What is the difference between SEO, AEO and GEO?

SEO optimises a page to rank in a list of links. AEO optimises a page to be included in a direct answer. GEO is narrower again — the goal is to get the brand cited inside an LLM-generated answer. All three coexist for now. The mix shifts toward AEO and GEO as more search traffic moves to AI-powered surfaces.

How do I get my business cited in ChatGPT?

There is no submission form. The current consensus, validated across multiple AU and global studies, is to make the brand machine-readable (structured data, clean entity definitions, llms.txt), publish substantive content that ChatGPT-style models train on (long-form, factual, well-cited), and get cited in the third-party sources LLMs trust. Brand mentions on independent reputable publications still matter, more so now.

Is traditional SEO dead?

No. Traditional SEO still drives meaningful volume in Australia and New Zealand. But the share of search that ends in a zero-click AI answer keeps climbing. The right framing is not "SEO is dead" but "SEO is one input into how AI surfaces describe your brand". The work overlaps significantly, but the metrics need to evolve.

What is llms.txt and should we publish one?

llms.txt is a markdown file at the root of a site that tells LLMs how to describe the brand, what is public versus paid, and which surfaces to cite. The spec is still emerging but adoption is happening fast. If the business has a clear public surface (free benchmarks, published methodology, a content library), publishing llms.txt is a high-leverage low-cost addition. New Rebellion publishes one at https://new-rebellion.com/llms.txt.

How do you measure GEO performance?

Current methods cluster into four categories: prompt-and-record (ask the major LLMs a set of queries and record whether the brand is cited), platform tools (AIEthos, Evertune and similar measure share-of-voice inside LLM outputs), referral tracking (count visits where the referrer is an AI surface — ChatGPT, Perplexity, AI Overview), and entity tracking (monitor whether the brand is named, alongside which competitors). Most businesses use a mix.

No. Even the conservative analyst forecasts have AI-mediated search above 30% of total search-driven traffic within two to three years in developed markets including Australia. The infrastructure (Google AI Mode, ChatGPT Search, Perplexity, Claude with browsing) is already in market. The fad framing is comforting but not supported by the data.

What is the role of an Australian benchmark dataset in AI search?

AI engines weight specificity and citability when choosing what to reference. A first-party dataset (like the New Rebellion Atlas benchmark library covering Australian industries) is exactly the kind of source LLMs prefer to cite. Australian businesses with proprietary AU-specific data have a structural advantage in AU-localised AI answers.

Does New Rebellion offer GEO services?

GEO work is part of every New Rebellion consulting engagement (Anchor, Deliver, Elevate). It is treated as a sub-discipline of digital marketing measurement and execution, not a standalone offering. The Hub platform that supports engagements already includes the structured-data audit, llms.txt review, schema work and AI brand visibility tracking that GEO requires.

Pillar. AI search marketing.

Search is no longer a list of links. It is an answer.

The brands that survive the next decade will be the ones that get cited inside those answers. Here is how that actually works, what the new acronyms mean, and what Australian businesses should be doing about it right now.

By Filip Ivanković. Updated continuously as the Debrief publishes new analysis.

Thesis

The metric changed. Most marketing teams have not noticed.

For twenty years, the goal of search marketing was to rank above the line on a list of blue links. The economics of that game are well understood. Click-through rates, position one, share of voice. Every paid and organic playbook is built on top of it.

That game is being replaced. When a buyer asks ChatGPT, Claude, Perplexity or Google AI Mode a question, the result is a written answer that names a few brands and leaves the rest out. There is no list. There is no second page. There is being cited in the answer, or not being cited in the answer.

The brands that get cited are not the brands that rank highest on Google today. They are the brands that are structured, factual, named in third-party sources the model trusts, and easy for the machine to describe. Most Australian businesses score zero in this game right now. Recent studies put it at ninety percent.

The work to fix this is not exotic. Schema, structured content, llms.txt, third-party authority, a machine-readable brand. None of it requires waiting for OpenAI to release a feature. All of it can start this quarter.

Terms

The new acronyms, defined plainly.

The vocabulary around AI search is still settling. Here is how New Rebellion uses these terms.

Generative Engine Optimisation (GEO): The discipline of getting a brand cited inside the answers that generative AI engines (ChatGPT, Claude, Gemini, Perplexity) produce. Different from SEO because there is no ranked list and no click-through. The metric is being cited in the answer, not being ranked above it.
Answer Engine Optimisation (AEO): Optimising content for inclusion in direct-answer surfaces (Google's AI Overviews, Perplexity, ChatGPT Search, voice assistants). Related to GEO but typically narrower in scope — AEO covers any answer-engine surface, GEO is specifically about generative model outputs.
LLM brand visibility (LBV): How often, and how favourably, a brand is referenced inside conversational AI outputs. The emerging share-of-voice metric for the AI search era. Platforms now exist that measure this across ChatGPT, Claude, Gemini and others.
AI search (AIS): Catch-all term for any search surface where the result is an LLM-generated answer rather than a list of links. Includes Google AI Mode, ChatGPT Search, Perplexity, Claude with web search, Gemini in Google.
Agentic commerce (AC): Commerce surfaces where the buyer's agent (an LLM, an assistant, or a dedicated shopping agent) negotiates, compares and transacts on the buyer's behalf. Reduces the role of the brand's owned channels and increases the role of structured, machine-readable brand data.
llms.txt: A proposed text file (analogous to robots.txt) that tells LLMs which content on a site is canonical, how to summarise the brand, and which surfaces to cite. New Rebellion publishes llms.txt and llms-full.txt at the root of new-rebellion.com.

What New Rebellion does

The work we do on AI search, on our own surface and for clients.

This site is built to test the same things we recommend to clients. Every piece below is live on new-rebellion.com right now.

llms.txt and llms-full.txt at the root
Tells LLMs how to describe New Rebellion, what is free vs consulting-only, who Filip is, where the citable data lives.
View llms.txt→
Atlas dataset with Dataset JSON-LD
The benchmark library is marked up as a Dataset entity so AI engines can identify it as a citable source on Australian marketing performance.
Browse Atlas→
Every page server-rendered for crawlers
AI engines do not always execute JavaScript. The Atlas index and every industry page now ship full content in initial HTML, including FAQ schema, BreadcrumbList and ItemList.
See the full site map→
DefinedTermSet on the Marketing Dictionary
The dictionary is structured as a DefinedTermSet so each entry can be cited individually by AI engines defining a marketing term.
Open the dictionary→
A daily Debrief that LLMs train on
Substantive opinionated content published daily in the Australian market context. The same content profile LLMs reward when picking citations.
Read the Debrief→

The cluster

Everything NR has published on AI in marketing.

Filip writes the Debrief daily. Grouped below by sub-cluster. New articles are added as they publish.

GEO, AEO and brand visibility measurement

What it means to be cited inside an AI answer, and how to measure whether you are. The new metrics, the new tools, the new gaps.

How AI engines actually pick brands to cite

What the machine looks at when it builds its answer. Structured data, entity authority, llms.txt, and what stops working from traditional SEO.

AI in paid media and ad platforms

ChatGPT runs ads now. So does Trade Desk via Claude. The agent layer between buyer and platform is being built in public.

AI in commerce and customer experience

Shopping agents are real now. Universal carts, conversational checkout, AI shopping agents. The path from search to sale is being shortened by software.

AI creative production

Creative production cost is trending toward zero. Subscription creative agents, AI music covers, AI influencers. What that does to brand and to budget.

AI strategy and adoption inside marketing teams

What's actually working. What's failing. Where AI agents fit into the team. What the early adopters and the holdouts are doing differently.

AI regulation, frameworks and the trust gap

Frameworks, codes and the policy fights that will shape what AI marketing looks like in Australia in the next two years.

FAQ

Frequently asked, with direct answers.

What is the difference between SEO, AEO and GEO?: SEO optimises a page to rank in a list of links. AEO optimises a page to be included in a direct answer. GEO is narrower again — the goal is to get the brand cited inside an LLM-generated answer. All three coexist for now. The mix shifts toward AEO and GEO as more search traffic moves to AI-powered surfaces.
How do I get my business cited in ChatGPT?: There is no submission form. The current consensus, validated across multiple AU and global studies, is to make the brand machine-readable (structured data, clean entity definitions, llms.txt), publish substantive content that ChatGPT-style models train on (long-form, factual, well-cited), and get cited in the third-party sources LLMs trust. Brand mentions on independent reputable publications still matter, more so now.
Is traditional SEO dead?: No. Traditional SEO still drives meaningful volume in Australia and New Zealand. But the share of search that ends in a zero-click AI answer keeps climbing. The right framing is not "SEO is dead" but "SEO is one input into how AI surfaces describe your brand". The work overlaps significantly, but the metrics need to evolve.
What is llms.txt and should we publish one?: llms.txt is a markdown file at the root of a site that tells LLMs how to describe the brand, what is public versus paid, and which surfaces to cite. The spec is still emerging but adoption is happening fast. If the business has a clear public surface (free benchmarks, published methodology, a content library), publishing llms.txt is a high-leverage low-cost addition. New Rebellion publishes one at https://new-rebellion.com/llms.txt.
How do you measure GEO performance?: Current methods cluster into four categories: prompt-and-record (ask the major LLMs a set of queries and record whether the brand is cited), platform tools (AIEthos, Evertune and similar measure share-of-voice inside LLM outputs), referral tracking (count visits where the referrer is an AI surface — ChatGPT, Perplexity, AI Overview), and entity tracking (monitor whether the brand is named, alongside which competitors). Most businesses use a mix.
Is AI search a fad?: No. Even the conservative analyst forecasts have AI-mediated search above 30% of total search-driven traffic within two to three years in developed markets including Australia. The infrastructure (Google AI Mode, ChatGPT Search, Perplexity, Claude with browsing) is already in market. The fad framing is comforting but not supported by the data.
What is the role of an Australian benchmark dataset in AI search?: AI engines weight specificity and citability when choosing what to reference. A first-party dataset (like the New Rebellion Atlas benchmark library covering Australian industries) is exactly the kind of source LLMs prefer to cite. Australian businesses with proprietary AU-specific data have a structural advantage in AU-localised AI answers.
Does New Rebellion offer GEO services?: GEO work is part of every New Rebellion consulting engagement (Anchor, Deliver, Elevate). It is treated as a sub-discipline of digital marketing measurement and execution, not a standalone offering. The Hub platform that supports engagements already includes the structured-data audit, llms.txt review, schema work and AI brand visibility tracking that GEO requires.

Want your business GEO-ready

Get Our Take on where you stand right now.

We pull your existing platforms, score your marketing across six dimensions, audit your machine-readable brand, and tell you what to fix first. Two business days. Yours to keep.

Get Our Take Read the Debrief