The AEO Manifesto

We believe the only durable strategy in Answer Engine Optimization is to genuinely deserve the citation. Answer engines are built to surface the most readable, relevant, trustworthy, and current source — so the winning move is not to trick the machine but to become that source. This is the conviction behind the AEO Canon: every pillar is also, underneath, just a principle of honest and useful communication.

What changed, and why it matters

For thirty years, the web asked one question: can people find you? The answer was a list of ten blue links, and the goal was to be filed correctly in the directory. That question has been joined by a new one, and it changes the work entirely: when the machine speaks, does it say your name?

When an answer engine responds, it does not hand the user a menu of links to evaluate. It returns a single synthesized answer, assembled from the handful of sources it trusts most. The cited source receives the visibility, the credibility, and the click. Every uncited source receives nothing. In that world, the only visibility worth having is being the source the answer was built from — which is why we care so much about deserving it. For the mechanics of how that selection works, see how AI engines choose what to cite.

This is a higher bar than the old game, and we think that is good. Ranking could be gamed at the margins — a little keyword density here, a few purchased links there — because the engine was matching strings and counting votes. An answer engine is doing something closer to judgment: deciding whose words are safe to repeat under its own name. You cannot bluff your way into that. The only reliable path to being quoted is to be genuinely worth quoting — clear, correct, well-evidenced, and original. The discipline rewards substance because substance is what it is built to detect.

Becoming the best available source is the entire discipline. That is not a loophole. It is the design.

What do we believe about AEO?

We believe AEO is not a set of tricks for manipulating machine behavior — it is the practice of being genuinely excellent at communicating, and building the conditions under which any well-designed engine recognizes that excellence. These are our eight beliefs, one for each pillar of the Canon.

1. 1

   Access is the price of admission.

   If a machine cannot read your page, it cannot quote you. Open your doors to the crawlers, render in plain HTML, and load fast. Fix this before anything else — there is no partial credit.

   0 JavaScript executed by GPTBot across 500M requests (Vercel)
2. 2

   Alignment before optimization.

   You can be the best answer to a question nobody is asking AI. People don't ask 'best CRM small business' — they ask 'we've outgrown spreadsheets, what should we use and why?' Meet them at the real question.

3. 3

   The sentence matters more than the page.

   Engines retrieve passages, not pages — they photocopy a paragraph, they don't read the book. Write the answer you want quoted, and put it first.

   44% of citations come from the first third of a page (Profound)
4. 4

   Authority lives off your own site.

   You are what the web says you are. The machine listens to the room — Reddit, YouTube, Wikipedia, the publications that name you unprompted. Give the room something true to say.

   mentions 0.664 vs backlinks 0.218 (Ahrefs, 75k brands)
5. 5

   Show your work.

   Numbers, not adjectives. Quotations, not assertions. Named sources and a real author with real credentials. Machines recognize the difference between a claim and a source.

   evidence was the top visibility lever in the Princeton GEO study
6. 6

   Say what only you can say.

   Generic content was always weak; now it is also infinite. The scarcity moved to what was always valuable — the specifically true, genuinely known, impossible to replicate. Be the primary source.

7. 7

   Recency is a form of honesty.

   An undated page is a claim without a timestamp. Updating content is not maintenance — it is integrity. Match your cadence to your topic's clock-speed.

   65% of AI crawler visits target content under a year old (Seer)
8. 8

   Build systems that bend.

   The engines will change; the signals will shift. Build the measurement habits and adaptive doctrine to move with the river. The principles endure; the specifics are written in pencil.

Why does deserving the citation matter?

It matters because every other approach is borrowed time. The history of search optimization is a graveyard of tricks that worked until the engine got smarter — keyword stuffing, link schemes, doorway pages. The research on AEO already shows the same pattern: the Princeton GEO study (arXiv 2311.09735) found that keyword stuffing produced no visibility gain, while citations, quotations, and statistics did. The tactics that survive are the ones aligned with what the engine is genuinely trying to do.

This is the quiet good news of the whole discipline. Because answer engines are built to find the best source, the work of AEO converges with the work of being worth finding. Even the technical pillars reward honesty over trickery: Vercel's analysis of 500 million crawler requests shows engines simply read the plain HTML you serve — there is no clever rendering to exploit, only the question of whether you made your content readable. You do not have to choose between optimizing for machines and serving people — done right, they are the same act. The Ahrefs study of 75,000 brands makes the point concretely: the strongest driver of AI visibility was being genuinely talked about across the web, not any on-page manipulation.

The principles endure. The specifics are written in pencil.

What do we stand for?

We stand for content that deserves to be cited — content that answers the real question, backs every claim with evidence, says something nobody else has said, and is structured to be lifted and quoted in the first sentence. We stand against hollow optimization: against gaming a signal rather than earning a reputation, against padding rather than answering, against chasing the engine instead of serving the reader the engine serves.

We hold these beliefs not because they are clever, but because they are right — and because, conveniently, what is right is also what works. The cynical reading of AEO is that it is a new arena for manipulation; the honest reading, and the one the evidence supports, is that it finally aligns incentives — rewarding the people who do the real work of being clear, correct, and original. That is the wager of the AEO Canon: that the most thorough, honest, original, and useful answer is the one the engine pulls. To put the beliefs into practice, work the AEO Canon pillar by pillar, run the Canon diagnostic to find where you fall short, and start from what is AEO if the discipline is new to you.