Two businesses in the same category. Same size. Same age. Same product quality, more or less. Roughly the same marketing budget. Both have done the AEO tactical work - schema markup, structured data, AI-friendly content, the standard playbook.
One of them appears in roughly twenty percent of category-level AI prompts when prospects search for businesses like theirs. The other appears in zero percent. Across hundreds of prompts, on five different AI engines, one is a steady part of the conversation and the other is invisible.
Why?
If you cannot answer that question with confidence, you cannot do AEO. You can do tactics. You can produce content. You can implement schema. You can wire up llms.txt files and submit to AI directories. You can do every single thing the current AEO playbook recommends, and if the underlying question stays unanswered, your work will produce visibility for the businesses that already had the answer and zero for those that did not.
The question sounds simple and is not. What does the corpus believe about you, and why does it believe that? The answer requires a framework. The framework is the rest of this series, and this piece is the map of it.
The Diagnostic Question
Strip every tactic away and there is one question left. Every meaningful AEO outcome flows from it.
When the corpus is asked about your category, what does it say about you - and is what it says even close to true?
Five sub-questions hide inside that one. Does the AI know your business exists at all? When it describes you, are the details accurate? When it places you in a category, is the placement specific or generic? When it ranks you against competitors, does it rank you accurately or default to whoever has been mentioned more in trade press? When it recommends businesses like yours, are you in the recommendation set or outside it?
Every AEO question that matters reduces to one of those five. Every AEO tactic that does not address one of them is decorative. This is uncomfortable to say because it invalidates a lot of work that a lot of operators are doing right now in good faith. It is true anyway.
The framework is built around answering those five questions. It has three pillars and one parallel layer. The pillars describe the brand qualities that determine corpus-level beliefs about your business. The layer describes how those beliefs reach users through different engine architectures. We will introduce all of them here, then go deep on each in the pieces that follow.
Three Pillars
The first pillar is Identity. Does the corpus know you exist as a distinct entity? Identity is the prerequisite to everything else. Without it, no amount of work on the other pillars produces visibility. Below the threshold of recognition, you might as well not exist as far as AI engines are concerned. Above the threshold, Identity becomes a gradient, and elevation determines how often and how confidently the AI surfaces you in answers.
Identity is built off-site, through accumulated mentions of your business on third-party authoritative sources over time. Wikipedia. Trade press. Podcasts. Industry directories. Reddit and forum discussions. None of this lives on your website. None of it is fixable through schema markup. Identity is the pillar most AEO advice is not addressing because the work to build it does not look like AEO work - it looks like PR, brand-building, and category presence accumulated over years. We will go deep on what Identity is and how to build it.
The second pillar is Authority. Does the corpus have a settled position about your rank in your category? Authority is different from Identity in a precise way. Identity is about whether you exist. Authority is about whether the corpus has absorbed a comparative claim about you - that you are top three in your space, that you handle the largest volume in your specialty, that you are recognized as the leader by industry rankings repeated annually.
Authority is not reputation. It is not customer reviews. It is not general respect from peers. It is something narrower and more specific: trade-press repetition of comparative ranking claims with quantitative signals attached, accumulated over enough years that the corpus has settled on a position. A firm that has been described as "the leader in its category" by dozens of trade publications across a decade will have Authority. A firm that is well-respected but not described in those terms will not, even if its product is equally strong. We will go deep on Authority - what it is, why most operators misunderstand it, and how to build it deliberately.
The third pillar is Specificity. Do you own vocabulary in your category that the AI reaches for when prompted? Specificity is the most counterintuitive of the three pillars and the one most often confused with content quality. It is not about publishing more detailed content. It is about owning named, distinctive terminology that exists at the intersection of your brand and your category, such that when the AI is prompted about your category, your terminology is part of the path of least resistance for the model to follow.
A national fitness chain that has branded its training methodology with proprietary scoring vocabulary and proprietary technology terms gets cited heavily when AI is asked about heart-rate-based interval training, because the model has absorbed that vocabulary and reaches for the brand alongside it. A competitor that uses generic category language - the same terminology everyone else in the space uses - does not get this lift, regardless of how much content they publish. Specificity is what separates brands that became inseparable from a slice of category vocabulary from brands that just published a lot of words. We will go deep on Specificity - what counts, what does not, and how to build it deliberately over the years it actually takes.
The Hierarchy
Three pillars do not produce visibility additively. They produce visibility hierarchically, and getting the hierarchy wrong is one of the most common mistakes operators make when they first encounter the framework.
Identity is upstream. Authority and Specificity are parallel and downstream of Identity.
What this means in practice: if Identity is below the threshold, work on Authority and Specificity produces no measurable lift. The AI cannot reach for your branded vocabulary if the AI does not know you exist. The AI cannot rank you in a category if the AI does not have a representation of you to rank. We have audited businesses with strong vocabulary discipline and clear positioning that nonetheless registered zero visibility because Identity was not yet established. The work was not wasted, but it was not yet harvesting. The harvest requires Identity first.
Above the threshold, Authority and Specificity start producing differential outcomes. They are not identical and they do not collapse into each other. Authority produces visibility on questions about category leadership and rank. Specificity produces visibility on questions about category-defining terminology and methodology. A business strong on Authority but weak on Specificity will appear in answers about "top firms in X" but not in answers using the technical vocabulary of the category. A business strong on Specificity but weak on Authority will appear in answers triggered by category vocabulary but not in ranking-style questions. Strong on both is dominant. Strong on one is competitive. Strong on neither, with Identity above the threshold, is visible but unremarkable.
This hierarchy has implications for sequencing. If Identity is your weakest pillar, the highest-leverage work for the next twelve to eighteen months is Identity-building, regardless of what your competitors are doing on the other pillars. There is no shortcut around it. If Identity is solid and Authority is the gap, the next six to twelve months are Authority work - getting placed on the rankings that matter, getting trade press repetition of comparative claims, niching down if necessary so that you can credibly own a narrower category. If Identity and Authority are both solid and Specificity is the gap, the work is vocabulary discipline and brand-language consistency over years.
Most businesses get the sequencing backward. They reach for Specificity work first because it feels concrete - we can pick three coined terms tomorrow morning - then wonder why nothing changes. Specificity does not move the needle from below the Identity threshold. Sequence matters.
The Distribution Layer
There is a fourth element to the framework that is not a pillar. Calling it a fourth pillar would be wrong because it does different work, operates on different timescales, and has different decision-makers in most organizations. We call it the Distribution Layer.
The pillars describe brand quality at the corpus level. The Distribution Layer describes how brand quality reaches users through different engine architectures. ChatGPT and Gemini are hybrid engines that draw on training data plus retrieval. Perplexity is retrieval-heavy and treats every query as a search-and-summarize task. AI Overviews sits inside Google's ecosystem with its own behavioral patterns. Claude tends to give comprehensive answers with multiple options. Grok pattern-matches aggressively on brand vocabulary. Same brand, same corpus presence, can produce visibility ranging from ten percent on one engine to seventy percent on another, purely because of how each engine surfaces information.
Distribution-layer work is what most current AEO tactical advice is actually addressing. Schema markup. Structured data. llms.txt files. Local directory presence. Content freshness. AI crawler access. None of these change what the corpus believes about your business. All of them affect how that belief is delivered to users when they ask AI engines questions. They are necessary - the floor of credible AEO practice - but they are not sufficient on their own. A business with strong Distribution and weak pillars gets pulled into hyper-specific local or technical queries and stays invisible everywhere else. A business with strong pillars and weak Distribution underperforms on retrieval-heavy engines despite being well-known to the corpus.
The framework treats Distribution as parallel to the pillars rather than downstream of them, because the work to build Distribution is genuinely different work. Pillars are multi-year strategic projects owned by founders, marketing leaders, and PR. Distribution is operational work owned by web teams, agencies, and SEO functions. Conflating them confuses the org chart and conflicts the priorities. A later piece treats Distribution at full operational depth, engine by engine.
Citation as Readout
There is one more concept to introduce before we go deep on the pillars. The concept is Citation, and the temptation is to treat it as a fourth pillar. We did not, for reasons that matter.
Citation is what most operators see when they assess AEO performance. The AI mentions you. With a link, ideally. In a positive context. On a category-relevant prompt. The mention is the visible win, and operators understandably want to know how to produce more of them.
The trap is treating citation as a buildable skill. Operators come to us asking how to optimize for citation, how to produce "citable" content, how to win the AI's recommendation. The honest answer, after a year of running diagnostics across hundreds of businesses, is that citation is not a skill. It is the readout. It is the integrated measurement that tells you whether your three pillars and your Distribution Layer are doing their job collectively.
Brands with strong pillars and adequate Distribution get cited frequently across engines, often with links, with positive sentiment, on category-level prompts. Brands with weak pillars do not get cited regardless of how much they optimize for citability. There is no separate "citation strategy" that produces citations from a state of weak underlying brand presence. Citation tracks the system. It does not move the system.
This reframe matters because it changes what readers should be doing on Monday morning. Stop trying to engineer citations directly. Start tracking them as the dashboard for the rest of the work. Citation rate by engine, mention type with-link versus without-link, sentiment, prompt category coverage - these are the metrics that tell you which pillar is the bottleneck and which Distribution-layer issue needs attention. A later piece treats citation as a measurement framework in operational depth. For now, the relevant point is that citation is the integrated readout of the framework, not a fourth pillar of it.
Why Three Pillars and Not Twelve
The AEO field is currently full of frameworks that list seven, ten, or twelve factors. List density is not insight. Most of those frameworks are inventories of tactics dressed up as architecture. They tell you what to do without telling you what to think. The result is operators with checklists and no model, executing tactics in random order against a system they do not understand.
Three pillars are sufficient because each one names a distinct mechanism, and the three mechanisms together cover the question of what determines corpus-level visibility. Identity names the prerequisite mechanism: existence in the corpus. Authority names the ranking mechanism: settled comparative position. Specificity names the vocabulary mechanism: ownership of category language. Each pillar is independent enough to be diagnosed and built separately. Together they are exhaustive enough to explain the visibility outcomes we see across hundreds of audits.
The framework does not need a fourth pillar because the question "what determines corpus-level visibility" does not need a fourth answer. Adding one would not improve the model. It would dilute it. The Distribution Layer captures everything that is not corpus-level visibility but still affects what users see. Citation captures the readout. The architecture is complete at three pillars, one layer, one readout. Anything more is decoration.
Three is also memorable in a way that twelve is not. The reader who finishes this should be able to name the three pillars without looking. They should be able to use them as a diagnostic lens on any AEO conversation, any vendor pitch, any tactical recommendation. "Which pillar does this address?" is a useful question. "Which of the twelve factors does this address?" is not. The framework is built to be carried in the head, applied in the moment, and used to evaluate every piece of advice you encounter for the rest of your AEO career.
What Comes Next
The pieces that follow take each pillar in turn and go deep: Identity first, then Authority, then Specificity. Each pillar gets a paired treatment: what it is and why it matters first, then how to build it operationally. A closing piece addresses why citation is the readout rather than a fourth pillar, and how the Distribution Layer fits into the framework as a parallel rather than a competing concept.
The next piece runs a diagnostic exercise. Most readers will discover their Identity state is different from what they assumed. The discovery is uncomfortable for some businesses and clarifying for all. Either way, it is the starting point for everything that follows.
If you remember nothing else from this, remember this. Identity is the gate. Authority and Specificity are what happens after the gate is open. Distribution is how brand quality reaches users. Citation is the readout that tells you whether the system is working. Three pillars, one layer, one measurement. That is The Visibility Code.