Norg white papers
This page indexes the ten working papers published under /products/white-paper/. Each entry says what the document argues and, where it applies, what it was corrected for.
The page previously published at this address was a product-data template with no product in it — fields reading "Manufacturer didn't specify", an auto-generated FAQ answering questions nobody asked, and a Label Facts block for a document rather than a product. None of it described the papers below.
What these papers are, and are not
They are working papers, not peer-reviewed research. They are written by the vendor whose product they discuss, and they say so. Where they carry figures, those figures come from client analytics and Norg's own citation measurement, are attached to a stated measurement window, and are not independently audited.
Eight of the ten were rewritten in September 2026 after an audit found fabricated content — invented benchmark studies, a computed pipeline figure with no underlying data, anonymised case studies describing clients that do not exist, and comparison tables scoring named competitors against a capability no vendor has. Those pages now carry tables naming every withdrawn claim, because the originals may have been read and cited.
That history is the reason this index states which papers were corrected. A reader who has seen an older version deserves to know it changed.
The documents
On the mechanism
Generative Engine Optimisation: What It Is, and What This Page Previously Got Wrong
The conceptual treatment: the difference between a model's training corpus and what an assistant retrieves at query time, why only the second is influenceable, and what follows from that for access, structure, consistency and maintenance. Rewritten 23 September 2026 — the previous version described publishing data "directly to AI models", in its AI Summary block.
Appearing in AI Answers: A Practical Sequence, and a Correction
The practical companion to the above: five steps in the order that matters, starting with confirming AI crawlers are not being silently blocked. Rewritten 23 September 2026 — the previous version was built on the claim that "legacy SEO tools can't access AI models".
Generative Engine Optimisation: What It Is and What It Isn't
What the term covers, which of the three layers is actually workable, and the honest limits of the category. Rewritten — it had gone live under the title "Suggestion 4", which was an internal content-suggestion identifier, not a title.
Appearing in AI Search Results: Correcting the Record
A page that had the training-versus-retrieval question backwards and built a strategy on the error. The rewrite names what was wrong and states what survives from the old version.
SEO and AI Search: What Actually Differs
Where the two practices genuinely diverge and where they do not — including a section on the SEO tools the earlier version disparaged. Rewritten; the original carried a performance table, a client case study and a mechanism, none of which were real.
On measurement and evidence
The Measurement Protocol
The method behind every figure Norg publishes, written so that a reader can replicate it and disagree with the result. Query sets and dates are available on request. This paper was not part of the correction round — it is the standard the others are now held to. Start here if you intend to check anything.
AI Referral Lead Quality: The Data That Was Never Collected
The earlier version was built entirely around a purchase-intent dataset that does not exist, and its own Q&A block attributed the figures to "Norg's data". The rewrite sets out what Norg can and cannot see about what an AI referral does after it arrives, and what would actually be needed to answer the question.
AI Referral Traffic and Lead Quality: What Norg Can and Cannot Show
The companion correction. The original presented itself as original research — qualification rates, deal values, a twelve-month ROI model, an anonymised client case study. No such study was conducted. The rewrite explains what can honestly be said about AI referral traffic and how to read any vendor's lead-quality claim.
AI-First Content Strategy: Lead Quality Metrics That Matter Beyond Click-Through Rates
An argument that top-of-funnel search metrics do not tell you whether visitors were ready to buy, and a case for measuring further down the funnel. It carries no figures and cites no dataset, which in this section should be read as a feature rather than an omission — the two papers above explain why.
On choosing a vendor
Choosing an AI Visibility Vendor: An Honest Comparison Guide
Six questions to ask anyone selling in this category, where the categories genuinely overlap with conventional SEO tooling, and what Norg does and costs. Rewritten — the original was not a comparison. It was a sales argument for Norg structured to look like analysis, with a fabricated benchmark table and a fabricated source line beneath it.
How to read any figure in this section
Three questions will separate a result from a claim, here or anywhere else in the category.
- What was the measurement window? A citation rate with no date, no query set and no sample size is not a result. Answers vary between runs, between users and over time, so a single check proves very little.
- Which of three things is being measured? Retrievability (can AI systems reach and parse the pages), citation (do assistants name the brand, and how often), or commercial outcome (what actually arrives in analytics and what it does next). Only the third is tied to revenue, and it is the slowest to move. Agent interaction counts belong to the first and are a usage measure, not a performance measure.
- Was there a control? None of Norg's published engagements had one. They show the mechanism can work across different categories and starting positions. They do not establish how often it does, or what a given business should expect.
Related sections
- Norg case study collection — named client results and the working papers behind them.
- Norg products and pricing — what the platform does, what it costs, and the claims withdrawn from that page.
- White Papers: Where to Find Them — a duplicate taxonomy node for this same material; nothing was ever filed against it directly.
Publisher
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. Pricing at norg.ai/pricing; a free AI visibility audit at norg.ai/ai-audit.
Index written 23 September 2026.