LLM Optimisation
SEOAlso: LLMO · LLM SEO · Generative Engine Optimisation · GEO
Quick definition
LLM optimisation is the practice of structuring your content, brand signals and authority so that large language models (LLMs) like ChatGPT, Claude and Gemini are more likely to cite, recommend or describe your business accurately when users ask relevant questions. It is sometimes called Generative Engine Optimisation (GEO) or LLMO.
How it varies across Australia
Most Australian businesses have no deliberate LLM optimisation strategy. The ones appearing most consistently in AI-generated answers tend to be those with strong editorial authority, structured data on their sites and genuine third-party coverage rather than those who recently added AI-specific copy.
See digital maturity and acquisition signals across Australian industries →The main levers
How clearly your brand, products and expertise are defined across the web, structured data and your own content.
Mentions, citations and coverage from sources LLMs already trust: publications, directories, review platforms.
Content written to directly answer questions, not to rank for a keyword, using the language your audience uses.
Schema markup that helps models parse what your business does, where it operates and who it serves.
Outdated or contradictory claims on your own properties reduce the probability a model cites you correctly.
What it actually means
Traditional search engine optimisation (SEO) works by getting pages to rank on a results page a human then clicks through. LLM optimisation works differently. There is no results page. The large language model reads across a vast body of training data and live retrieval sources, synthesises an answer, and either names you or does not. The click may never happen.
This changes the game in ways most marketers have not fully absorbed. The goal is not to rank. The goal is to be the entity the model reaches for when a relevant question comes up. That means the model needs to understand clearly what you do, who you serve and why you are credible. If your entity is fuzzy, contradicted by third-party sources, or simply absent from the sources models draw on, you get omitted.
The tactics that appear to help are familiar to anyone who has done serious technical SEO or content strategy. Clear, well-structured content that directly answers specific questions. Consistent entity definition across your site, your Google Business Profile, structured data markup, and external mentions. Genuine coverage from publications and directories that models weight heavily in their training and retrieval. None of this is a hack. It is the unglamorous work of being a legible, credible presence on the internet.
What is genuinely new is the question of measurement. In SEO, rankings and organic traffic tell you whether the work is landing. In LLM optimisation, the feedback loop is messier. You can probe models manually, use emerging monitoring tools, or track whether branded search and direct traffic shift over time. Attribution here is harder than anything you will encounter with paid media or traditional organic search.
LLM optimisation is not a new channel. It is the consequence of doing the basics well enough that machines trust you.
How it shows up
LLM optimisation shows up in whether AI assistants mention your brand accurately when asked relevant questions. Test it by prompting ChatGPT, Claude, Gemini and Perplexity with the questions your prospective customers are most likely to ask. Does your brand appear? Is the description accurate? Are competitors named instead?
For businesses with direct-response goals, the downstream signal is whether branded search volume or direct traffic grows as AI tool usage grows in your category. These are imperfect proxies, but they are the closest measurable signal most businesses have access to today.
The Australian context
Australian businesses face a compounding challenge with LLM optimisation. The training data for most major models is heavily weighted toward US and UK sources. Australian brands, regulators and market dynamics are underrepresented, which means models have weaker priors about Australian entities to begin with.
This cuts both ways. It means Australian businesses that do invest in editorial authority, structured data and third-party coverage from Australian publications can stand out more sharply in a less-crowded field. Local context matters. An Australian financial services brand with consistent coverage in Australian Financial Review, ASIC-adjacent directories and well-structured schema may achieve stronger model presence than a generic global competitor, even without the global brand recognition.
The Privacy Act and Australian Consumer Law (ACL) implications of AI-generated recommendations are still being worked out. Businesses in regulated categories like financial advice, health and legal services should be especially careful about what model-cited content says about their services.
Where people get this wrong
Related terms
Common questions
Is LLM optimisation the same as SEO?
Related but not identical. Traditional SEO targets ranking algorithms that return a list of pages. LLM optimisation targets the training data and retrieval systems behind AI assistants that synthesise answers directly. The underlying signals overlap heavily, but the feedback loop, measurement and end state are different.
How do I know if an AI tool is recommending my business?
Manual prompting is the most accessible starting point. Ask ChatGPT, Claude, Gemini and Perplexity the questions your best customers ask before buying. Note whether your brand appears, what it says and whether the description is accurate. Repeat monthly. Specialist monitoring tools are emerging but most are early-stage.
Can I get my business removed or corrected in an LLM's output?
Not directly. You cannot file a removal request with an LLM the way you can with Google. The indirect route is fixing inaccuracies at the source: update your own site, correct third-party listings and get accurate coverage published. Models re-train and re-retrieve over time, so corrections propagate, slowly.
How much should I invest in LLM optimisation right now?
Enough to do the fundamentals well, clear entity definition, accurate structured data, genuine third-party coverage, and answer-shaped content on your core topics. Significant incremental spend on LLMO-specific tactics beyond that is premature for most Australian businesses until measurement improves and the playbook stabilises.
Keep exploring
About New Rebellion
New Rebellion is a marketing intelligence consultancy. We build tools, score Australian businesses on how their marketing actually performs, and publish Debrief every day. This dictionary is part of how we work in the open.
How we think →