Competitor Research竞品研究

Setup

On first use, read setup.md for integration guidelines.

When to Use

User needs deep competitor analysis. Agent conducts thorough research on competitors in a niche, identifies gaps and opportunities, and delivers actionable strategies. Supports both new market entry and existing business competitive analysis.

Architecture

Memory lives in ~/competitor-research/. See memory-template.md for structure.

CODEBLOCK0

Quick Reference

Topic	File
Setup process	INLINECODE3
Memory template

memory-template.md | | Research frameworks | frameworks.md |

Core Rules

1. Define Scope Before Research

Never start without clarity on:

Question	Why It Matters
What decision will this inform?	Shapes depth and focus
New market entry or existing competition?

Different analysis needs | | Direct competitors only, or substitutes too? | Defines research boundaries | | Time constraint? | Determines depth level |

If user is vague, ask. Bad scope = wasted research.

2. Use Depth Levels

Level	Time	Output	Best For
Quick Scan	15-30 min	Top 5 competitors, key differentiators, obvious gaps	Initial exploration
Standard

1-2 hours | Full landscape, pricing matrix, positioning map, opportunities | Business planning | | Deep Dive | Half day+ | Individual competitor audits, detailed SWOT, strategic playbook | Serious competition |

Always confirm depth level before starting. Default to Standard if unsure.

3. Structure Every Competitor Analysis

For each competitor, cover:

CODEBLOCK1

4. Always Find Gaps and Opportunities

End every research session with:

GAP ANALYSIS

• - What do customers complain about that nobody solves?
• What segments are underserved?
• What's overpriced in the market?
• What's missing that should exist?

OPPORTUNITIES

• - Where can user win? (price, features, positioning, audience)
• What would be the wedge to enter?
• What's the unfair advantage potential?

Research without actionable gaps is just a report. Make it strategic.

5. Iterate and Build Knowledge

Each research session builds on previous ones:

• - First session: Establish landscape, identify key players
• Follow-up sessions: Deep dive individual competitors
• Return visits: Update with new findings, track changes

Before researching a niche again, check niches/{niche}/ for prior work.

6. Cite and Date Everything

Mark all findings with:

• - Source: Where you found it (website, G2, LinkedIn, etc.)
• Date: When observed (pricing changes, features evolve)
• Confidence: High (direct source) / Medium (inferred) / Low (speculation)

Undated intelligence becomes unreliable fast.

7. Deliver Actionable Recommendations

Every research deliverable ends with:

CODEBLOCK2

Research Frameworks

Market Landscape

Start broad, then narrow:

1. 1. List all players (direct, indirect, substitutes)
2. Categorize by segment (enterprise, SMB, prosumer, etc.)
3. Map by positioning (premium vs budget, generalist vs niche)
4. Identify white space

Competitive Matrix

Compare on dimensions that matter:

Competitor	Price	Feature X	Feature Y	Target	Differentiator
Player A	$$$	✅	❌	Enterprise	Security
Player B

$ | ❌ | ✅ | SMB | Simplicity | | (User) | $$ | ✅ | ✅ | Mid-market | Best of both |

Win/Lose Analysis

For each competitor, answer:

• - Why would a customer choose them over user?
• Why would a customer choose user over them?
• What type of customer is a slam-dunk for each?

Positioning Audit

Analyze how competitors position:

• - Homepage headline and subhead
• Three main value props
• Social proof strategy
• Pricing presentation
• Comparison pages (if any)

Look for positioning gaps nobody owns.

Iterative Research Workflow

Session 1: Landscape
CODEBLOCK3

Session 2+: Deep Dives
CODEBLOCK4

Return Visit
CODEBLOCK5

Common Traps

• - No scope = bad research → Always clarify what decision this informs before starting
• Feature obsession → Business model and positioning often matter more than features
• Outdated pricing → Check pricing pages directly, don't trust cached data
• Missing substitutes → Direct competitors aren't the only threat. What else solves the same job?
• Analysis paralysis → Set time limits. Good-enough research beats perfect research never delivered
• No recommendations → A list of competitors isn't strategy. What should user DO with this?
• Forgot to save → Update memory and niche files after every session

Security & Privacy

Data that stays local:

• - All research stored in INLINECODE7
• Niche analyses and competitor profiles
• User preferences and context

This skill does NOT:

• - Access private competitor systems
• Create fake accounts for research
• Scrape content violating ToS
• Send your research externally
• Store any credentials

Related Skills

Install with clawhub install <slug> if user confirms:

• - market-research — broader market analysis
• INLINECODE10 — strategic frameworks
• INLINECODE11 — ongoing tracking after research

Feedback

• - If useful: INLINECODE12
• Stay updated: INLINECODE13