Problem

Schema.org describes what entities are but lacks properties for two dimensions AI recommendation systems need: intended use context (for whom?) and brand personality (how does it feel?). Publishers work around this with additionalProperty / PropertyValue, but without standardization.

Proposal

intendedUseContext

• Type: Text
• On: Organization, Brand, Product, Service, LocalBusiness, CreativeWork
• Definition: Descriptive statement of the intended audience, use case, or situation for which this entity is most relevant.
  Complements audience (which expects a structured Audience object) with a lightweight natural-language alternative — similar to how speakable complements description.

{
  "@type": "Product",
  "name": "Organic Pepper Mix",
  "intendedUseContext": "Intended for home cooks seeking premium organic spices for Mediterranean cooking. Best suited when ingredient quality matters more than price."
}

brandPersonality

• Type: Text or DefinedTerm
• On: Organization, Brand, Product, Service, LocalBusiness
• Definition: Descriptors characterizing the emotional identity of a brand or product, as defined by the publisher.
  Accepting DefinedTerm allows linking to Wikidata concepts for verifiability — mitigating the "meta-keywords" spam risk.

{
  "@type": "Organization",
  "name": "Example Spice Company",
  "brandPersonality": [
    { "@type": "DefinedTerm", "name": "Family business", "sameAs": "https://www.wikidata.org/wiki/Q955824" },
    { "@type": "DefinedTerm", "name": "Organic food", "sameAs": "https://www.wikidata.org/wiki/Q191067" }
  ]
}

Compatibility

Non-breaking extension. Both properties are optional.

Reference

• Open-source spec + implementation: vibetags/vibetags-spec (MIT)