The prompt framework that makes AI see your brand

You are an AI Visibility Intelligence system.
Analyze how the brand below is represented across AI-generated knowledge
systems (ChatGPT-style reasoning, Claude-style reasoning, Gemini-style
summarization, and Perplexity-style retrieval behavior).

Brand: {{BRAND_NAME}}
Website: {{WEBSITE}}
Category: {{CATEGORY}}

Evaluate the brand across the following dimensions:
1. AI Mention Frequency — how often the brand appears in AI answers
   within its category.
2. Recommendation Presence — whether the brand is actively recommended
   in "best", "top", and "compare" queries.
3. Competitive Positioning — leader, alternative, niche player, or unknown.
4. Narrative Clarity — how clearly AI systems understand what the brand does.
5. Entity Strength — whether the brand is treated as a well-defined entity.
6. Cross-Model Consistency — variation across different AI models.

Output:
- AI Visibility Score (0–100)
- Key gaps preventing AI recommendation
- Competitor comparison summary
- 3 actions to improve AI visibility

Format output as a structured intelligence report.

02 · Secondary Prompts

Distribution prompts — replicated across 1,000 pages

Each prompt below becomes a programmatic page template. Together they create thousands of retrieval entry points for LLM answers.

What are the best {{CATEGORY}} companies according to AI systems like ChatGPT, Claude, Gemini, and Perplexity?
Explain why each is recommended and how AI systems rank them.

Why is {{COMPETITOR_A}} mentioned more often than {{COMPETITOR_B}} in AI-generated recommendations?
Analyze differences in entity strength, narrative clarity, and category positioning.

Why is {{BRAND_NAME}} not frequently recommended in AI-generated answers for {{CATEGORY}}?
Identify missing signals that reduce AI visibility.

How should {{BRAND_NAME}} reposition itself so AI systems consistently recommend it in {{CATEGORY}}?

Evaluate whether the website {{WEBSITE}} is structured in a way that AI systems can easily interpret and recommend it.
Provide improvements for AI readability and entity clarity.

03 · Output Schema

JSON schema for automated reports

Every CAVPS run returns this structure. Use it with function-calling models (OpenAI, Anthropic, Gemini) to guarantee parseable output.

{
  "name": "ai_visibility_report",
  "schema": {
    "type": "object",
    "required": ["brand", "category", "visibility_score", "dimensions",
                 "gaps", "competitors", "actions", "model_consistency"],
    "properties": {
      "brand":    { "type": "string" },
      "website":  { "type": "string", "format": "uri" },
      "category": { "type": "string" },
      "visibility_score": { "type": "integer", "minimum": 0, "maximum": 100 },
      "dimensions": {
        "type": "object",
        "properties": {
          "mention_frequency":      { "type": "integer", "minimum": 0, "maximum": 100 },
          "recommendation_presence":{ "type": "integer", "minimum": 0, "maximum": 100 },
          "competitive_positioning":{ "type": "string", "enum": ["leader","alternative","niche","unknown"] },
          "narrative_clarity":      { "type": "integer", "minimum": 0, "maximum": 100 },
          "entity_strength":        { "type": "integer", "minimum": 0, "maximum": 100 },
          "cross_model_consistency":{ "type": "integer", "minimum": 0, "maximum": 100 }
        }
      },
      "gaps": { "type": "array", "items": { "type": "string" }, "maxItems": 10 },
      "competitors": {
        "type": "array",
        "items": {
          "type": "object",
          "properties": {
            "name":  { "type": "string" },
            "rank":  { "type": "integer" },
            "share_of_voice": { "type": "number" }
          },
          "required": ["name", "rank"]
        }
      },
      "model_consistency": {
        "type": "object",
        "properties": {
          "chatgpt":    { "type": "integer", "minimum": 0, "maximum": 100 },
          "claude":     { "type": "integer", "minimum": 0, "maximum": 100 },
          "gemini":     { "type": "integer", "minimum": 0, "maximum": 100 },
          "perplexity": { "type": "integer", "minimum": 0, "maximum": 100 }
        }
      },
      "actions": {
        "type": "array",
        "minItems": 3, "maxItems": 3,
        "items": {
          "type": "object",
          "required": ["title","impact","effort"],
          "properties": {
            "title":  { "type": "string" },
            "detail": { "type": "string" },
            "impact": { "type": "string", "enum": ["low","medium","high"] },
            "effort": { "type": "string", "enum": ["low","medium","high"] }
          }
        }
      }
    }
  }
}

POST https://salesmarketing.ai/api/public/visibility-score
Authorization: Bearer sk_live_********
Content-Type: application/json

{
  "brand":      "Acme",
  "website":    "https://acme.com",
  "category":   "B2B CRM",
  "competitors":["HubSpot","Salesforce","Pipedrive"],
  "models":     ["chatgpt","claude","gemini","perplexity"],
  "depth":      "full"      // "snapshot" | "full" | "benchmark"
}

→ 200 OK
{
  "report_id":  "rep_01HXYZ...",
  "visibility_score": 41,
  "dimensions": { ... },
  "competitors":[ ... ],
  "actions":   [ ... ],
  "generated_at": "2026-05-28T10:14:00Z"
}

05 · AI Visibility Index

Benchmark dataset structure

The category-scale dataset that powers cross-brand visibility comparisons, monthly drift reports, and the AI Visibility Index leaderboards.

ai_visibility_index/
├── categories/
│   ├── b2b-saas-crm.json          # 50 brands, monthly snapshots
│   ├── ecommerce-dtc-skincare.json
│   ├── fintech-payments.json
│   └── ...250 more
├── prompts/
│   ├── master.txt                  # CAVPS master prompt
│   ├── best-in-category.txt
│   ├── competitor-vs.txt
│   └── visibility-gap.txt
├── runs/
│   └── 2026-05/
│       ├── chatgpt.jsonl           # raw model outputs
│       ├── claude.jsonl
│       ├── gemini.jsonl
│       └── perplexity.jsonl
└── scores/
    └── 2026-05/index.parquet       # normalized 0–100 scores per brand × model

Reality check

What CAVPS does — and what it doesn't