The framework for being found by machines

One question, five answers: why AEO exists

In Answer Engine Optimization (O'Reilly, 2026), Rodrigo Stockebrand describes a simple experiment: the same set of questions, run across ChatGPT, Perplexity, Google's AI Mode, Claude and Copilot, minutes apart.

Five platforms. One question. And almost zero overlap in who won.

Each system searches differently, trusts differently, cites differently. Being visible to one says nothing about the others. Answer engine optimization — AEO — is the discipline of being found and cited by these systems, and Stockebrand's framework organizes it into three pillars:

1. Technical foundation — can machines find and read your content at all?
2. Content optimization — does your content survive being cut into passages?
3. Brand and entity management — what does the rest of the web say about you?

Why a design generalist follows this field

I wrote earlier about designing this site to be read by machines. The premise was blunt: when someone asks an AI who you are, the answer is assembled from whatever machines managed to read. If they read nothing, you don't exist.

That thought doesn't let you go once you've had it. As a design generalist I've spent my career in the space between art and technology, and AEO is the newest room in that space: making people and products legible to the machines that increasingly mediate discovery. The framework is worth knowing even if you never optimize anything — it explains how those machines decide.

Pillar one: technical foundation

The first pillar answers one question: can answer engines find, crawl and understand your content at all? It covers crawler access, structured data that identifies you as an entity, site architecture, and pages whose content exists in the HTML itself rather than appearing only after JavaScript runs.

It's plumbing, and it's unforgiving. As Stockebrand puts it:

Content that can't be crawled can't be retrieved. Content that can't be retrieved can't be cited. Simple as that.

Pillar two: content optimization

The second pillar is writing that survives retrieval. Answer engines don't read pages the way people do. They cut them into passages — chunks of a few paragraphs — score each one alone, then run every candidate through a pipeline of retrieval, ranking and selection. Stockebrand's image for it is a relay race:

Your content has to successfully clear each handoff to reach the finish line. Get filtered out at any stage, and you're done.

The practical shift: "The goal shifts from ranking to citation. The unit of optimization moves from the page to the passage." Every section has to answer a question on its own, lead with the answer, and carry a trustworthy date. Writing this way is a structural skill — closer to information design than to copywriting.

Pillar three: brand and entity management

The third pillar is everything the rest of the web says about you — and by the numbers, the biggest one. Research cited in Answer Engine Optimization puts roughly 85% of what influences AI visibility in offsite signals: coverage, mentions and consistency across the open web, not anything on your own site.

That's because models carry a second kind of knowledge, baked in during training. Long before any search happens, the model already believes something about you, assembled from every mention of your name it has ever seen.

The web is talking about you whether you like it or not, and the model is listening.

Managing that means consistency: the same story about who you are on your site, your profiles, and everywhere else your name appears.

A system, not a bag of tricks

What made the framework click for me is that it isn't a list of hacks:

Answer engines are systems, and systems respond to the coherent combination of signals more than to any individual optimization.

Foundation, content, reputation — each pillar useless without the others, all of them measured and adjusted over time. Designers have a name for that kind of thinking. We've been doing it for decades: making complex systems legible, coherent and trustworthy for an audience. The audience just changed. It reads faster than we do, and it never stops.