Ask yourself a question that would have sounded absurd five years ago: when an AI assistant answers a homeowner's question about who to call, what exactly is it reading about your company? Not your billboard. Not your truck wrap. Not the logo you paid five figures for. Machines don't see any of it. They read your entity — the structured, verifiable data that tells every AI system who you are, what you do, what others say about you, and whether the facts agree with each other.
Branding is not dying. It is gaining a second body. The first body is the one humans experience — design, voice, feel. The second is the one machines parse. Most companies have spent decades polishing the first and do not know the second exists. I've been saying on stages and podcasts for years that traditional websites are becoming obsolete — this is the mechanism. The audience for your web presence is increasingly not a person scrolling. It is a machine deciding whether to mention you.
The framework: data, soul, systems
I think about the entity era in three layers, and the framework runs through everything this publication will say about branding:
- Data — what machines can verify about you. Structured markup on every page. Consistent facts everywhere your business appears. Machine-readable indexes that hand AI systems your story instead of making them guess it.
- Soul — the voice, values, and story encoded in everything you publish. This used to live in a brand book nobody opened. Now it is effectively training data: every AI that writes about you, or for you, learns your soul from what you've published — or substitutes a generic one.
- Systems — how your brand behaves when software interacts with it. Can an agent get a price, check availability, book a job? The "brand experience" of the machine era includes the experience machines have.
A live example you can inspect right now
I don't ask anyone to take this on theory — this website is the case study. Open jenniferbagley.com/llms.txt. That file is a curated, machine-readable briefing about me as an entity: who I am, what I've built, every page on this site with a description, a question-and-answer corpus, licensing terms for AI systems. Under the hood, every page here carries a structured-data graph — one connected entity: the person, the book, the company, the speaking practice — so that machines don't have to reconcile fragments. It is the difference between hoping AI systems figure you out and handing them the file.
When I say branding is evolving into entity file data, soul, and systems, this is what I mean, running in production, on the site you are reading.
What to do with this if you run a company
- Audit the data layer. Search your company the way a machine would. Do your name, services, hours, and claims agree across every surface? Entity confusion is the new broken link.
- Write the soul down. If your voice and values aren't published, every AI touchpoint defaults to beige. Publish the founder's actual thinking — it becomes the corpus machines learn you from. (It's also, not coincidentally, why this blog exists.) Where does a soul worth encoding come from? From actually understanding your customer's world — the client-first inversion.
- Test the system layer. Try to be your own customer through a machine: can anything short of a phone call get a question answered or a job booked? Every dead end a machine hits is a recommendation you won't get.
The companies that treat their entity as seriously as their logo will be the ones AI systems can confidently recommend — and in a market where recommendations increasingly come from machines, that is not a technical detail. That is the brand. My team at CI Web Group builds this layer for home service companies every day, and the essays that follow in this pillar will break down each layer in depth.