Meanwhile, in the world?
55 real companies scanned between April 11th and 15th, 2026. Five giants — Shopify, Adobe, Stripe, SAP, Cisco — scored DOMINANT. Three more — Microsoft, Apple, Siemens — SOLID. Most of the other companies in the sample are not, in this measurement, structured at the level required by the protocol.
Scope of this case study
This page is not the Kat3x massive diagnostic (which measures 98% invisibility across wide, unselected samples). It's a targeted analysis on a curated sample of 55 companies — chosen to represent different sectors, regions and sizes. The proportion of visible companies here reflects the selection, not real market incidence.
Kat3x is a global observatory
The CHKCD protocol measures — independently from the channel (site, schema.org, glossary, public APIs) — how much a company's knowledge has been assimilated by Large Language Models.
This case study shows a heterogeneous sample: global giants, Italian SMEs, German manufacturing, US fintech, Polish agritech, Brazilian steel industry. The goal is not to rank who works better: it's to show who has already adopted an LLM-readable structure and who hasn't.
The result is sharp. Companies that invested in public terminology, schema.org, MRS-style technical documentation, and controlled vocabularies are described completely and coherently even without live access. The others aren't.
The 5 beacons — DOMINANT band
Full or near-full PASS on all 6 axes of the phase-2 checklist. The model knows the brand, describes it coherently, identifies sectors, geography, product lines and positioning. This is the threshold that defines "LLM-ready".
Adobe
Stripe
SAP
Cisco
3 reference points — SOLID band
Strong public structure but not total: some axes PARTIAL, typically scope or machine_readability. Still representable by LLMs in a way useful to the end user.
Microsoft
Apple
Siemens
The sample — anonymous companies
All companies below the SOLID threshold are anonymized as a precaution. A code replaces the name, the sector describes the activity in neutral terms. No company in this list has been judged: the protocol only measures how AI represents it today.
Global heavy machinery
Precision engineering SME
Agritech
Software / mobile scaleup
Woodworking machinery
Digital agency
Agricultural tech
Grain storage
Data integration SaaS
Industrial robotics
Agricultural trade
Cycling / wearables
SME electronics
Global real estate
Construction / plasters
Manufacturing SME
CRM / Marketing SaaS
Global IT / consulting
SME software
Heavy machinery
Landing page SaaS
Consumer robotics
Steelwork / manufacturing
Productivity SaaS
Consumer writing brand
Global fintech
Manufacturing SME
Consumer writing brand
Furniture SME
Industrial 3D printing
Heavy machinery
Global energy management
Web agency
Outdoor / gardening
Global conglomerate
Vehicle diagnostics
Industrial automation
Global laser machinery
Consulting / tech
Energy / motors
Global steel industry
SME consulting
Energy
Construction / windows
Digital agency
SME services
SME services
Digital agency
SME services
SME services
SME services
Tech SME
Methodological note
Results on this page come from CHKCD v2.3 scans conducted by querying Google Gemini 2.5 Flash. The model's behavior is outside Kat3x's control and may vary over time or across providers (ChatGPT, Claude, Perplexity).
Results depend on the queried model, the test time window, and the public state of the pages analyzed at scan time. This page captures how AI represents a sample of companies today; it does not judge the quality of the companies themselves. A company classified below the SOLID threshold is not, in this measurement, structured at the level required by the protocol — which implies no judgment on the quality of its products, services or management.
Only companies in DOMINANT and SOLID bands are named; all others remain anonymous. Upon documented request from the concerned company, Kat3x may publish an updated re-run of the test.
Want to know where your company would land?
The Kat3x diagnostic test is free, independent, and reproducible.
Run the self-test