Gen Visibility

Gen Visibility is an AI-powered tool that enhances human-written content to be cited as a source in AI-generated answers. It reverse-engineers how AI systems select sources and generates structured content that matches those patterns.

This is not traditional SEO.

Traditional SEO optimizes for search engine rankings. Generative Visibility optimizes for being cited as a source by ChatGPT, Claude, Perplexity, and Google AI Overviews.

🆕 March 2026 Release

• 3-Step Pipeline: Query Generation → Pattern Analysis → Content Layer
• Cost-Efficient: $0.01–0.05 per analysis using GPT-4o-mini + GPT-4o
• Location-Aware: Optional local content generation for geographic targeting
• Citation Tracking: Monitor whether AI systems cite your content over time

How It Works

When someone asks ChatGPT "What's the best Mercedes for families?", the AI assembles an answer from its training data and any retrieved sources. Gen Visibility analyzes what those queries look like, how AI systems structure their answers, and what patterns trigger source citation — then generates content that matches.

The Core Insight

🔍

Traditional SEO

Optimize for search ranking

→

🧠

Generative Visibility

Optimize for AI citation

AI systems prefer to cite sources that are structured, specific, authoritative, and directly answer the queries users are asking. Gen Visibility ensures your content matches all four criteria.

The Pipeline

Every analysis runs through a 3-step pipeline, each powered by AI:

1

Query Generation

GPT-4o-mini

Generates 30–50 natural language queries that real users would type into ChatGPT, Claude, Perplexity, or Google AI Overviews about your topic.

Example queries for "Mercedes GLE family car":

• "best family Mercedes SUV"
• "which Mercedes has 3 rows"
• "Mercedes GLE vs BMW X5 for families"
• "is the Mercedes GLE safe for kids"
• "Mercedes GLE cargo space with car seats"

2

Pattern Analysis

GPT-4o-mini

Analyzes how AI systems would typically answer these queries. Identifies key entities, answer structure patterns, citation triggers, and content gaps.

🏷️ Entities

GLE 350, GLE 450, GLS 580, MBUX, 4MATIC

🎯 Citation Triggers

Specific prices, safety ratings, cargo dimensions

📐 Patterns

Comparison tables, bullet lists, Q&A format

🕳️ Content Gaps

Car seat compatibility, real-world MPG data

3

Content Layer Generation

GPT-4o

Generates structured, AI-friendly content sections to append to your original article. Uses higher-quality model for authoritative, cite-worthy output.

Key principle: The original human-written content is never modified. Gen Visibility only generates supplementary sections designed to be appended.

How to Use

1

Enter Your Topic

The subject of your article. Be specific — "Mercedes GLE family car" works better than just "Mercedes".

2

Add Location (Optional)

If your content targets a specific area, add it. This generates location-specific queries and local insights (e.g., "Chicago winters" or "Palm Springs luxury").

3

Paste Your Article

Paste the full text of your human-written article. The tool uses this to understand context and generate relevant supplementary content.

4

Click "Analyze & Generate AI Layer"

The 3-step pipeline runs (typically 15–30 seconds). You'll see real-time progress updates.

5

Review & Copy

Review the generated content across three tabs (Content Layer, Queries, Analysis). Click "Copy Markdown" to copy the AI layer and append it to your article.

Output Sections

The generated AI Layer contains these structured sections:

📋 Quick Answers

Bullet-point answers to the most common user queries. Designed for AI systems to extract and quote directly.

❓ Frequently Asked Questions

5–10 Q&A pairs matching real user queries. Each answer is 2–3 sentences, specific, and authoritative.

⚖️ Comparisons

Structured comparison blocks (e.g., "GLC vs GLE"). AI systems frequently generate comparison content and cite sources that already have it structured.

📊 Key Specifications

Data tables and structured lists with specific numbers, features, and specs. Specificity is a key citation trigger.

📍 Local Insights (when location provided)

Location-specific recommendations and context. Critical for local businesses and regional content.

🎓 Expert Summary

2–3 paragraph authoritative summary that AI systems would want to quote when synthesizing information.

Use Cases

🏖️ Travel & Luxury

Optimize travel guides and luxury lifestyle content for AI citation.

Topic: "Luxury Palm Springs resorts" • Location: "Palm Springs"

🚗 Automotive

Car reviews and comparisons structured for AI answer generation.

Topic: "Mercedes GLE family car" • Location: "Chicago"

🏢 B2B / Enterprise

Product pages and thought leadership optimized for AI-assisted research.

Topic: "Enterprise data analytics platforms"

🏠 Real Estate

Property listings and neighborhood guides for AI-powered home search.

Topic: "Best neighborhoods in Portland" • Location: "Portland, OR"

🍽️ Food & Restaurant

Restaurant reviews and food guides for AI recommendation engines.

Topic: "Best Italian restaurants" • Location: "San Francisco"

💻 SaaS / Technology

Software comparison pages and feature documentation for AI-assisted tool selection.

Topic: "CRM software for small business"

API Reference

POST /api/analyze

Run the full analysis pipeline on an article.

curl -X POST https://gen-visibility.scottfelten.com/api/analyze \
  -H "Content-Type: application/json" \
  -d '{
    "topic": "Mercedes GLE family car",
    "article": "The Mercedes GLE is one of the most popular...",
    "location": "Chicago"
  }'

Parameter	Type	Required	Description
`topic`	string	Yes	The subject of your article
`article`	string	Yes	Full article text (up to 5MB)
`location`	string	No	Geographic context for local optimization

GET /api/analyses

List all previous analyses with cost and timing data.

curl https://gen-visibility.scottfelten.com/api/analyses

GET /api/analysis/:id

Retrieve a specific analysis with full results.

curl https://gen-visibility.scottfelten.com/api/analysis/1

GET /api/health

Service health check.

curl https://gen-visibility.scottfelten.com/api/health
# {"status":"ok","service":"gen-visibility","version":"1.0.0"}

AI Models Used

Gen Visibility uses a dual-model strategy for cost efficiency and quality:

Pipeline Step	Model	Why	Cost/1M Tokens
Query Generation	GPT-4o-mini	Fast, creative, cost-efficient for brainstorming	$0.15 in / $0.60 out
Pattern Analysis	GPT-4o-mini	Good analytical capability at low cost	$0.15 in / $0.60 out
Content Generation	GPT-4o	Higher quality, authoritative writing	$2.50 in / $10.00 out

Model Recommendation: The default configuration balances cost and quality. Steps 1–2 use Mini for speed and cost savings. Step 3 uses GPT-4o because content quality directly impacts citation probability — this is where quality matters most.

Pricing & Costs

$0.01 – $0.05

Per analysis (typical range)

Article Length	Estimated Cost	Processing Time
Short (500 words)	~$0.01	10–15 seconds
Medium (1,500 words)	~$0.02–0.03	15–25 seconds
Long (3,000+ words)	~$0.03–0.05	20–30 seconds

Cost Efficiency: At $0.03 average per analysis, you can run 1,000 articles for approximately $30. Compare this to traditional SEO tools at $100–500/month.

Best Practices

✅

DO: Write for humans first

Your original article should be naturally written. The AI layer supplements — it doesn't replace.

✅

DO: Include specific data

Prices, dimensions, ratings, dates — AI systems strongly prefer specificity over vagueness.

✅

DO: Use location when relevant

Local queries are growing fast in AI search. Location context generates highly targeted content.

✅

DO: Re-run periodically

AI query patterns evolve. Re-analyze quarterly to keep your content aligned with current patterns.

❌

DON'T: Replace your original content

The AI layer is supplementary. Append it, don't substitute.

❌

DON'T: Keyword stuff

AI systems detect and penalize low-quality generation. The tool is designed to avoid this.

❌

DON'T: Use without review

Always review generated content for accuracy before publishing. You are the authority — the tool is your assistant.

Citation Tracking

Gen Visibility includes a citation tracking system to measure whether your content is being cited by AI systems over time.

Metrics Tracked

📊 Citation Rate

Percentage of tracked queries where your content appears as a cited source.

📈 Coverage Index

How many of the generated queries your content effectively answers.

🏆 Competitive Sources

Which other sources are being cited for the same queries.

Architecture

Article + Topic + Location
        │
        ▼
┌─────────────────────┐
│  Query Generator     │  GPT-4o-mini  →  30-50 queries
│  (Step 1)           │
└─────────┬───────────┘
          │
          ▼
┌─────────────────────┐
│  Pattern Analyzer    │  GPT-4o-mini  →  entities, triggers, gaps
│  (Step 2)           │
└─────────┬───────────┘
          │
          ▼
┌─────────────────────┐
│  Content Generator   │  GPT-4o      →  structured AI layer
│  (Step 3)           │
└─────────┬───────────┘
          │
          ▼
┌─────────────────────┐
│  SQLite Database     │  Persist results, track citations
└─────────────────────┘

Tech Stack

Backend	Node.js + Express
Database	SQLite (better-sqlite3)
AI	OpenAI API (GPT-4o-mini + GPT-4o)
Frontend	Vanilla HTML + Tailwind CSS (CDN)
Hosting	VPS + nginx + systemd + SSL
Repository	github.com/scottfelten71/gen-visibility

📡

Signal Engine

NEW

The Signal Engine measures whether your content is actually being cited by AI systems. It's the measurement companion to the Gen Visibility content optimizer — together they form a complete create → measure → improve loop.

This is AI SEO observability.

Gen Visibility creates AI-optimized content. Signal Engine tells you if it's working. Nobody else has this closed loop.

The Create → Measure → Improve Loop

🧠

1. Create

Gen Visibility optimizes your content for AI citation

📡

2. Measure

Signal Engine probes AI and scores your visibility

📈

3. Improve

Use insights to re-optimize and track score over time

Measurement Pipeline

Every Signal Engine analysis runs through a 4-step pipeline:

1

Content Ingestion

GPT-4o-mini

Scrapes the target URL (or accepts raw text), strips HTML, and uses AI to extract structured elements: headings, entities, Q&A sections, unique phrases, and key claims.

Extracts:

• Page title & headings

• Named entities

• Q&A sections

• Unique phrases (3-7 words)

• Structured blocks

• Key factual claims

2

AI Visibility Probe

GPT-4o-mini

Generates 15 realistic user queries from your content, then sends each query to an AI system. For each response, it checks for brand mentions, content phrase matches, and entity alignment.

For each probe, evaluates:

• Brand mention: Does the AI mention your brand/site by name?
• Phrase match: Does the AI response contain your content's unique phrases?
• Entity alignment: Does the AI reference the same entities as your content?
• Confidence score: 0-100% likelihood your content influenced the answer

3

Scoring Engine

Local computation

Aggregates all probe signals into a single 0-100 AI Visibility Score using weighted components. Identifies key drivers and confidence level.

4

Report Generator

GPT-4o

Generates an executive-ready visibility report with evidence, analysis, and specific recommendations. Written in measured, professional tone — never claims direct attribution.

Behavioral rules: The report uses language like "indicates", "suggests", and "correlates with" — never "proves" or "caused by". Signal triangulation over certainty.

Engine Modules

📥 Content Ingestion Module

Scrapes and structures content for analysis.

Input	URL or raw text
Output	Entities, questions, headings, unique phrases, key claims, word count
Model	GPT-4o-mini

🔍 AI Visibility Probe Module

Simulates real AI search queries and checks for content citation.

Input	Content analysis + brand name
Output	15 probe results with brand mentions, phrase matches, confidence scores
Model	GPT-4o-mini (for query gen + probes)

🧮 Scoring Engine Module

Aggregates signals into weighted visibility score.

Input	Content analysis + probe results
Output	0-100 score, confidence level, key drivers
Model	None (computed locally)

📊 Report Generator Module

Produces client-ready executive visibility report.

Input	All analysis data
Output	Markdown report with 6 sections
Model	GPT-4o (higher quality for client-facing output)

Extensibility: Each module is independently callable. Future integrations include Google Search Console data, GA4 traffic patterns, Perplexity API probing, and scheduled monitoring runs.

Scoring Model

The AI Visibility Score (0-100) is calculated from five weighted components:

Component	Weight	What It Measures
AI Brand Mention Rate	35%	Percentage of probe queries where AI mentions your brand by name
Content Phrase Reuse	25%	How many of your content's unique phrases appear in AI responses
Entity Alignment	20%	Whether AI references the same specific entities (products, places, numbers) as your content
Content Structure Quality	10%	How well-structured your content is for AI consumption (Q&As, lists, headings)
Average Probe Confidence	10%	Overall confidence that your content influenced AI answers

Score Interpretation

0–39: Weak

AI systems rarely reference your content. Significant optimization needed.

40–69: Moderate

Some AI visibility. Content is partially aligned with AI citation patterns.

70–100: Strong

Content is well-positioned for AI citation. Continue monitoring and iterating.

Confidence Levels

Level	Probes Run	Meaning
High	12+	Score is reliable — based on broad query coverage
Medium	8–11	Score is indicative — some queries may not have been tested
Low	<8	Score is directional only — limited probe data

Report Sections

Each Signal Engine report contains six executive-ready sections:

1. Executive Summary

2-3 sentence overview of AI visibility status and key finding.

2. AI Visibility Score

The 0-100 score with explanation of what's driving it up or down.

3. Evidence of AI Inclusion

Specific probe results showing where AI mentioned the brand or reused content phrases.

4. Content Structure Assessment

Analysis of how well the page is structured for AI consumption.

5. Competitive Signals

What AI systems are saying about the topic and which patterns they prefer.

6. Recommendations

3-5 specific, actionable items to improve AI visibility score.

Signal Engine API

POST /api/signal/analyze

Run a full signal analysis on a URL or content.

curl -X POST https://gen-visibility.scottfelten.com/api/signal/analyze \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://yourdomain.com/article",
    "brandName": "Your Brand",
    "rawText": "Optional: paste content directly"
  }'

Parameter	Type	Required	Description
`url`	string	Yes*	URL to analyze
`brandName`	string	No	Brand name to track in AI responses
`rawText`	string	Yes*	Raw content (alternative to URL)

* Either url or rawText is required.

GET /api/signal/reports

List all previous signal reports.

curl https://gen-visibility.scottfelten.com/api/signal/reports

GET /api/signal/report/:id

Retrieve a specific signal report with full probe data.

curl https://gen-visibility.scottfelten.com/api/signal/report/1

Full Workflow

The recommended workflow for maximizing AI visibility:

1

Baseline Measurement

Run Signal Engine on your existing content before optimization. Record the score.

2

Optimize with Gen Visibility

Use the Gen Visibility tool to generate an AI-optimized content layer. Append it to your article.

3

Publish & Wait

Publish the optimized content. Allow 1-7 days for AI systems to potentially index the changes.

4

Re-Measure

Run Signal Engine again on the same URL. Compare score to baseline.

5

Iterate

Use the report's recommendations to further optimize. Track scores over time to see trends.

Pro tip: Run Signal Engine monthly on your top-performing content. AI citation patterns change as models are updated — what works today may need adjustment in 3 months.

Cost Per Cycle

Step	Tool	Cost
Baseline measurement	Signal Engine	~$0.05–0.15
Content optimization	Gen Visibility	~$0.01–0.05
Post-optimization measurement	Signal Engine	~$0.05–0.15
Full cycle		~$0.11–0.35

At $0.25 average per cycle, you can optimize and measure 100 pages for about $25.

FAQ

Does this modify my original article?

No. Gen Visibility only generates supplementary content to append. Your original human-written content is never touched.

How is this different from traditional SEO?

Traditional SEO optimizes for search engine rankings. Generative Visibility optimizes for being cited as a source by AI systems like ChatGPT, Claude, and Perplexity. Different algorithms, different optimization strategies.

How quickly will I see results?

AI citation patterns change as models are updated. Results can appear within days for some queries, but tracking over weeks gives the most reliable data.

What's the best article length?

1,000–3,000 words gives the best results. Too short and there isn't enough context; too long and the pipeline costs increase without proportional benefit.

Can I use this for multiple articles?

Yes. Each analysis is independent and stored in your history. Run as many as you need — at $0.01–0.05 each, cost is minimal.

Does this work for any industry?

Yes. The pipeline is topic-agnostic. It works for travel, automotive, B2B, real estate, food, technology, healthcare, finance — any domain where people ask AI for recommendations.