The same underlying intent, phrased differently, produces different brand citations across AI engines. Brands that understand this pattern can optimize for the query types that drive their highest-intent traffic.
LLM citation is sensitive to query phrasing in ways that traditional SEO is not. Two queries with the same underlying intent — "best project management software" vs "top tools for managing engineering sprints" — can produce entirely different brand citation sets. This is not a bug; it is a feature of how retrieval-augmented generation works.
How query sensitivity affects brand citation
LLMs resolve queries by retrieving documents that closely match the query's semantic embedding. Narrow, specific queries retrieve from a smaller document pool — which means brands with content that precisely matches that framing have an outsized advantage. Broad queries retrieve from a large pool and favor brands with the highest overall entity authority.
The practical implication: your citation probability depends on which query types are most common for your category — and whether your content and entity signals are optimized for those specific phrasings.
The three query types that drive the most citations
- Category queries: "what is the best X" — retrieves based on entity authority. Brands with strong Wikidata, press coverage, and Common Crawl presence win here.
- Job-title queries: "what do [role] teams use for X" — retrieves based on ICP-specific content. Brands with structured, role-specific content and FAQPage schema win here.
- Feature queries: "what tool does X" — retrieves based on structured data precision and feature-specific landing pages. Brands with complete Organization schema and detailed feature content win here.
Measuring your query sensitivity
Run the same query five different ways across ChatGPT, Claude, and Perplexity. Compare your citation rate across phrasings. If you appear consistently on category queries but rarely on job-title queries, you have an ICP content gap. If you appear on narrow feature queries but not broad category queries, you have an entity authority gap.
| Query type | Primary signal | Fix if missing |
|---|---|---|
| Category ("best X") | Entity authority | Wikidata + press + CC presence |
| Job-title ("for [role]") | ICP-specific content | FAQPage schema + role-specific pages |
| Feature ("does X") | Structured data precision | Organization schema + feature markup |
| Brand-direct ("what is Y") | Entity graph completeness | Wikidata + Wikipedia |
| Comparison ("X vs Y") | Earned media + reviews | Press + Reddit + earned citations |
What this means
Query sensitivity makes citation monitoring more complex than a single score suggests. A brand that appears in 30% of category queries but 0% of job-title queries has a specific, fixable problem — not a general visibility problem.
The FAQPage schema playbook covers the fastest way to capture job-title and feature query traffic. The earned media playbook covers how to build the entity authority needed for broad category queries. Start by understanding which query type is driving your highest-intent prospects, and optimize for that layer first. Scan your domain to see your current structural citation signals.
Put this into practice
See how your domain scores on the signals covered in this edition. Veezow runs a free AI visibility scan — robots, sitemap, structured data, bot access, and off-site presence.
Run a free scan →