How to Analyze and Compare Content for AI Retrieval

Optimising content for AI retrieval requires examining it from multiple angles — how it chunks for RAG, which topics it covers, how readable it is, and how it compares against competing pages. Doing this across separate tools is fragmented and slow. A unified workspace that combines chunking, topic analysis, readability, and side-by-side comparison enables faster iteration.

Content Lab in QueryBurst provides this multi-view workspace, accepting indexed pages or external URLs with optional competitor comparison.

What Is QueryBurst Content Lab?

Content Lab is a comprehensive suite of content analysis tools that helps you understand, compare, and optimize your content for AI visibility. It combines multiple analysis views into a single workspace, letting you examine your content from different angles.

Whether you're optimizing for search, improving topical coverage, or comparing against competitors, Content Lab provides the insights you need.

Getting Started

1. Navigate to Content Lab in the sidebar
2. Select a page from your indexed pages, or enter an external URL
3. Choose a view from the navigation bar at the top
4. Optionally, add a competitor URL to compare side-by-side

Views

Content Lab includes six specialized views:

Chunks View

Break your content into semantic chunks and score them against search queries.

Features:

• Multiple chunking strategies (hybrid, heading-based, fixed token)
• Semantic similarity scoring against your query
• BM25 keyword matching scores
• Title, meta description, and full-page relevance scores
• AI response generation (optional)
• Semantic triple extraction showing entity relationships
• Manual chunk control using <chunk> tags (see below)

Use cases:

• Test which parts of your content are most relevant to specific queries
• Compare your content's relevance against competitors
• Identify weak sections that need strengthening

Structure View

Visualize your document's heading hierarchy at a glance.

Features:

• Tree view of all headings (H1-H6)
• Proper nesting visualization
• Quick overview of content organization

Use cases:

• Audit heading structure for SEO best practices
• Ensure logical content flow
• Compare structure against competitor pages

Document View

Read your full content with powerful entity highlighting.

Features:

• Full rendered markdown document
• Entity highlighting by type (Person, Organization, Location, etc.)
• Multiple highlight modes:
  • All Entities - Highlight everything detected
  • Knowledge Graph - Only entities linked to Google's Knowledge Graph
  • Shared - Entities that appear in both your page and competitor
  • Entity List - Show entities in document order
• Toggle individual entity types on/off
• Works best after running Entity analysis

Use cases:

• See what AI models recognize as important entities
• Identify which entities you share with competitors
• Audit entity coverage across your content

Entities View

Deep entity extraction with gap analysis.

Features:

• Extract named entities using Google Cloud NLP
• Entity types: Person, Organization, Location, Event, Product, Work of Art, etc.
• Knowledge Graph linking (Wikipedia, MID identifiers)
• Salience scoring (importance to the content)
• Gap analysis when comparing:
  • Shared entities - Appear in both documents
  • Your unique entities - Only in your content
  • Their unique entities - Only in competitor (opportunities!)
• Filter by entity type, KG status, salience threshold

Use cases:

• Discover entity gaps versus competitors
• Verify Knowledge Graph coverage
• Identify missing topics to add

Coherence View

Analyze how topics and entities flow through your content.

Features:

• Section-by-section coherence analysis
• Heading-to-content alignment scoring
• Entity distribution across sections
• Document comparison metrics

Use cases:

• Check if headings match their content
• Identify sections that drift off-topic
• Ensure consistent topic coverage

Heat Map View

Visual topic coverage comparison.

Features:

• Hierarchical topic tree
• Color-coded coverage intensity
• Side-by-side comparison with competitor
• Drill down into sub-topics

Use cases:

• Quickly spot coverage gaps
• Compare topic depth against competitors
• Prioritize content improvements

Editing Your Document

Content Lab allows you to edit your page content directly. This is useful for:

• Testing how content changes affect relevance scores
• Experimenting with different phrasings
• Adding or removing sections to see the impact

How to edit:

1. Click the Edit button in the sidebar (pencil icon)
2. Make your changes in the editor
3. Click Save to apply changes
4. Run scoring or analysis to see the impact

Note: Edits are session-only and don't modify your actual website. Use them to experiment before making real changes.

Manual Chunk Tags

For precise control over how your content is chunked, use manual <chunk> tags in your content:

<chunk>
This is the first chunk. It contains information about topic A.
All content between the tags stays together.
</chunk>

<chunk>
This is the second chunk. It covers topic B.
You control exactly where breaks occur.
</chunk>

When to use manual chunks:

• Important sections that shouldn't be split
• Q&A pairs that need to stay together
• Tables or lists with context-dependent headers
• Any content where automatic chunking doesn't work well

How it works:

• Use the Hybrid chunking strategy to enable manual chunks
• Content inside <chunk> tags is kept intact
• Content outside tags uses automatic chunking
• Tags are stripped from the final output

Comparing Pages

Add a competitor URL to enable comparison mode:

1. Click Add Competitor in the sidebar
2. Enter the competitor's URL
3. Click Go to fetch their content
4. All views now show side-by-side comparison

In comparison mode:

• Chunks shows both pages scored against the same query
• Structure compares heading hierarchies
• Document can highlight shared entities
• Entities provides full gap analysis
• Coherence/Heat Map shows topic coverage differences

Tips & Best Practices

For Chunk Analysis

• Test multiple queries to understand content relevance
• Look for high-scoring chunks that don't match your intent (may confuse AI)
• Low-scoring important sections need strengthening

For Entity Analysis

• Run entity analysis before using Document view highlighting
• Focus on "Their unique entities" to find content gaps
• High-salience competitor entities are priority additions

For Coherence

• Aim for 70%+ alignment between headings and content
• Sections with low coherence may need rewriting
• Compare against competitors to set benchmarks

For Structure

• Ensure single H1 per page
• Maintain proper heading hierarchy (no skipping levels)
• Use descriptive headings that match content

Keyboard Shortcuts