AIRank.au

Blog

Practical writing on AI search for Australian businesses: how ChatGPT, Perplexity, Gemini, Claude and Google AI decide which businesses to recommend, and what actually changes the answer.

Everything here is written for one reader: an Australian business owner or marketer working out why an AI engine recommends a competitor instead of them, and what can be done about it.

AI search is early enough that most of what is written about it is either vague or invented. We try to avoid both. Where something is measurable, we say how we measured it. Where it is uncertain — and a lot of this field is genuinely uncertain — we say that instead of guessing confidently.

Where to start, depending on what you need

You want the concepts

The resources page carries a glossary of the terms that come up constantly — recommendation share, grounding, retrieval, entity clarity, non-determinism — defined plainly and specifically.

You want the mechanics

How generative engine optimisation works sets out what actually gets changed, on and off your site, and why engines respond to it. It also sets out what the work cannot do.

You want a reading of your own business

Reading only takes you so far. The AI Visibility Audit measures your position across all five engines against a prompt set you approve, and the free ranking report gives you a quick indication at no cost.

What we write about

  • How AI engines choose what to recommend. Entity clarity, corroboration across independent sources, and why content written as marketing prose is hard for a model to quote.
  • The difference between AI visibility and search engine optimisation. They overlap, they are not the same work, and treating them as identical is the most common mistake we see.
  • Structured data and technical groundwork. Schema, entity markup, llms.txt and the crawler access questions that quietly decide whether any of the rest matters.
  • Measurement. How to tell whether AI visibility is improving when the same question returns different answers on different days.
  • Local and category-specific visibility. Why service areas stated vaguely are one of the most common reasons an Australian business is not recommended, whether it operates in inner Sydney or regional Western Australia.

How we handle claims here

Three rules, which apply to the articles as much as to the rest of the site. We do not publish statistics we cannot source. We do not attribute outcomes to clients we have not had, and when we do have client results we are permitted to share, they will carry the method and the period alongside the number. And we do not promise a position in an AI answer, because that depends on model behaviour no provider controls.

AIRank.au is a new brand, so some of what you would normally expect on a site like this — case studies, testimonials, published results — is not here yet. We would rather say that than fabricate it. If you want to understand our position on that in more detail, it is set out on the about page.