AI Visibility Evidence Model: Five Factors, Graded by Evidence | Stefan Petschinka | richresults.ai

richresults.ai<br>what is AEO<br>how AEO works<br>who is AEO for<br>case studies<br>about<br>contact &rarr;<br>DE | EN

Reference Document

The AI Visibility<br>Evidence Model.

01 Definition

What the AI Visibility Evidence Model is.

The AI Visibility Evidence Model is a reference model that orders the publisher-side factors behind AI visibility by strength of evidence. It defines five factors, Topical Relevance, Machine Access, Entity Consistency, Extractability and Independent Corroboration, and assigns each a documented evidence grade based on peer-reviewed research, controlled preprints and official platform documentation.

The model exists because the field still lacks a concise publisher-side reference that maps the main actionable factors to explicit evidence grades and primary sources. Every factor in the model carries its grade and its sources, so every statement on this page can be checked against the primary literature listed in the source register below.

02 Purpose and Boundary

A map of the evidence, not a methodology.

The AI Visibility Evidence Model maps what the evidence shows works. The AEO Mastery Framework describes how richresults.ai implements it. The model is descriptive: it reports the state of the research. The framework is prescriptive: it defines a working method. Neither replaces the other.

The model is not a ranking system, not a score and not a promise of results. AI visibility is a distribution across repeated, non-deterministic answers, and no factor in this model guarantees a citation. What the model provides is priority: it tells you which work is supported by evidence, which work is hygiene, and which work the evidence contradicts. That is less modest than it sounds. Knowing which factor rests on which evidence tells you what to start with and what can wait. A promise of results cannot tell you that.

One boundary is deliberate: the model orders the factors a publisher can work on. The retrieval stage itself, which engine selects which sources, how it ranks them and where it places them in the model context, is system-side. Controlled work shows this stage dominates citation outcomes [2], and platform documentation describes engine-specific retrieval decisions, including when a system grounds at all [13]. All five factors work toward that stage; none of them controls it.

03 The Evidence Scale

Four grades, defined before use.

Grade A: peer-reviewed and controlled, or independently replicated.

Grade B: controlled with limited transferability to open production systems, or an official platform statement.

Grade C: correlational or triangulated across independent datasets, without causal proof.<br>Grade D: unsupported or contradicted by evidence.

Each factor additionally carries its mechanism type. A gate is a binary precondition, a driver influences outcomes gradually, and hygiene reduces errors without creating an advantage. Mechanism type and evidence grade are separate dimensions: a factor can be a hard gate on weak empirical evidence, or a soft driver on strong evidence.

These four grades are this model's own scale, and it is deliberately conservative. A factor rated C is not unimportant; it means no one has isolated its causal effect in production, and anyone claiming otherwise is claiming more than the data supports. Grades move as evidence accumulates, and this page carries a visible update date for that reason.

04 The Five Factors

Factor 1: Topical Relevance

Content that directly addresses the actual question is the strongest documented content-side driver of citation. In the largest controlled citation study to date, 252,000 trials across six language models, topic match to the query and position in the model context were the dominant factors, and off-topic content was practically never cited first [1]. Controlled work confirms the same pattern from the model side: when weighing conflicting evidence, models rely heavily on a page's relevance to the query while largely ignoring stylistic authority signals such as scientific-looking references or neutral tone [14]. No entity work, no markup and no authority signal compensates for content that does not answer the question being asked.

Mechanism: driver. Evidence grade: A. Peer-reviewed, controlled, convergent across models and study designs [1, 2, 14].

Factor 2: Machine Access

A source that crawlers cannot reach cannot be retrieved, and a source that cannot be retrieved cannot be used for the content of an answer. Machine Access covers crawl permissions for the relevant bots, index presence, crawlable and renderable main content, and firewall configurations that do not silently block AI crawlers. Platform documentation is explicit on both sides of this gate. OpenAI requires OAI-SearchBot access for a site's content to be used in ChatGPT search answers; excluded pages can still appear as navigational links [12]. Google requires...