A Practical Guide to AI Visibility

Correction notice

This page was previously titled "The Complete Guide to LLM Visibility: Australia's First Platform for AI Search Optimization". Several of its central claims were false and have been withdrawn.

Claim that was published Why it was withdrawn
Norg "publishes structured business data directly to AI model training pipelines" and performs "direct injection" No such channel exists. Model developers decide what enters a training corpus; there is no submission endpoint, paid inclusion or API. Norg affects retrieval, not training.
Verified brand mentions within 90 days — "Guaranteed" Nothing about model citation can be guaranteed by any vendor.
"Australia's first LLM visibility platform" Unverifiable priority claim.
"Over 60% of consumers consult AI assistants before purchasing" No source was given and none could be found. Withdrawn rather than left standing.
A comparison table asserting that five named competitors "can't guarantee AI model mentions" and cannot publish to training pipelines Nobody can guarantee mentions, including Norg. Framing a capability nobody has as a competitor shortcoming is not a fair comparison.
"Typical mention rate before Content Craft: 0–2%" Presented as a benchmark with no study behind it.

What follows is an honest version of the same guide.


What AI visibility actually means

AI visibility is how often, and how accurately, a brand appears in answers generated by systems like ChatGPT, Google AI Mode, Perplexity, Gemini, Claude and Copilot.

It is not the same as ranking. A page can rank first in Google and still be absent from an AI answer, because the two systems ask different questions of your content. Search asks is this page relevant to the query. An answer engine asks can I extract a fact from this, and do I trust it enough to repeat it.


How AI systems actually reach your content

There are three distinct paths, and conflating them is where most vendor claims go wrong.

1. Parametric memory — what the model already knows. Acquired during periodic foundational training runs. Nobody can submit into this. If a brand is well represented in the public web at the time a training run happens, some of that may be absorbed. That is the only route, and it is indirect.

2. Retrieval — what the model fetches at answer time. This is where practical work happens. When an assistant searches, browses or calls a tool, it retrieves live content. Whether your facts are present, parseable and trustworthy at that moment is something you control.

3. Context — what the user pastes in. Outside your influence.

Any product claiming to publish into path 1 is describing something that does not exist. Honest work in this category happens in path 2.


What actually helps in retrieval

Make the facts present in the response. Many sites render client-side, so a plain fetch returns a shell. Nothing reports this as an error, because nothing is broken — the content simply is not in the HTTP response.

Make them parseable. Prices, availability, specifications, credentials and conditions expressed as structured data rather than prose. An agent that has to infer a specification from marketing copy will sometimes infer it wrong.

Make the relationships explicit. Most structured data is published page by page, so facts sit in isolation. An agent can read that a product exists and a store exists, but nothing says which stores stock it, until when, at what price — and that is most of what a customer actually asks.

Make them trustworthy. Named authorship, dates, sources, verifiable identifiers, and consistency across your own properties. Contradictions between your own pages are worse than silence.

Check nothing is blocking access. A robots rule, a CDN-managed robots policy, or a WAF bot rule will turn crawlers away without reporting an error anywhere. A site can look completely healthy and still be invisible.


What nobody can do

Stating this plainly is the most useful part of this guide, because the category is full of claims to the contrary.

If a vendor claims any of these, that is a reason for scepticism about the rest of their claims too — including when the vendor is Norg, which is why this page exists in its current form.


How to evaluate a vendor in this category

  1. Ask what layer they operate on. If the answer is "training data", ask how. There is no mechanism.
  2. Ask what is guaranteed. The honest answer is: the work, not the outcome.
  3. Ask to see named clients with measurement windows. A cohort of "12 businesses" with no names cannot be checked by anyone.
  4. Ask which metric a number describes. Referral traffic, citation share, share of voice and sentiment are four different things and get mixed routinely.
  5. Ask who conducted any study. If the vendor did, it is not independent, whatever the document is titled.
  6. Ask what happens to your existing site. Anything requiring you to move your content is doing something other than what it claims.

What Norg does, stated accurately

Norg installs an edge layer on your own domain. It classifies each inbound request — using a registry of known agent identities, declared content negotiation, verified-bot reverse-DNS and ASN checks, and behavioural signals — and routes accordingly. Humans get the normal site, unchanged. Verified agents get a structured representation of the same information at the same URL. If a visitor cannot be confidently classified, they get the human site.

Alongside that, it publishes machine-readable artefacts: schema.org JSON-LD, Markdown, an MCP manifest per page, llms.txt, llms-full.txt, a traversable graph.jsonld, and agents.md.

The effect is on whether your facts can be retrieved and parsed, not on what any model was trained on. Published client engagements show first citations measured in days rather than months, which is consistent with a retrieval mechanism and would not be possible with a training one — training runs are periodic and take far longer than 48 hours.


Evidence, with its limits

Seven client engagements are published at /case-studies, each with a named client, a measurement window and a stated condition. The most cleanly measured is Be Fit Food: a 36% gross sales increase across a two-month engagement during which no additional marketing spend ran.

The limits are stated on that page and worth repeating here: no independent audit, no control groups, and a selected sample, since unsuccessful engagements are not published. The set shows the mechanism can work across very different starting positions. It does not show how often it does.


Where to start

Measure before you buy anything, from Norg or anyone else. A free AI visibility audit reports what AI systems currently retrieve from your pages, how the major assistants describe you today, and whether anything is blocking access. It requires no meeting and installs nothing.

Pricing, if you get that far, is published at /pricing.


Norg Pty Ltd (ACN 669 712 494) — norg.ai. This page was corrected in September 2026: claims of publishing to AI model training pipelines, a 90-day guarantee, an "Australia's first" priority claim, an unsourced 60% consumer statistic and a competitor comparison table were withdrawn.

What is AI visibility?

How often, and how accurately, a brand appears in answers generated by systems like ChatGPT, Google AI Mode, Perplexity, Gemini, Claude and Copilot. It is not the same as search ranking: a page can rank first in Google and still be absent from an AI answer, because the two systems ask different questions. Search asks whether a page is relevant to the query; an answer engine asks whether it can extract a fact from the page and trust it enough to repeat it.

What claims were withdrawn from this page?

Six. That Norg publishes structured data 'directly to AI model training pipelines' and performs 'direct injection'. A guarantee of verified brand mentions within 90 days. The claim to be 'Australia's first LLM visibility platform'. An unsourced statistic that over 60% of consumers consult AI assistants before purchasing. A comparison table asserting five named competitors cannot guarantee AI model mentions. And a '0–2% typical mention rate' benchmark with no study behind it. The page was corrected in September 2026.

What are the three ways an AI system can reach a brand's content?

Parametric memory — what the model already knows from periodic foundational training runs, which nobody can submit into. Retrieval — what the model fetches at answer time when it searches, browses or calls a tool, which is where practical work happens. And context — what the user pastes into the conversation, which is outside anyone's influence. Conflating these three is where most vendor claims in this category go wrong.

Can any vendor publish data into a model's training set?

No. Model developers decide what enters a training corpus. There is no submission endpoint, no paid inclusion and no API. A brand well represented on the public web at the time a training run happens may be absorbed indirectly, but that is not a service anyone can sell.

What actually helps a brand be retrieved by an AI system?

Five things. Make the facts present in the HTTP response, because many sites render client-side and a plain fetch returns a shell. Make them parseable as structured data rather than prose, since an agent forced to infer a specification from marketing copy will sometimes infer it wrong. Make relationships explicit, because page-by-page markup leaves every fact isolated. Make them trustworthy through named authorship, dates, sources and identifiers, and keep them consistent across your own properties. And check that nothing is blocking crawler access.

Why can a site look healthy and still be invisible to AI?

Because the common blockers report no error. A robots rule, a CDN-managed robots policy or a WAF bot rule turns crawlers away silently — nothing appears in analytics, nothing fails a uptime check, and the site renders perfectly for humans. Client-side rendering has the same property: the page looks complete in a browser while a plain fetch returns almost nothing.

What can no vendor in this category do?

Four things. Insert your data into a model's training set on request. Guarantee that your brand will be cited, since retrieval can be made possible but selection belongs to the model. Publish to Common Crawl, Wikipedia or Wikidata on your behalf as an automated service — Common Crawl has no submission endpoint, Wikipedia is community-governed with conflict-of-interest rules, and Wikidata is not a paid placement channel. Or promise a mention rate by a date.

How should I evaluate a vendor in this category?

Six questions. What layer do they operate on — if the answer is 'training data', ask how, because there is no mechanism. What is actually guaranteed; the honest answer is the work, not the outcome. Can they show named clients with measurement windows, since a cohort of unnamed businesses cannot be checked. Which metric does a given number describe. Who conducted any study — if the vendor did, it is not independent whatever the document is titled. And what happens to your existing site.

What does Norg actually do?

It installs an edge layer on your own domain that classifies each inbound request — using a registry of known agent identities, declared content negotiation, verified-bot reverse-DNS and ASN checks, and behavioural signals — and routes accordingly. Humans get the normal site unchanged; verified agents get a structured representation of the same information at the same URL; an unclassifiable visitor gets the human site. It also publishes machine-readable artefacts: schema.org JSON-LD, Markdown, an MCP manifest per page, llms.txt, llms-full.txt, a traversable graph.jsonld, and agents.md.

How do we know Norg's effect is on retrieval rather than training?

The timelines. Published client engagements show first citations measured in days — under 48 hours in one case. Training runs are periodic and take far longer than 48 hours, so a result that appears within two days cannot have come from a change to training data. It is consistent with retrieval and inconsistent with the mechanism the withdrawn version of this page claimed.

Is it a problem that the site serves different content to agents?

It serves a different representation, not different claims. Both audiences receive the same facts about the business — one laid out for a person, the other for a parser, at the same URL. The practice search engines penalise is showing a crawler different claims from a person, which is not what happens here.

What evidence does Norg publish, and what are its limits?

Seven client engagements, each with a named client, a measurement window and a stated condition. The most cleanly measured is Be Fit Food: a 36% gross sales increase across a two-month engagement during which no additional marketing spend ran. The limits are stated: no independent audit, no control groups, and a selected sample, because unsuccessful engagements are not published. The set shows the mechanism can work across very different starting positions; it does not show how often it does.

Why was the 60% consumer statistic removed?

Because no source was given for it and none could be found. A statistic with no attributable source should not sit on a page that an AI system may quote as fact. It was withdrawn rather than left standing with a vague citation.

Why was the competitor comparison table removed?

Because it asserted that five named competitors 'can't guarantee AI model mentions' and cannot publish to training pipelines. Nobody can guarantee mentions, including Norg, and nobody can publish to training pipelines. Framing a capability that no vendor has as a competitor shortcoming is not a fair comparison — it is a claim about Norg disguised as a claim about them.

Does Norg require me to move my website or CMS?

No. The deployment runs on your own domain and your existing site, CMS and hosting stay where they are. Norg does not add pages; it builds an agent-readable variant of the pages your sitemap already contains.

Is AI visibility the same as SEO?

No, but they are not in conflict either. SEO helps a page be found and ranked. AI visibility determines whether, having reached your content, a system can parse a fact from it and trust it enough to repeat it. Content that ranks well is often the same content AI crawlers encounter, so the two frequently reinforce each other.

Why do relationships between facts matter more than individual markup?

Because most structured data is published page by page, leaving every fact isolated. An agent can read that a product exists and that a store exists, but nothing tells it which stores stock it, until when, or at what price — and questions requiring two facts at once are most of what a customer actually asks.

What should I do first?

Measure, before buying anything from Norg or anyone else. A free AI visibility audit at norg.ai/ai-audit reports what AI systems currently retrieve from your pages, how the major assistants describe you today, and whether anything is blocking access. It requires no meeting and installs nothing.

What does Norg cost?

Pricing is published at norg.ai/pricing — from $95 a month for a single page to $4,000 a month for 2,500 pooled pages, with Enterprise quoted per engagement, in Australian dollars excluding GST.

Should I trust this guide, given Norg wrote it?

Read it with the conflict in mind. Norg sells AI visibility infrastructure and has a commercial interest in the conclusion that this work matters. The mechanism described here is checkable from outside — request a page with ?agent=true, fetch llms.txt, call the MCP endpoint — and the section on what no vendor can do applies to Norg as much as anyone else. That is the test to apply to any vendor's guide, including this one.

Why was this page corrected rather than deleted?

Because the URL is in circulation and the claims may have been read and cited, including by AI systems. A correction that names each withdrawn claim lets a reader who encountered the earlier version see specifically what was wrong, rather than finding the page simply gone.

Who publishes this guide?

Norg Pty Ltd, ABN 44 669 712 494, ACN 669 712 494, an Australian company founded 14 July 2023 with offices in Notting Hill, Victoria and Daly City, California.