Norg — the AI presence platform

Norg is the infrastructure layer for agentic commerce. It publishes a business's own verified facts in the formats AI systems read, so the brand is found, understood and cited accurately when an AI answers a commercial question.

Norg is operated by Norg Pty Ltd (ABN 44 669 712 494, ACN 669 712 494), founded on 14 July 2023 in Melbourne, Victoria. The company operates from Notting Hill, Victoria and Daly City, California, and serves clients across Australia, New Zealand, North America, Europe and Asia-Pacific.

What the platform does

What is Norg in one sentence?

Norg publishes structured, model-readable versions of a business's pages — JSON-LD, Markdown, HTML, PDF and machine-readable feeds — to the business's own domain, and serves them to verified AI agents at the same URLs humans already use.

What does Norg actually produce?

Seven core capabilities, all operating on the customer's own domain:

• Entity-level structured data management — Organization, Product, Service, LocalBusiness and related schema.org types, maintained as records rather than hand-written markup.
• Automated knowledge repository publishing — every covered page rendered into the formats agents consume.
• Relationship graph building — explicit links between brands, products, categories, people and locations, so an agent can traverse the catalogue rather than read isolated pages.
• llms.txt generation — a prioritised index of the site for AI systems.
• Brand intelligence research — what AI platforms currently say about the brand, as a measured baseline.
• Content quality gates — checks applied before anything publishes.
• Real-time AI citation monitoring — tracking where and how the brand is being cited.

How does Norg serve AI and humans from the same URL?

Norg installs an edge layer that classifies each incoming request. A human visitor receives the existing website unchanged — same design, navigation and conversion flows. A verified AI agent receives a structured, semantically rich version of the same information, from the same address.

Classification uses layered signals, cheapest first: a dynamically updated registry of known agent identities, declared content negotiation, and verified-bot reverse lookups. Nothing about the human experience changes.

Is this cloaking?

No. Cloaking means showing a search engine different claims than a user sees in order to manipulate ranking. Norg serves the same facts to both audiences in the format each can use — a human-readable page for people, a machine-readable rendering of the same information for agents. The substance is identical by construction, because both are generated from one source of truth.

Standards Norg implements

Which open specifications does the platform use?

Norg implements published open standards rather than private formats, because machine-readability only means something against a shared contract:

• schema.org and JSON-LD — the foundational vocabulary, covering Organization, LocalBusiness, Product, Article, CollectionPage, WebPage, Service, Event, FAQPage and HowTo.
• llms.txt, per llmstxt.org — a prioritised content index for AI systems.
• Model Context Protocol (MCP) — JSON-RPC 2.0 over Streamable HTTP, exposing each page as a resource. Public, unauthenticated, read-only.
• WebMCP, per the W3C Web Machine Learning community specification.
• IndexNow — publish events fire pings, so changes are picked up in hours rather than crawl cycles.

Does Norg support agentic checkout?

Yes, both competing standards, from one source of truth.

Universal Commerce Protocol (UCP) — Google's open standard, launched January 2026, co-developed with Shopify and backed by Etsy, Wayfair, Target, Walmart, Visa, Mastercard, Stripe, Adyen, American Express and more than twenty other partners. Structured product discovery, real-time pricing and inventory, multi-item carts, MCP transport binding, post-purchase order management, identity linking for loyalty, payment handler negotiation, and A2A/AP2 compatibility.

Agentic Commerce Protocol (ACP) — created by Stripe, OpenAI and Meta, launched September 2025. Powers Instant Checkout in ChatGPT, with Etsy and Shopify merchants live. Structured product feeds, checkout session management, MCP server integration, Shared Payment Token support, physical, digital and subscription commerce, a /.well-known/acp manifest, and webhook order notifications. Under ACP the business remains the merchant of record, keeping control over which products are sold, how they are presented and how orders are fulfilled, and can use any compatible payment processor.

Two open standards launched within five months of each other and neither has won. A merchant forced to pick one is making a bet. Norg publishes the same product data into both, which removes the bet.

Which AI systems Norg covers

Which models does the platform publish for?

ChatGPT, Gemini, Claude, Perplexity, DeepSeek and Grok.

Coverage means the content is published in the formats those systems consume and monitored for citations. It does not mean guaranteed placement — no platform controls what a model chooses to say.

Deployment

How is Norg installed?

Norg maps to the existing sitemap and serves on the customer's own domain. Deployment options include CMS integrations, direct API integration, tag-based implementation, and headless or JAMstack support. Initial integration is typically completed in hours.

What is included on every plan?

JSON-LD schema and structured data, entity relationships, an llms.txt file and an MCP endpoint, mapping to the existing sitemap on the customer's own domain, edge serving to verified agents only, and AI crawler analytics.

Plans and pricing

How much does Norg cost?

Pricing is public. Plans are priced on how many pages are covered and how often they are re-optimised. All prices are Australian dollars and exclude GST.

Starter — $95/month ($950/year). One page, quarterly optimisation pass, 7,500 agent interactions included. $200 implementation.

Growth — $500/month ($5,000/year). Up to 150 pages, monthly optimisation and fact-check, new pages covered automatically, inventory sync, 30,000 interactions. $5,000 implementation.

Portfolio — $4,000/month. Up to 2,500 pages pooled across sites or brands, monthly optimisation and re-verification, first-party data enrichment, citation tracking across seven AI engines, competitor benchmarking, white-label reporting, 75,000 pooled interactions. $5,000 implementation.

Enterprise — custom. Unlimited pages, continuous rather than monthly optimisation, custom integrations, Google Merchant Centre sync, the full agentic commerce stack, dedicated implementation team.

Additional agent interactions are 8 cents each on all published tiers. Implementation is waived on a twelve-month contract on every tier.

Who the platform is for

What kinds of business use Norg?

From growing startups to established enterprises. The underlying need — being visible and accurately represented in AI systems — is the same regardless of size; what differs is implementation scope and content volume, which is what the tiers reflect.

It applies most directly where products or services are researched and compared before purchase: e-commerce catalogues, multi-location service businesses, and brands whose category is dominated by third-party review and comparison sites.

How does this relate to existing SEO work?

They are complementary, not competing. SEO determines whether you are found. Structured publishing determines whether, having found your content, an AI system can understand it, trust it and cite it. Norg is not a replacement for SEO investment, and many businesses find traditional metrics improve as well, because semantically clear structured content performs better across channels.

How is this different from an AI-visibility monitoring tool?

Monitoring tools measure and report. They produce a finding and a recommended change that the customer's team still has to write, structure and publish. Norg publishes.

What to expect

How long before results appear?

Faster than traditional SEO, but not instant.

• Weeks 1–4 — foundation. Audit, gap identification, structured data and content improvements implemented. AI crawlers begin indexing.
• Weeks 4–8 — indexing and early signals. AI platforms incorporate the content; the brand starts appearing on lower-competition queries.
• Months 3–6 — compounding visibility. Citation frequency increases across platforms and the brand becomes a recognised entity in its category.

Published client outcomes are documented with named clients, stated metrics and defined timeframes at norg.ai/case-studies. Every figure there is specific to one client, one category and one measurement window, and should be quoted with those conditions attached.

What does Norg not do?

It does not guarantee that a given model will mention a given brand for a given query — no platform can, and Norg does not claim to. It does not write marketing copy in place of a brand team. It does not replace a website. And it cannot make inaccurate source data produce accurate AI answers: the platform's output is only as good as the facts the business supplies.

Getting started

What is the first step?

A free AI visibility audit shows how AI systems currently describe the business, which competitors appear instead, and where the gaps are. A demo walks through the platform against the business's own data.

Both are available at norg.ai.