Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.linkstonic.com/llms.txt

Use this file to discover all available pages before exploring further.

The LLM citability score is Linkstonic’s central predictive metric: a single number from 0 to 100 that estimates how likely AI models are to cite a specific page or brand in their generated answers. It’s not a guarantee of citation — no score can be — but it is a reliable leading indicator. Pages that improve their citability score consistently see higher mention rates in AI Tracking data within a few weeks of making the changes the score identifies as gaps.

​
What the score measures

The citability score is a composite of five signal categories, each reflecting a different dimension of what makes content trustworthy and useful to AI language models. Content quality signals assess whether your page provides clear, accurate, and comprehensive answers to the questions its target queries imply. AI models favor content that directly answers a question without requiring the reader — or the model — to infer the answer from surrounding context. Content structure evaluates whether your page is organized in a way AI models can parse and excerpt. This includes the presence and relevance of FAQ schema markup, the hierarchy and specificity of H2 and H3 headings, and whether section titles match common query patterns for your topic. Authority signals (E-E-A-T) measure whether your content demonstrates Experience, Expertise, Authoritativeness, and Trustworthiness. Concretely: named authors with visible credentials, a robust About page, external references and citations within your content, and inbound links from authoritative domains. Specificity and depth compare the granularity of your content against the pages AI models currently cite for the same queries. Thin content — pages that cover a topic at a surface level without data, examples, or nuanced explanation — scores poorly here even if it’s well-structured. Link profile checks for broken outbound links, crawlability issues (robots.txt blocks, noindex tags, redirect chains), and whether your internal linking gives AI crawlers a clear path to your most important pages.

​
How to read your score

RangeLabelWhat it means
0–49LowAI models rarely cite this content; significant structural or authority gaps are present
50–74ModerateContent is citation-eligible but is outcompeted by better-structured or higher-authority pages
75–89GoodContent is regularly cited; targeted improvements can push it into the top band
90–100ExcellentContent is among the most citation-friendly in its topic area; focus on maintaining and expanding coverage
A score below 50 typically means there are foundational issues — missing FAQ schema, no author attribution, thin content depth — that need to be resolved before incremental improvements matter. A score between 50 and 74 usually means your content is structurally sound but is losing citation races to competitors with more authoritative signals or better question-answer alignment.

​
Top improvements to raise your score

These are the changes that produce the largest citability score lifts, ordered by the impact-to-effort ratio Linkstonic observes across audited pages.
  1. Add FAQ schema with five or more relevant questions. AI models are trained on and retrieve structured FAQ content at a higher rate than unstructured prose. Each FAQ entry should answer a specific question your target audience asks, not a question that exists to pad the count. Use FAQ schema markup (FAQPage in JSON-LD) so both AI crawlers and Google can parse it cleanly.
  2. Strengthen E-E-A-T signals. Add a named author byline with a link to an author bio page. Ensure your About page names your team, describes your company’s credentials, and includes verifiable contact information. Add citations and references within your content where you make factual claims. These signals are among the strongest predictors of whether an AI model treats your content as authoritative.
  3. Restructure headings to match query patterns. Audit your H2 and H3 headings against the questions people actually ask about your topic. If your headings read like an internal outline (“Overview,” “Key Features,” “Conclusion”) rather than question-aligned sections (“What is [topic]?”, “How does [topic] work?”, “Which [topic] is best for [use case]?”), restructure them. AI models use heading text as a primary signal for what question a section answers.
  4. Add a dedicated “What is X?” section for your main topic. This is the single most reliably cited content pattern across AI platforms. A clear, factual, jargon-free definition section for your core topic — placed early in the page — appears in AI-generated answers at a disproportionately high rate relative to how little effort it takes to add.
  5. Fix broken links and verify crawlability. A page with broken outbound links signals to AI models that the content may be outdated or poorly maintained. Check that your robots.txt doesn’t block AI crawlers (Anthropic’s anthropic-ai, OpenAI’s GPTBot, Google’s Googlebot), that your page doesn’t carry a noindex tag unintentionally, and that your JSON-LD schema validates without errors.
  6. Match the content depth of top-cited competitor pages. Use the citation source list from a TrueTrace Hybrid audit to identify the pages AI models currently cite for your target queries. Read them. If they contain data tables, original research, worked examples, or more granular explanations than your page does, that depth gap is what you need to close.

​
The feedback loop: measuring real improvement

The citability score is an estimate based on content signals. The only way to know whether improvements translated into actual citations is to observe real AI answers and record the results. Use TrueTrace’s feedback loop — available on every audit report — to mark whether AI models actually cited your page after you’ve made changes. Over time, this feedback calibrates Linkstonic’s model to your domain and gives you a real citation rate to compare against the predicted score. After implementing improvements, re-run a TrueTrace audit (Hybrid mode for the most complete picture) and compare the new score against your baseline. Score increases of 5–15 points within two weeks are typical after resolving foundational FAQ and E-E-A-T gaps.
Run TrueTrace in Hybrid mode after implementing each batch of changes. Quick mode gives you a fast directional read, but only Hybrid mode surfaces the full cross-platform citation gap analysis that shows whether a specific improvement landed — and on which AI platforms it had the most impact.

​
Frequently asked questions

No. The citability score is a predictive estimate based on content signals that correlate with AI citation behavior — it is not a direct measure of whether any AI model will cite your page in any given answer. AI-generated answers are probabilistic: the same query can produce different outputs across sessions, and citation behavior varies by platform, query phrasing, and model version. A high citability score means your content is well-positioned to be cited; it doesn’t guarantee citation in every answer. Use the score as a directional guide for content investment, not as a definitive outcome measure.
Your citability score updates each time you run a TrueTrace audit for a URL or brand. Linkstonic does not automatically re-score pages on a fixed schedule — the score reflects the content state at the time of the most recent audit. To track score changes over time, run TrueTrace audits at a consistent cadence (weekly or after significant content changes) and use the historical tracking view to compare results. Your AI Tracking mention rate data, which updates on a rolling basis, is a complementary real-world signal you can monitor between audits.
Several factors can explain this gap. First, AI citation behavior is probabilistic and platform-specific — a score of 80 does not mean you appear in 80% of answers, only that your content is well-structured for citation. Second, your score may be high relative to your previous baseline but still below the threshold of the pages AI models currently cite for your target queries; check the citation source list in your TrueTrace report to see the actual competition. Third, some AI platforms have knowledge cutoffs or retrieval policies that limit which pages they draw from, regardless of content quality. Finally, if your page was recently published or significantly updated, it may not yet be indexed or weighted by all AI platforms. Use the feedback loop to track real citation events and identify whether the gap is a content issue, an indexing issue, or a platform-specific behavior.