our-claude-skills/custom-skills/32-ourdigital-presentation/code/references/research_patterns.md

Research Patterns Guide

Overview

Common patterns for extracting and structuring research content from Notion.

Content Extraction Patterns

Hierarchical Extraction

Extract content respecting the document hierarchy:

1. Main Topics (H1 headers)
   - Key Points (H2 headers)
     - Supporting Details (H3 headers)
       - Data points
       - Examples

Data Point Extraction

Identify and extract quantitative information:

• Metrics: Numbers with units (e.g., "25% growth")
• KPIs: Key Performance Indicators
• Comparisons: Year-over-year, benchmarks
• Trends: Directional indicators

Action Item Detection

Look for action-oriented language:

• Keywords: "recommend", "suggest", "propose", "should"
• Imperatives: "implement", "deploy", "analyze"
• Future tense: "will", "plan to", "intend to"

Synthesis Patterns

Theme Clustering

Group related content by theme:

1. Technical themes: Technology, infrastructure, systems
2. Business themes: Strategy, market, competition
3. Operational themes: Process, efficiency, workflow
4. Customer themes: Satisfaction, feedback, needs

Priority Scoring

Rank content by importance:

• Critical: Urgent, high-impact, deadline-driven
• Important: Strategic, long-term value
• Relevant: Supporting information, context
• Optional: Nice-to-have, future consideration

Evidence Mapping

Connect claims to supporting evidence:

Claim → Data Point → Source
"Market is growing" → "25% YoY growth" → "Q4 Market Report"

Notion-Specific Patterns

Database Queries

When extracting from Notion databases:

1. Filter by date range for recent content
2. Sort by priority or importance fields
3. Group by category or project
4. Aggregate metrics across entries

Page Relationships

Follow page links strategically:

• Parent pages: For context and background
• Child pages: For detailed information
• Linked pages: For related topics
• Mentioned pages: For cross-references

Content Types

Handle different Notion content types:

• Text blocks: Extract as-is
• Toggle lists: Expand and include all content
• Tables: Convert to structured data
• Embeds: Note source and type
• Code blocks: Preserve formatting

Quality Checks

Completeness

Ensure extraction captures:

• All main sections
• Key data points
• Action items
• Recommendations
• Links and references

Accuracy

Verify extracted content:

• Numbers match source
• Quotes are exact
• Context is preserved
• Relationships are maintained

Relevance

Filter content by:

• Topic relevance
• Time relevance (recent vs outdated)
• Audience relevance
• Objective relevance

Advanced Patterns

Sentiment Analysis

Identify tone and sentiment:

• Positive indicators: Success, achievement, growth
• Negative indicators: Challenge, risk, decline
• Neutral indicators: Stable, maintained, unchanged

Temporal Analysis

Track changes over time:

• Historical context: Past performance
• Current state: Present situation
• Future projections: Plans and forecasts

Cross-Reference Analysis

Connect information across sources:

• Confirmatory: Multiple sources agree
• Contradictory: Sources conflict
• Complementary: Sources add different perspectives

Output Formatting

Executive Summary Pattern

1. High-level conclusion (1-2 sentences)
2. Key findings (3-5 bullets)
3. Critical metrics (top 3-5)
4. Immediate actions (top 3)

Detailed Report Pattern

1. Background and context
2. Methodology
3. Findings by category
4. Data analysis
5. Conclusions
6. Recommendations
7. Appendices

Quick Brief Pattern

1. What: Core message
2. Why: Importance/impact
3. How: Key data/evidence
4. When: Timeline/urgency
5. Who: Stakeholders/owners