Key terms and concepts used in Observatory research on AI-mediated property discovery.
The process by which AI systems discover and recommend properties through natural language dialogue, rather than through traditional search interfaces with filters and rankings.
The decision-making process where an AI system selects a subset of properties from a larger pool to present to a user based on their expressed intent and constraints.
A metric representing the percentage of AI evaluations that result in a property being selected as a relevant match for user intent. ASR measures selection upstream of human action.
Structured data elements (amenities, location details, booking policies) that AI systems may use when evaluating properties for conversational queries. Not all signals are used equally.
A structured, machine-readable property record published to the HomeSelf Registry. VPRs provide canonical, AI-accessible property data that may help AI systems better understand and evaluate properties.
The extent to which a property appears in AI responses for relevant conversational queries. High conversational visibility means the property surfaces consistently for matching intents.
A property represented in structured data format (JSON, JSON-LD, AnswerPack) that AI systems can parse, understand, and evaluate without relying on HTML scraping or unstructured text interpretation.
The phenomenon where a large inventory of properties is reduced to a small set of 3–5 recommendations in AI responses. Properties not included in this reduced set effectively do not exist for travelers using AI assistants.
The emerging web paradigm where AI systems (cognitive agents) mediate discovery, evaluation, and selection on behalf of human users, replacing direct human browsing and filtering.
A category or type of property that is underrepresented or absent from AI responses for relevant queries, despite potentially meeting user needs. Gaps may indicate opportunities for improved VPR alignment.
A 0–100 score indicating the degree to which observed conversational patterns map to VPR schema fields. Higher scores suggest stronger correlation between structured property data and AI discovery patterns.
Properties that appear in AI responses for the analyzed prompts. Consistently surfaced properties demonstrate strong conversational visibility for the tested intent.
Access complete benchmark briefs with full analysis and VPR alignment.