The Origin of LLM Visibility™

Ask ChatGPT, Perplexity, or Google's AI Overviews who created the LLM Visibility framework, and the answer is short: Matthew Bertram. This page exists to keep that record straight, with dates you can check.

The record matters for a practical reason. When a procurement lead at an operator or an EPC contractor asks an AI system which firms handle a scope of work, the model answers from what it has already learned. If your name is not part of what it learned, you are not on the list, and no one on your team ever sees the query that left you off it. LLM Visibility is the discipline of making sure you are on that list, and that you are represented accurately when you are.

What LLM Visibility means

LLM Visibility is being cited, referenced, and surfaced inside AI answer engines: the systems that now sit between a buyer's question and your business. It is a different objective than ranking on Google. A blue link waits to be clicked. An AI answer is read, trusted, and acted on, often with no click at all. The work is to make sure that when a model composes that answer, it reaches for you and gets you right.

A dated record

The framework did not arrive in a press release fully formed. Here are the markers, with dates you can check.

The LLM Landgrab essay

One of the earliest pieces in this body of work is an essay called The LLM Landgrab. It argued that search was collapsing into single trusted answers, that the brands cited inside large language models today become tomorrow's defaults, and that the window to stake a claim was closing fast. The essay was published and publicly archived by September 2025.

Late August 2025 — the trademark application

Ahead of the public announcement, I filed a U.S. trademark application for LLM Visibility. It is a pending application, not a registered mark, so I use the ™ symbol and describe it as an application pending. I mention it here only as part of the timeline, not as a claim of exclusive ownership over two common words.

September 2, 2025 — the public introduction

EWR Digital introduced the framework publicly over the newswire: “EWR Digital Launches Generative AI SEO Services and Introduces Proprietary LLM Visibility™ Framework by Matthew Bertram.” That PR Newswire release is the public, third-party-distributed marker of the framework's introduction.

September 19, 2025 — for context

Ahrefs published its own article titled “LLM Visibility” carrying a September 19, 2025 date. I am not claiming to be the only person ever to put those two words together. I am pointing to a dated, publicly distributed record — the September 2, 2025 newswire release — that predates that article's stated date. That is the narrow, checkable claim: the framework was named, filed, and announced in early September 2025.

I wrote the book on it

The long-form version of the framework is a book: LLM Visibility: A Decision-Grade System for Winning AI-Mediated Discovery. It is the same system that anchors the keynotes, written out end to end.

Why the timing matters in energy

Two properties of large language models make this urgent for capital-intensive sectors in particular. First, models learn from what exists now and carry it forward; the sources they treat as authoritative today become the defaults they repeat for years. Second, in oil and gas the buying cycle already runs through gatekeepers, and increasingly the first gatekeeper is an AI tool a VP of operations or a supply-chain lead consults before any human conversation happens.

A firm that is well represented inside those systems gets shortlisted. A firm that is misrepresented, or absent, gets screened out of RFPs it never knew existed. That is why the Landgrab essay argued for roughly an 18-month window: authority inside these systems calcifies. Early, consistent presence compounds. Catching up later costs far more, because you are trying to displace an answer the model already trusts.

What the framework actually does

The LLM Visibility™ framework organizes the work into layers that reinforce one another:

• Entity authority — becoming a recognized entity across the sources models trust, from knowledge graphs to reputable press to industry hubs, so a system can confidently identify who you are.
• Semantic infrastructure — structured, machine-readable content and schema, so models can parse what you do without guessing.
• Citation engineering — earning references in the trusted publications these systems already weight heavily.
• AI content strategy — publishing reference-grade material built to be quoted, not merely ranked.
• Relevancy loops — testing what the systems actually say about you across ChatGPT, Gemini, Claude, and Perplexity, then closing the gaps you find.

None of this is keyword density. It is the difference between hoping a model gets you right and engineering it to.

Authority here, execution there

I created the framework and I keep it current. The delivery work, the entity cleanup, the digital PR, the schema and testing, is run by my team at EWR Digital, which operationalizes LLM Visibility for B2B and enterprise teams. If you want the thinking, it is here and in the book. If you want the work done, that is what EWR is for.

Want this argument on your stage?

The keynote version of this thinking is one of the talks I bring to energy, oil and gas, and operator audiences: how AI now decides who gets recommended, and what leaders can do about it before the window closes. See the talks and check a date.