brando:VerbalIdentity

Reusable configuration of the brand’s voice, language, and behavioural prompt rules.

rdfs:comment
Reusable configuration of the brand's voice, language, and behavioural prompt rules.

---

Term definition

• IRI: Brando:VerbalIdentity
• Type: owl:Class
• Superclasses:
  • brando:BrandExpression
  • schema:Intangible
• Defined in: brando: (Brando Schema Vocabulary v1.3)

A Brando:VerbalIdentity is a reusable verbal expression preset. It tells AI systems how the brand speaks in a given scope:

• tone and emotional texture,
• conversational / writing style,
• vocabulary do/don’t lists,
• key messages,
• reusable prompt scaffolds,
• hard “must do / must not do” rules at the language level.

You typically do not call a Verbal directly at runtime; instead, you attach it to:

• a brando:Brand, and/or
• a brando:Context,

via brando:usesVerbal. Your Brand OS or AI middleware then reads the token and turns it into prompts, system messages, and constraints.

---

Relationships

Inbound: where Verbals are used

Verbals are attached from brands and contexts via:

• brando:usesVerbal
  • Type: owl:ObjectProperty
  • Domain: brando:Brand, brando:Context
  • Range: Brando:VerbalIdentity
  • Comment: “Attaches a verbal expression token defining how the brand speaks in this scope.”

Example (brand & context attaching the same token):

{
  "@id": "https://example.com/brand/northstar",
  "@type": "brando:Brand",
  "schema:name": "Northstar Bank",
  "brando:usesVerbal": {
    "@id": "https://example.com/brand/northstar/tokens/verbal/default"
  }
}
````

```jsonld
{
  "@id": "https://example.com/brand/northstar/context/retail-chatbot",
  "@type": "brando:Context",
  "brando:usesVerbal": {
    "@id": "https://example.com/brand/northstar/tokens/verbal/support"
  }
}

A common pattern (non-normative):

• Brand-level Verbal → default voice.
• Context-level Verbal → specialisation or override (e.g. “support voice”, “investor voice”).

---

Key properties (domain = Brando:VerbalIdentity)

The following properties have Brando:VerbalIdentity in their rdfs:domain.

1. Tone, dialogue, writing, and persona

Property IRI	Range	Type	Comment (from vocab)
brando:toneOfVoice	xsd:string	owl:DatatypeProperty	Emotional texture and attitude of the brand's voice in this token.
brando:dialogueStyle	xsd:string	owl:DatatypeProperty	Guidance on conversational patterns, pacing, and pronoun usage.
brando:writingStyle	xsd:string	owl:DatatypeProperty	Preferences for grammar, sentence structure, and clarity.
brando:namingConvention	xsd:string	owl:DatatypeProperty	Rules for how products, features, dates, and other entities are named and formatted.
brando:personaCard	xsd:string	owl:DatatypeProperty	Humanised brand-as-a-person description used to guide tone and behaviour.

Usage notes (non-normative):

• Keep these fields focused and operational – they should be easy for an LLM or agent to turn into actual behaviour.
• personaCard is often a short paragraph (“If this brand were a person…”).
• dialogueStyle and writingStyle can be brief bullet-like sentences or more structured text.

Example:

{
  "@id": "https://example.com/brand/northstar/tokens/verbal/default",
  "@type": "Brando:VerbalIdentity",
  "brando:toneOfVoice": "Calm, empathetic, precise. Never rushed, never flippant.",
  "brando:dialogueStyle": "Use second person ('you'), ask clarifying questions before answering complex queries.",
  "brando:writingStyle": "Short paragraphs, concrete examples, no jargon without explanation.",
  "brando:namingConvention": "Use full product names on first mention, then short names. Dates in ISO (YYYY-MM-DD).",
  "brando:personaCard": "If Northstar were a person, they’d be a thoughtful financial coach: calm, patient, and pragmatic."
}

---

2. Vocabulary and messaging

Property IRI	Range	Type	Comment (from vocab)
brando:approvedTerms	xsd:string	owl:DatatypeProperty	Words and phrases that are encouraged in this verbal token.
brando:prohibitedTerms	xsd:string	owl:DatatypeProperty	Words and phrases that must not be used in this verbal token.
brando:keyMessages	xsd:string	owl:DatatypeProperty	Core value propositions and ideas that should recur across outputs.
brando:tagline	xsd:string	owl:DatatypeProperty	Primary tagline plus any sanctioned variants for this verbal token.

Usage notes (non-normative):

• In JSON-LD, you can supply multiple values by:

  • repeating the property; or
  • using a JSON array (depending on your tooling).
  • approvedTerms / prohibitedTerms are per Verbal – you can vary them across contexts or campaigns.
  • keyMessages are useful raw material for RAG / prompt construction.

Example:

{
  "@id": "https://example.com/brand/northstar/tokens/verbal/default",
  "@type": "Brando:VerbalIdentity",
  "brando:approvedTerms": [
    "interest rate",
    "fees",
    "savings goals"
  ],
  "brando:prohibitedTerms": [
    "get rich quick",
    "guaranteed returns"
  ],
  "brando:keyMessages": [
    "We help you understand trade-offs, not hide them.",
    "We prioritise long-term financial health over short-term hype."
  ],
  "brando:tagline": "Clarity, not shortcuts."
}

---

3. Prompt scaffolds and behavioural rules

Property IRI	Range	Type	Comment (from vocab)
brando:promptScaffold	xsd:string	owl:DatatypeProperty	Reusable instruction block or template for priming language models.
brando:mustDos	xsd:string	owl:DatatypeProperty	Required behaviours and inclusions this verbal token must follow.
brando:mustNotDos	xsd:string	owl:DatatypeProperty	Behaviours and constructions this verbal token must never produce.

Usage notes (non-normative):

• promptScaffold is designed to be machine-consumed:

  • think of it as a reusable system-prompt fragment you can slot into LLM calls.
  • it can include placeholders your runtime fills in.

  • mustDos / mustNotDos are best expressed as short, imperative sentences:

  • easier for both humans and models to interpret.

Example:

{
  "@id": "https://example.com/brand/northstar/tokens/verbal/support",
  "@type": "Brando:VerbalIdentity",
  "brando:promptScaffold": "You are speaking as Northstar Bank's support assistant. Be calm, empathetic, and precise. Explain options simply and always disclose trade-offs.",
  "brando:mustDos": [
    "Acknowledge the user's concern before giving instructions.",
    "Offer a next-best action if you cannot fully resolve the issue."
  ],
  "brando:mustNotDos": [
    "Do not speculate about regulatory or legal outcomes.",
    "Do not tell the user what they 'should' feel."
  ]
}

At runtime, your Brand OS might:

1. Load the relevant Verbal.
2. Insert promptScaffold into the system prompt.
3. Convert mustDos / mustNotDos into explicit instructions or post-check rules.

---

4. Files and supporting material

These are shared properties used across several Brando types; for Verbals they typically point at voice/tone guidelines and reference materials.

Property IRI	Range	Type	Comment (from vocab, summarised)
brando:downloadableFile	schema:MediaObject	owl:ObjectProperty	Links a brand, token, context, campaign, or policy node to the most up-to-date creative files exposed by the Brand OS or DAM.
brando:supportingFile	schema:MediaObject	owl:ObjectProperty	Links a brand, token, context, campaign, or policy node to additional files helpful for production but not the primary source-of-truth asset.

Usage notes (non-normative):

• For Verbals:

  • downloadableFile → canonical voice/tone guidelines (e.g. “Northstar Voice & Tone Guidelines (PDF)”).
  • supportingFile → examples, workshop decks, annotated transcripts, etc.

  • These schema:MediaObject nodes can themselves carry:

  • schema:contentUrl,

  • schema:encodingFormat,
  • and other standard Schema.org metadata.

Example:

{
  "@id": "https://example.com/brand/northstar/tokens/verbal/default",
  "@type": "Brando:VerbalIdentity",
  "brando:downloadableFile": {
    "@id": "https://assets.example.com/northstar/voice/northstar-voice-tone.pdf",
    "@type": "schema:MediaObject",
    "schema:name": "Northstar Voice & Tone Guidelines (PDF)",
    "schema:contentUrl": "https://assets.example.com/northstar/voice/northstar-voice-tone.pdf"
  },
  "brando:supportingFile": {
    "@id": "https://assets.example.com/northstar/voice/example-conversations.pdf",
    "@type": "schema:MediaObject",
    "schema:name": "Example Conversations – Retail Support"
  }
}

---

Example: Brand + Context + Verbals

A combined JSON-LD example showing how Brando:VerbalIdentity works with brando:Brand and brando:Context:

{
  "@context": {
    "schema": "https://schema.org/",
    "brando": "https://brandoschema.com/"
  },
  "@graph": [
    {
      "@id": "https://example.com/brand/northstar",
      "@type": "brando:Brand",
      "schema:name": "Northstar Bank",
      "brando:hasContext": {
        "@id": "https://example.com/brand/northstar/context/retail-chatbot"
      },
      "brando:usesVerbal": {
        "@id": "https://example.com/brand/northstar/tokens/verbal/default"
      }
    },
    {
      "@id": "https://example.com/brand/northstar/context/retail-chatbot",
      "@type": "brando:Context",
      "brando:audienceSegment": "Retail banking customers",
      "brando:domainContext": {
        "channel": "chatbot",
        "surface": "public-website",
        "region": "UK"
      },
      "brando:usesVerbal": {
        "@id": "https://example.com/brand/northstar/tokens/verbal/support"
      }
    },
    {
      "@id": "https://example.com/brand/northstar/tokens/verbal/default",
      "@type": "Brando:VerbalIdentity",
      "brando:toneOfVoice": "Calm, empathetic, precise. Explain trade-offs clearly.",
      "brando:approvedTerms": [
        "interest rate",
        "fees",
        "savings goals"
      ],
      "brando:prohibitedTerms": [
        "get rich quick",
        "guaranteed returns"
      ],
      "brando:keyMessages": [
        "We help you understand your options.",
        "We are transparent about risks and costs."
      ],
      "brando:tagline": "Clarity, not shortcuts."
    },
    {
      "@id": "https://example.com/brand/northstar/tokens/verbal/support",
      "@type": "Brando:VerbalIdentity",
      "brando:toneOfVoice": "Reassuring and efficient. Focus on clear next steps.",
      "brando:dialogueStyle": "Acknowledge feelings, then move quickly to options.",
      "brando:promptScaffold": "You are Northstar Bank's support assistant for UK retail customers...",
      "brando:mustDos": [
        "Confirm understanding of the user's question before giving instructions."
      ],
      "brando:mustNotDos": [
        "Do not blame the user for mistakes.",
        "Do not minimise concerns."
      ]
    }
  ]
}

Runtime pattern (non-normative):

1. Resolve Brand + Context for the request.

2. Load Verbals from:

   • Brand (/tokens/verbal/default),
   • Context (/tokens/verbal/support).

   • Merge or layer properties according to your precedence rules:

   • e.g. context-specific promptScaffold takes precedence over brand default.

3. Compile:

   • system prompts using promptScaffold, toneOfVoice, dialogueStyle, writingStyle,
   • lexical constraints using approvedTerms, prohibitedTerms, mustDos, mustNotDos.

---

Related documentation

• Concepts / overview

  • Brando types overview
  • Getting started
  • Quickstart (JSON-LD)
  • Quickstart (YAML)

• Related types

  • brando:Brand
  • brando:Context
  • brando:BrandExpression
  • Brando:VisualIdentity
  • Brando:AudioIdentity
  • brando:Policy