| name | market-research |
| description | 统一的市场与用户研究入口:市场面收集市场趋势、行业报告并识别机会(TAM-SAM-SOM);用户面进行用户访谈、调研与画像创建。当用户需要市场研究、行业分析、趋势识别、市场规模分析,或需要了解用户需求、创建用户画像、进行用户访谈,或说"X 市场发生了什么"、"用户想要什么"、"用户研究"时使用。即使没有明确说"市场研究"或"用户研究",当用户正在研究某个市场或行业,或试图理解用户行为、需求与痛点时也应激活。 Also triggers on: market research, what is happening in the X market, market sizing, what do users want, user research. |
| layer | perception |
| input-from | user |
| output-to | product-strategy,competitive-analysis,prd-gen |
Market Research
统一研究入口:市场面看清赛道,用户面看懂用户。
What This Skill Does
按研究交付物分为两种模式:
- 市场面 (Market Intelligence):搜索驱动的市场趋势、市场规模(TAM-SAM-SOM)、关键玩家、机会与威胁分析,交付
market-analysis.json + Markdown 报告
- 用户面 (User Research):访谈提纲、调研设计、用户画像与洞察,交付
user-research.json + Markdown 报告
两种模式可独立运行,也可组合执行(先市场面定赛道,再用户面定人群)。
When to Use
市场面 — activate when:
- User asks about market trends or industry developments
- Phrases like "market research", "industry analysis", "market size", "X 市场发生了什么"
- Exploring new markets or product categories
- Need to understand market dynamics before product decisions
用户面 — activate when:
- User asks about understanding user needs or behaviors
- Phrases like "user research", "talk to users", "用户想要什么", "create personas"
- Creating or updating user personas
- Analyzing user feedback or interview data
- Need to validate product ideas with real users
How It Works
- Determine mode — 根据用户意图选择市场面、用户面或两者组合;意图不明时询问用户
- 市场面: Define market scope → Search industry sources → Analyze trends → Assess market size (TAM/SAM/SOM) → Identify opportunities → Generate insights
- 用户面: Define research goals → Select methods → Create instruments (interview guides, surveys) → Gather data → Analyze findings → Create artifacts (personas, journey maps, insights)
Research Modes
市场面 (Market Intelligence)
搜索驱动:所有市场数据必须来自实际搜索,禁止仅凭训练记忆输出(见 Anti-Hallucination Rules)。
分析深度:
| Depth | When to Use | What You Get |
|---|
overview | Quick market check | Market size, 3 trends, 5 key players |
standard | Regular analysis (default) | Full sizing, trends, opportunities, threats, recommendations |
deep | Strategic decisions | Segmentation analysis, buying patterns, vendor landscape, 10+ recommendations |
用户面 (User Research)
研究方法:
| Method | When to Use | Output |
|---|
interviews | Deep understanding of needs | Interview transcripts, key quotes |
surveys | Broad pattern validation | Survey results, statistical insights |
personas | User archetype creation | Persona profiles |
journey-mapping | Experience optimization | User journey maps |
【未验证】规则(关键约束):没有真实用户数据(真实访谈记录、调研结果或用户提供的一手资料)时,不得把推断包装成结论:
- 所有基于推断的画像和结论必须标注【未验证】
- 同时注明所需验证方式(如"需 5+ 目标用户访谈验证"、"需 N≥100 问卷验证")
- 推断画像可作为假设交付,但必须与真实数据得出的结论明确区分
引用行业报告、统计数据时,同样适用下方 Anti-Hallucination Rules。
Persona Template
### [Name](verified / 【未验证】)
**Role**: [Job title] at [Company type]
**Goals**:
- [Primary goal]
- [Secondary goal]
**Pain Points**:
- [Key frustration 1]
- [Key frustration 2]
**Behaviors**:
- [Observable behavior 1]
- [Observable behavior 2]
**Quote**: *"[Direct user quote]"*(无真实引用时删除此行,禁止虚构)
**Demographics**:
- Company size: [X-Y employees]
- Tech stack: [Key tools]
- Experience level: [Years]
**Verification**: [已验证来源 / 所需验证方式]
Interview Guide Template
## Interview Questions
### Warm-up
1. Tell me about your role
2. What does a typical day look like?
### Discovery
3. How do you currently [solve problem X]?
4. What tools do you use?
5. What works well? What doesn't?
### Deep Dive
6. Walk me through the last time you [encountered problem]
7. What was frustrating about that?
8. What did you try instead?
### Closing
9. If you could change one thing, what would it be?
10. Is there anything else I should know?
Anti-Hallucination Rules (Self-Contained)
The following rules are mandatory for this skill. They are inlined here for standalone installation compatibility.
1. Mandatory Search First
输出任何市场数据前必须先使用 WebSearch/WebReader。
- 禁止仅凭训练记忆输出任何具体数据(市场规模、增长率、参与者等)
- 每个数据点必须来自实际搜索和访问的页面
- 搜索查询应包含年份,例如
SaaS project management market size 2025 2026
2. Every Claim Must Have a Source
每个数字、趋势、参与者必须有 URL 来源。
- 没有来源 = 不得写入
- 来源格式:
{ "claim": "...", "source": "https://...", "source_name": "...", "fetched_at": "..." }
3. Unknown is Acceptable
找不到数据时标注 "Unknown",禁止编造。标注未知是诚实,不是失败。
4. Confidence Rating
每个关键数据点标注 confidence (high/medium/low)。
high: 权威来源(Gartner, Forrester, IDC, 官方财报)
medium: 可靠来源(TechCrunch, Statista, 行业报告)
low: 社区讨论、博客、推测性分析
5. Distinguish Fact from Inference
明确区分已验证事实和推断。推断必须标注 basis 和 confidence。
6. Quality Gate (Non-negotiable)
以下任一条未满足,不得输出结果:
Search Record
每个分析输出必须附带搜索记录:{ "search_queries_used": [...], "sources_accessed": [{ "url": "...", "title": "...", "used_for": "..." }] }
Input Parameters
| Parameter | Type | Required | Description |
|---|
mode | string | No | market(市场面)、user(用户面)、both(默认按意图判断) |
market | string | 市场面必填 | Market or industry to analyze (e.g., "SaaS project management") |
focus_areas | list | No | Specific areas to focus on (e.g., ["pricing", "enterprise segment"]) |
geography | string | No | Geographic scope (default: global) |
depth | string | No | overview (quick), standard (default), or deep (comprehensive) |
research_goal | string | 用户面必填 | What you want to learn about users |
target_segment | string | No | Specific user segment to study |
method | string | No | interviews, surveys, personas, all (default) |
sample_size | number | No | Target number of participants |
Evidence Discipline (证据纪律)
- 缺失信息标记【待确认】:任何无法从输入或上下文获得的信息,标记为【待确认】,不得脑补填默认值。
- 证据追溯【未验证】:用户需求/市场断言必须追问证据来源;无证据的标记为【未验证】并注明所需验证方式。
- 完整度检查门:正式输出交付物前,先输出信息完整度清单(已确认 ✅ / 待确认 ⚠️ / 未验证 ❓),经用户确认后才生成正文。
- 可追溯:正式输出中的关键结论必须能追溯到以下来源之一:已确认的用户输入、引用的证据(含 URL)、或用户的显式决策;无法追溯的结论视为脑补,删除或改标【待确认】。
Output Structure
每种模式生成两个输出;JSON 文件名为下游契约(contracts/ 与其他 skill 按此引用),不得改名:
- 市场面 →
docs/product/.ompm/market-analysis.json(结构化数据)+ Markdown 报告
- 用户面 →
docs/product/.ompm/user-research.json(结构化数据)+ Markdown 报告
market-analysis.json Format
{
"analysis": {
"id": "uuid",
"timestamp": "2026-03-12T...",
"market": "SaaS project management",
"market_size": {
"tam": { "value": "$10B", "confidence": "high", "source": { "url": "https://...", "source_name": "Gartner", "fetched_at": "2026-03-12T..." } },
"sam": { "value": "$3B", "confidence": "high", "source"
市场面 Markdown Report Structure
# Market Intelligence Report
## Market Overview
- **Market**: SaaS Project Management
- **Date**: YYYY-MM-DD
- **Geography**: Global
## Market Size
| Segment | Size | Confidence | Source |
|:--------|-----:|:-----------|:-------|
| TAM | $10B | high | Gartner, 2025 |
| SAM | $3B | high | IDC, 2025 |
| SOM | $300M | medium | Internal estimate |
## Key Trends
1. **AI-Powered Automation** - Teams are demanding AI assistance for task management...
2. **Remote Work Collaboration** - Post-pandemic shift continues driving demand...
3. **Industry-Specific Solutions** - Generic tools losing ground to vertical products...
## Key Players
| Company | Market Share | Strength | Weakness |
|:--------|-------------:|:---------|:---------|
| Asana | 20% | Enterprise features | Expensive |
| Monday.com | 18% | UX simplicity | Limited customization |
## Opportunities
- **Small Business Segment**: Underserved by enterprise-focused tools
- **Integration Ecosystem**: Gaps in connecting with specialized tools
## Threats
- Market saturation in mid-market segment
- Platform vendors (Microsoft, Google) adding features
## Strategic Recommendations
1. Focus on small business segment for growth
2. Build deep integrations with developer tools
3. Consider vertical specialization (e.g., for agencies)
user-research.json Format
{
"research": {
"id": "uuid",
"timestamp": "2026-03-12T...",
"goal": "Understand onboarding pain points",
"method": "interviews",
"participants": 8,
"personas": [
{
"name": "Product Manager Alice",
"role": "PM at B2B SaaS",
"goals": ["Ship features faster", "Align team"],
"pain_points": ["Too many tools", "Unclear priorities"],
"behaviors": ["Checks Jira first thing"
用户面 Markdown Report Structure
# User Research Report
## Research Overview
- **Goal**: Understand onboarding pain points
- **Method**: User Interviews
- **Participants**: 8 users
- **Date**: YYYY-MM-DD
## User Personas
### Product Manager Alice(verified)
**Role**: PM at B2B SaaS company (50-200 employees)
**Goals**:
- Ship features faster without breaking things
- Keep team aligned on priorities
**Pain Points**:
- Too many tools to check (Jira, Slack, GitHub, Notion)
- Unclear priorities lead to context-switching
**Behaviors**:
- Checks Jira first thing every morning
- Prefers async communication over meetings
**Quote**: *"I just wish everything was in one place instead of checking 5 different apps."*
**Verification**: 8 场真实访谈交叉验证
### Hypothetical Startup PM【未验证】
**Role**: PM at early-stage startup
**Pain Points**:
- [推断项] 资源有限,一人多角
**Verification**: 基于市场面用户细分推断,需 5+ 目标用户访谈验证
## Key Insights
### Insight 1: Context-Switching Friction
**Finding**: Users spend 40% of time switching between tools
**Evidence**: 6/8 participants mentioned this as top frustration
**Opportunity**: Unified workspace could save hours daily
## Recommendations
1. **Short-term**: Add status page showing all work in one view
2. **Medium-term**: Build integrations with Jira, GitHub, Slack
: Create AI-powered priority recommendations
[ ] 验证【未验证】画像:完成 5+ 目标用户访谈
[ ] Run A/B test on simplified onboarding
Quality Standards
市场面:
- Cover market size (TAM/SAM/SOM) — use "Unknown" if not found, do NOT fabricate
- Identify real trends found through search (no minimum count required)
- List major players found in search results
- Provide actionable opportunities based on real data
- Note significant threats
用户面:
- Include 3+ personas (if applicable) — base on real data, do NOT fabricate
- 无真实用户数据时,所有推断画像/结论已标注【未验证】并注明所需验证方式
- Provide real insights with evidence counts from actual research
- Include direct user quotes with source references(无真实引用则不引用,禁止虚构)
- Link insights to actionable recommendations
通用:
- ALL numeric claims, market sizes, growth rates, industry benchmarks have source URLs
- ALL data points have confidence rating (high/medium/low)
- Facts and inferences are clearly distinguished
- Search record is included proving searches were executed
- 完整度检查门已通过(Evidence Discipline 第 3 条:已确认 ✅ / 待确认 ⚠️ / 未验证 ❓ 清单经用户确认)
- Be valid JSON for downstream skills
Context Integration
Reads:
docs/product/.ompm/market-analysis.json — 用户面复用市场上下文(组合执行时)
Writes:
docs/product/.ompm/market-analysis.json — 市场分析结果(契约文件名,供下游 skill 使用)
docs/product/.ompm/user-research.json — 用户研究结果(契约文件名,供下游 skill 使用)
Read By:
product-strategy — Uses market data and personas for positioning strategy
competitive-analysis — References key players for comparison
prd-gen — References user needs and market opportunities in requirements
Example Usage
User: "What's the market for AI writing assistants?"
→ 市场面:搜索驱动分析该市场,输出 market-analysis.json
User: "Research the project management software market"
→ 市场面,focus on SaaS PM tools
User: "Is the fitness app market growing?"
→ 市场面,depth=overview
User: "Create personas for our product"
→ 用户面,method=personas;无真实数据时画像标注【未验证】并注明验证方式
User: "Analyze these user interviews"
→ 用户面,分析用户提供的访谈数据(真实数据,可直接得出结论)
User: "先研究下市场规模,再帮我做用户画像"
→ 组合执行:市场面 → 用户面
Execution Profile (执行建议)
- 模型建议:轻量(haiku 级)——宿主支持多模型/子代理时按此分配;单模型宿主忽略
- 工具姿态:读写 + 联网搜索——遵循最小权限
- 记忆文件:
docs/product/.ompm/memory/market-research.md——跨会话积累经验(优质信源、领域基线、分析框架);执行开始时读取、结束时更新;文件不存在则新建
- Claude Code 增强:本 profile 在该宿主由 market-researcher subagent 以隔离上下文实现
下一步
本 skill 完成后,如果用户没有明确下一步,引导用户使用 ompm skill 做意图路由——它会读取本轮产出和当前状态,判断最有价值的下一步。不要替用户预设固定长链;output-to 声明的是数据流向,不是强制路径。