Competitive Intelligence Dossier

This recipe walks through building a competitive intelligence dossier — a structured analysis of a competitor's online presence, strategy, and positioning. It uses the discover, fan-out, synthesize pattern that generalizes to any multi-step research workflow.

The Pattern

DISCOVER  → Find all relevant sources and appearances
FAN-OUT   → Deep-dive into each source in parallel
SYNTHESIZE → Roll up findings into a single dossier

This pattern works because it breaks an overwhelming research task into manageable, parallelizable steps.

Step 1: Discovery

Set Up the Research Task

Create a task group called "Competitive Intel" to organize the work:

@Assistant create a task group called "Competitive Intel"

Then create the discovery task:

@Researcher create a task titled "Discover all online appearances for Acme Corp"
and start working on it. Search for:
- Company website and blog
- Press mentions and news articles
- Social media presence
- Podcast appearances
- Job postings (signal of growth areas)
- Product review sites (G2, Capterra, etc.)
- GitHub/open-source presence

Write a document titled "Acme Corp - Discovery" listing every 
appearance with URL, date, platform, and a one-line summary.

The Researcher agent will use web search, news search, and URL fetching to compile a comprehensive list.

What You Get

A structured document listing 15-30+ appearances organized by platform type. Each entry includes the source URL, date found, and a brief description of the content.

Step 2: Fan-Out (Deep Dives)

For each significant appearance, create a sub-task for detailed analysis:

@Researcher for each entry in the "Acme Corp - Discovery" document,
spawn a subtask to analyze that source. For each one:
- Visit the URL and extract key information
- Note the audience and reach
- Identify key quotes or claims
- Flag any competitive positioning against us
- Rate relevance (high/medium/low)

Update the discovery document with your findings.

The agent will create multiple sub-tasks and work through them, updating the document as it goes. Tasks in the "waiting" state indicate work in progress.

Parallel Execution

If you have multiple agents, you can assign different sub-tasks to different agents for faster completion:

@Researcher handle the press and news sub-tasks
@Analyst handle the product review and comparison sub-tasks

Step 3: Synthesis

Once the deep-dives are complete, synthesize everything into a final dossier:

@Writer create a document titled "Acme Corp - Competitive Dossier" 
based on the discovery document and all deep-dive findings. Include:

1. Executive Summary (3 sentences)
2. Company Overview (size, funding, market position)
3. Product Analysis (features, pricing, positioning)
4. Online Presence Assessment (reach, engagement, sentiment)
5. Competitive Threats (where they beat us)
6. Competitive Advantages (where we beat them)
7. Key Quotes and Claims (with sources)
8. Recommended Actions (what should we do?)

Storing Results

The final dossier lives as a document in your workspace, searchable and editable. You can:

• Link it from relevant task groups or channels
• Update it periodically by re-running the discovery step
• Share it with stakeholders by posting in a channel
• Build a database of competitor profiles for cross-company comparison

Variations

Person Research

Same pattern, but for researching a person (potential hire, speaker, partner):
- Discovery: LinkedIn, social media, publications, speaking engagements
- Fan-out: Detailed analysis of each appearance
- Synthesis: Professional profile with recommendations

Market Landscape

Research an entire market segment:
- Discovery: Identify all players in a space
- Fan-out: Profile each company
- Synthesis: Market map with positioning, gaps, and opportunities

Due Diligence

For M&A or partnership evaluation:
- Discovery: Financial data, news, legal filings, employee reviews
- Fan-out: Detailed analysis of each data source
- Synthesis: Risk assessment with recommendations