the adult population is already asking AI tools questions they used to put into Google. A decent chunk of that is happening during the workday, which means local business searches, supplier research, service provider comparisons. Exactly the kind of searches a small business owner in Fitzroy or Coburg wants to show up for.

In my experience auditing Melbourne businesses, roughly half the sites I look at haven’t claimed their Google Business Profile properly, or they claimed it years ago and never touched it again. The foundations that feed AI citations for local queries (GBP, reviews, accurate address and hours) are wide open. If you’re reading this and your plumbing business or café hasn’t sorted its GBP out, the local SEO guide is where I’d start. GEO sits on top of that work, not instead of it.

There’s a real window here. Most Melbourne SMBs haven’t done anything about GEO yet because most haven’t heard of it. Competing on attention is almost always harder than competing on being early.

GEO vs SEO: what actually changed?

A lot of articles on this subject draw a sharp line between GEO and SEO. They’re closer than that. GEO is what you get when you apply a decade of structured-content best practice to a new audience.

The practical overlap is larger than the table suggests. A site that’s fast, structured, well-written, and factually specific tends to do well in both. A site that hides its point under paragraphs of vague marketing copy tends to do badly in both.

If you care about site speed and mobile experience (you should), the same Core Web Vitals work that helps SEO also helps AI crawlers parse your site. If your Google Ads spend is leaking (it probably is), the self-audit checklist in the Google Ads guide will tell you where. None of this is wasted when you start thinking about GEO. It’s all compounding.

What does the research say about GEO?

The foundational paper is Aggarwal et al., “GEO: Generative Engine Optimization”, from a Princeton and IIT Delhi team, published at KDD 2024. It’s the first serious attempt to measure which content techniques actually improve citation rates in generative search, tested across 10,000 queries.

They found that certain techniques boosted visibility in AI responses by up to 40%. These are the nine techniques they tested, strongest performers first:

1. Citing sources. Adding credible references inside your content.
2. Adding statistics. Replacing “a lot of people” with “49% of Australians” (with a source).
3. Using quotations. Including direct quotes from experts or studies.
4. Fluency optimisation. Clear, well-structured writing.
5. Unique wording. Specific phrasing the AI hasn’t seen a thousand times elsewhere.
6. Technical terms. Precise terminology where it suits the topic.
7. Authoritative tone. Writing like someone who actually knows the subject.
8. Easy-to-understand language. Simplifying dense writing so the answer is legible.
9. Keyword stuffing. The one old-school SEO trick they tested — and it didn’t work. Cramming query keywords in barely moved the needle, which is the whole point: AI engines reward substance, not density.

The top three (citing sources, adding statistics, using quotations) did most of the heavy lifting, each adding roughly 30-40% visibility. None of this is a shock if you’ve read good writing advice for the last twenty years. The interesting bit is that a machine learned to prefer it for the same reasons humans do. Specific is better than vague. Sourced is better than asserted. Clear is better than clever.

The practical takeaway: the paper measured content techniques, not tooling. Answer-first formatting, schema markup, and clean HTML structure are all worth doing — I get to them below — but they’re your own good practice, not something this study ranked. Get the writing right first; the rest is incremental on top of that.

What does Google now say about AI search? (May 2026)

In May 2026, Google did something it hadn’t before. It published an official guide to optimising for AI features in Search. For a year the SEO world had been guessing. Now we have Google’s actual position, and it’s blunter than the “AI SEO” industry would like.

The headline: for Google specifically, there is no separate AI ranking system. AI Overviews and AI Mode pull from the same index and the same ranking as regular search. In Google’s words, to be eligible to appear in an AI Overview, “a page must be indexed and eligible to be shown in Google Search with a snippet… There are no additional technical requirements”. Translation: if you’ve done good SEO, you’ve done your Google GEO. There’s no secret AI checklist.

Google also spelled out what you don’t need, which is the more useful half. You don’t need an llms.txt file (“Google Search itself doesn’t use them”). You don’t need to chop your content into tiny chunks for the AI to read. You don’t need to rewrite everything in some special “AI-friendly” voice. And you don’t need special schema markup to show up in AI answers. Most of what gets sold as “GEO” is, for Google, unnecessary.

So where do the GEO tactics actually earn their keep? The other engines. ChatGPT, Perplexity, and Claude don’t follow Google’s ranking, and they don’t behave like each other. ChatGPT leans on consensus sources and clear answers. Perplexity weights recency hard and cites several sources per answer. Claude likes depth and well-structured content. Answer-first formatting, specific sourced statistics, freshness, clean structure: those genuinely help you get cited there, and they make the page better for humans, which is why I’d do them anyway. Just don’t believe anyone who calls them a secret Google hack. Google has now said, on the record, that they aren’t.

The honest version hasn’t really changed. Do excellent SEO for Google, and the same fundamentals (clear answers, specific facts, fresh content) are what get you cited across ChatGPT, Perplexity, and the rest.

What actually moves the needle in GEO?

This is the section where I tell you what I’d actually bother with, because a lot of the AI SEO packages being sold right now are 80% fluff on top of standard fundamentals.

What I’d prioritise

Answer-first formatting in every section. If someone could only read the first two sentences of each H2, they should get the answer. Everything after is supporting evidence. This matches the research and, honestly, just makes the article better.

One-paragraph definitions of what you do. Your homepage, your service pages, your About page. Each should open with a single sentence an AI tool can quote verbatim. “We’re passionate about delivering solutions” is not quotable. “Nathan Schram is a Melbourne-based digital marketing consultant specialising in SEO, GEO, and Google Ads for small businesses” is.

FAQ sections with direct answers. FAQ content maps almost exactly to the kind of question people ask AI tools. A Relixir analysis of 50 domains found FAQPage schema produced the highest citation lift of any schema type at around 28% median. Yes, the firm publishing those numbers sells schema services, so take the magnitude with a grain of salt. The direction is consistent across other studies. One thing to know: Google retired the FAQ rich result in May 2026, so FAQPage markup no longer earns you the expandable Q&A box in Google’s results. The value now is the AI-citation angle and the clean structure, not a Google SERP feature. Write FAQs. Add the schema where the questions are real. The cost is low.

Fresh content with clear dates. Every article I publish has a “last updated” date. AI tools weight recency. A 2024 article about AI search is stale by 2026.

Specific, verifiable numbers. “We’ve helped many businesses” is worthless. “I’ve audited Google Ads accounts for trades, healthcare, real estate, and professional services since 2013” is fact-checkable and citable.

What I’d skip for now

Buying an “AI SEO” package for $2,000 a month. Most of what’s in those is either standard SEO or basic schema implementation. If your site has good structure, clear writing, and current schema, you’ve already done 90% of the work. The rest is measurement, which barely any tool does well yet.

Skip llms.txt. The llms.txt proposal from Jeremy Howard at Answer.AI in September 2024 is a tidy idea: a text file telling AI crawlers what matters on your site. Plenty of sites added one through 2025. Then the evidence came in. In May 2026 Google stated flatly that its Search systems don’t use it, and an Ahrefs analysis of 137,000 domains found 97% of llms.txt files got zero bot requests. John Mueller compared it to the discredited meta-keywords tag. I had this filed under “cheap insurance” a few months ago. It’s now closer to “don’t bother for most sites.” If you want one for the agentic-browsing future, fine, but don’t pay anyone to strategise about it.

Chasing every new “GEO tool.” The tooling market is a mess. Most tools show you mentions in some subset of AI tools, extrapolate wildly, and sell you a dashboard. Wait a quarter. The honest ones will still be around.

The schema you should actually add, with code

Schema markup is where most small business websites miss the easiest win. Microsoft confirmed at SMX Munich in March 2025 that Bing’s AI tools use schema in their indexing. Google was clearer in its May 2026 guide: structured data isn’t required for AI search and there’s no special markup to add, but it’s still worth using because it helps you qualify for rich results in regular search. So add schema because it’s cheap and well-supported for SEO and for engines like Bing, not because it’s a magic AI-citation lever. The cost of adding it is one hour. The downside is zero.

Drop these into the <head> of the relevant pages, inside <script type="application/ld+json"> tags. Test them with Google’s Rich Results Test and Schema Markup Validator. If either throws a red error, fix it before shipping.

Person schema (for a consultant or personal brand)

{
  "@context": "https://schema.org",
  "@type": "Person",
  "name": "Nathan Schram",
  "url": "https://nathanschram.com",
  "jobTitle": "Digital Marketing Consultant",
  "description": "Melbourne-based digital marketing consultant specialising in SEO, GEO, and Google Ads for small businesses.",
  "address": {
    "@type": "PostalAddress",
    "addressLocality": "Melbourne",
    "addressRegion": "VIC",
    "addressCountry": "AU"
  },
  "sameAs": [
    "https://www.linkedin.com/in/nathanschram",
    "https://bsky.app/profile/nathanschram.bsky.social",
    "https://github.com/littlebearapps"
  ]
}

LocalBusiness schema (for a Melbourne SMB)

{
  "@context": "https://schema.org",
  "@type": "LocalBusiness",
  "name": "Your Business Name",
  "image": "https://example.com/storefront.jpg",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "123 Brunswick St",
    "addressLocality": "Fitzroy",
    "addressRegion": "VIC",
    "postalCode": "3065",
    "addressCountry": "AU"
  },
  "geo": {
    "@type": "GeoCoordinates",
    "latitude": -37.7963,
    "longitude": 144.9780
  },
  "telephone": "+61-3-9000-0000",
  "url": "https://example.com",
  "openingHoursSpecification": [{
    "@type": "OpeningHoursSpecification",
    "dayOfWeek": ["Monday","Tuesday","Wednesday","Thursday","Friday"],
    "opens": "09:00",
    "closes": "17:00"
  }],
  "priceRange": "$$"
}

FAQPage schema (still useful for AI, no longer a Google rich result)

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [{
    "@type": "Question",
    "name": "What is GEO?",
    "acceptedAnswer": {
      "@type": "Answer",
      "text": "Generative Engine Optimisation (GEO) is the practice of structuring website content so AI tools like ChatGPT, Perplexity, and Google AI Overviews cite it in their answers."
    }
  }]
}

Organization schema (for a company homepage)

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Your Company",
  "url": "https://example.com",
  "logo": "https://example.com/logo.png",
  "sameAs": [
    "https://linkedin.com/company/your-company",
    "https://twitter.com/yourcompany"
  ]
}

If you want a hand adding schema or doing a proper GEO audit rather than a quick pass, get in touch. I do this for small businesses across Melbourne.

The 2-hour Melbourne SMB minimum viable GEO plan

Here’s what I’d actually do if you gave me a Saturday morning and a Melbourne small business website that’s currently invisible to AI search.

1. Rewrite the homepage opening paragraph (15 min). One sentence that says what you do, who you do it for, and where. No “passionate.” No “innovative.” Just the facts an AI tool could quote.
2. Add Person or Organization schema to every page (20 min). Use the templates above. Test in the Rich Results Test.
3. Add LocalBusiness schema to the homepage and contact page (15 min). Especially if you have a physical location.
4. Write one proper FAQ section (30 min). Five to eight questions real customers actually ask. Direct answers. Add FAQPage schema to the page it lives on.
5. Claim and optimise your Google Business Profile (20 min if you haven’t, 5 min to double-check if you have). Hours, categories, photos, description. This feeds AI citations for local queries directly.
6. Add a “last updated” date to every piece of content (10 min). Recency is a signal.
7. Check your key pages are crawlable and indexable (10 min). No accidental noindex, nothing important blocked in robots.txt, and your main content sitting in the HTML rather than hidden behind JavaScript. This is the one technical thing Google’s own guide says actually gates whether a page can appear in an AI answer.

That’s about two hours. It will not solve every problem. It also won’t leave you any worse off than you are now, and it covers the techniques the Princeton research actually measured as high-impact.

What won’t save you?

Before you hand money to anyone selling “GEO packages,” three honest caveats.

Schema alone isn’t magic. Some studies report strong citation lifts from structured data and FAQ blocks. AccuraCast’s September 2025 analysis found only 1.8% of cited sources used FAQPage schema, and Reddit (the single most-cited source in AI answers) uses almost no Schema.org markup at all. The studies disagree on magnitude. The sensible move is to add schema because it’s cheap and well-supported for regular SEO, not because a single stat is going to transform your citations.

The “AI-first” funnel is still tiny. Most small businesses I work with still get the majority of their traffic from Google, direct visits, and paid ads. AI referral traffic is real and growing, but for a Melbourne plumber right now it’s probably 3-8% of sessions, not 50%. Don’t neglect your Core Web Vitals or your Google Ads setup chasing an AI-only strategy.

The tooling is changing monthly. OpenAI adds features, Google re-pulls AI Overviews from industries they’d previously rolled out to, Perplexity launches new products. The foundations (clear writing, direct answers, structured data, fresh content, real expertise) have survived a decade of Google algorithm changes and will survive this one. The specific tooling around them is going to keep moving.

Do the fundamentals, don’t panic-buy an AI SEO package, and check back in six months. That’s how I’d approach it, anyway.

FAQ

Is GEO different from AEO (Answer Engine Optimisation)?

Mostly not. AEO came first, referring to optimisation for featured snippets and voice assistants. GEO is the broader term for generative AI search. Some practitioners use them interchangeably. I use GEO because it’s the term the Princeton paper established.

Do I need to do GEO if I already rank well in Google?

Yes, because ranking well in Google and being cited in Google AI Overviews are not the same thing. A page can be in position 1 for a keyword and still not appear in the AI Overview for the same query. That said, if your SEO fundamentals are strong, your GEO starting point is already better than most.

What’s the single highest-impact GEO change I can make?

Rewrite the first paragraph of your homepage and key service pages to open with a direct, specific, citable answer to “what do you do and who do you do it for.” Everything else compounds on top of that.

How do I measure if GEO is working?

It’s harder than SEO. Look at referral traffic from AI tool domains (chat.openai.com, perplexity.ai, gemini.google.com) in Google Analytics, brand mentions in AI answers (test by asking ChatGPT and Perplexity questions a customer would ask), and whether your pages appear in Google AI Overviews for your target keywords. No tool does this perfectly yet.

Does llms.txt actually do anything?

For Google, no. In May 2026 Google stated its Search systems don’t use llms.txt, and an Ahrefs study of 137,000 domains found 97% of the files got zero bot requests. The proposal is sensible in theory and might matter for the agentic-browsing future, but as a 2026 SEO or GEO lever it does almost nothing. Treat it as optional, not a growth move.

Can I just use ChatGPT to write my GEO content?

You can, and plenty of people do, but the Search Engine Land 16-month study of 2,000 AI-written articles found they collapsed to just 3% of Google’s top 100 within three months (they clawed back some ground later, but most never returned). Write it yourself, or get a human to edit AI drafts properly. The Princeton research is very clear that specific, factual, well-cited writing wins. Generic AI copy is the opposite of all of that.

What’s the minimum budget for GEO?

Two hours of your time and zero dollars, if you follow the weekend plan above. If you want an expert audit and implementation, a decent local consultant (shameless plug: me) will do a thorough GEO pass on a small business site for a few hundred to a couple of thousand dollars, depending on size. Don’t pay ongoing $2,000-a-month “AI SEO” fees unless someone’s actually proving the traffic.

Is GEO still going to matter in a year?

The specific tooling might change. The fundamentals have mattered for traditional SEO for two decades: clear writing, structured data, direct answers, verifiable facts, current content. Nothing in the research suggests those are going away. If anything, they’re about to matter more.

The bottom line

GEO matters more than the “AI SEO” crowd has earned the right to charge for it, and less than most of the panicked blog posts about “the death of Google” are suggesting. It’s an extension of the work a good SEO should already be doing: clear answers, specific facts, structured markup, current content. Sort those out and you’ll be fine.

If you’d like a proper GEO audit of your Melbourne business website, or help with the schema implementation, get in touch. I work with small and medium businesses across Melbourne on SEO, Google Ads, web design, and the whole mess AI has made of digital marketing strategy. Results, not reports.

Nathan Schram is a Melbourne-based digital marketing consultant with 13 years in tech marketing, sales, and business development across startups and ASX-listed organisations. He writes about Core Web Vitals, Google Ads, and local SEO for Melbourne businesses. Read more at About.

Originally published April 2026; updated June 2026 to fold in Google’s first official AI-search guidance (and to retire the llms.txt advice Google has since said it doesn’t use). Statistics were verified against their original sources at publication. AI search data moves fast, so check the linked sources for the latest figures.