关于 persona-generator
persona-generator 是用于客户研究、营销和产品决策的技能,依据访谈证据结合 Buyer Personas 与 Jobs-to-be-Done 构建买家画像,关注购买行为而非虚构的人口特征。
主要功能
- 01基于研究生成买家画像
- 02识别购买触发因素
- 03梳理决策标准与顾虑
- 04描述买家决策旅程
- 05生成可执行的客户细分
适用场景
- 01整理客户访谈并生成画像
- 02为客户发现研究制定待验证画像
- 03评估并改进现有画像
- 04规划营销内容与产品路线图
安装
复制下面的命令,在终端中运行。
npx skills add https:/ / github.com/ guia-matthieu/ clawfu-skills --skill persona-generator安装后在 Claude Code 中输入/persona-generator确认已加载。
详细介绍
Persona 生成器
创建有研究依据的买家角色画像,推动切实的营销和产品决策。结合买家角色画像方法与待完成任务(Jobs-to-be-Done),基于实际行为而非虚构的人口统计信息构建档案。
何时使用此技能
- 开始客户探索,确定要向哪些人验证想法
- 规划营销活动,向合适的人传递合适的信息
- 制定内容策略,创作能引起特定受众共鸣的内容
- 确定产品路线图的优先级,为真实用户构建功能
- 赋能销售团队,帮助销售人员了解交谈对象
- 统一团队认知,让每个人都使用相同的客户语言
方法论基础
| 方面 | 详情 |
|---|---|
| 来源 | Adele Revella - 《Buyer Personas》(2015)+ Clayton Christensen - 待完成任务(Jobs-to-be-Done) |
| 核心原则 | “基于真实研究构建的买家角色画像,能够揭示购买决策背后的思维过程——不仅是人口统计信息,还有动机、焦虑和决策标准。” |
| 重要原因 | 大多数角色画像都是虚构的人口统计信息(“营销玛丽,35 岁,喜欢瑜伽”)。有用的角色画像会解释某人为何购买——他们的焦虑、触发事件和决策过程。 |
Claude 做什么,您决定什么
| Claude 做什么 | 您决定什么 |
|---|---|
| 组织制作工作流 | 最终创意方向 |
| 提出技术方案 | 设备和工具选择 |
| 创建模板和检查清单 | 质量标准 |
| 识别最佳实践 | 品牌和语气决策 |
| 生成脚本大纲 | 最终脚本审批 |
此技能的功能
- 生成基于研究的角色画像 - 依据行为而非人口统计信息构建
- 识别购买触发因素 - 哪些事件会促使某人寻求解决方案
- 梳理决策标准 - 哪些因素会影响购买决策
- 揭示焦虑和障碍 - 哪些因素会阻止他们购买
- 记录买家旅程 - 他们如何调研并做出决策
- 创建可执行的细分群体 - 能推动实际决策的角色画像
使用方法
根据客户访谈生成角色画像
I've completed [X] customer interviews. Here are my notes: [summary]
Generate buyer personas using the Buyer Personas + JTBD methodology.
Focus on buying triggers, decision criteria, and anxieties.
创建假设角色画像(调研前)
I'm building [product] for [market].
Generate hypothesis personas I should validate through customer discovery.
Include questions to ask to validate each persona.
分析并改进现有角色画像
Here are our current personas: [paste personas]
Analyze them against best practices.
What's missing? What questions should we research to make them useful?
说明
生成角色画像时,请遵循以下循证方法:
第 1 步:了解角色画像为何有用
## The Persona Problem
### BAD Personas (Demographic Fiction)
"Marketing Mary"
- Age: 35
- Income: $80K
- Lives in suburbs
- Likes yoga and organic food
- Uses Instagram
**Why this fails:**
- Describes demographics, not motivations
- Doesn't explain why she would buy
- Can't drive marketing/product decisions
- Could describe millions of people
### GOOD Personas (Behavioral/JTBD-Based)
"The Overwhelmed First-Timer"
- **Trigger:** Just got promoted to manager, now responsible for [task]
- **Job-to-be-Done:** Make me look competent to my boss
- **Current Behavior:** Using spreadsheets, asking colleagues, stressed
- **Decision Criteria:** Easy to learn, makes me look good, won't fail publicly
- **Anxiety:** "What if I choose wrong and look incompetent?"
**Why this works:**
- Explains what triggered the search
- Reveals what they're really hiring the product to do
- Shows how they decide
- Surfaces what might stop them
- Drives specific marketing and product decisions
第 2 步:购买洞察的五个方面
## Adele Revella's Five Rings
### Ring 1: PRIORITY INITIATIVE
**What event triggered their search?**
Questions to research:
- What was the trigger event?
- Why now vs. 6 months ago?
- What finally made this urgent?
- What was the breaking point?
Example insight:
"They start looking when they get a new boss who asks 'why don't we have...' or when a competitor does something that makes them look behind."
---
### Ring 2: SUCCESS FACTORS
**What outcome are they trying to achieve?**
Questions to research:
- What does success look like?
- How will they measure results?
- What would make them a hero internally?
- What would make them regret the decision?
Example insight:
"They don't actually want [product feature]. They want to be seen as innovative by their CEO while not risking a high-profile failure."
---
### Ring 3: PERCEIVED BARRIERS
**What could stop them from buying?**
Questions to research:
- What concerns came up during evaluation?
- What almost made them walk away?
- What would cause them to delay?
- What do they fear will go wrong?
Example insight:
"Their biggest fear isn't that it won't work—it's that their team won't use it and they'll have wasted budget on something that sits unused."
---
### Ring 4: DECISION CRITERIA
**How do they evaluate options?**
Questions to research:
- What features/capabilities are must-haves?
- How do they compare options?
- Who else influences the decision?
- What trade-offs are they willing to make?
Example insight:
"They create a spreadsheet comparing 3-5 options. Integration with existing stack is #1. If it doesn't connect to Salesforce, they won't consider it."
---
### Ring 5: BUYER'S JOURNEY
**How do they research and decide?**
Questions to research:
- How did they first learn about the category?
- What resources did they use to research?
- Who did they talk to?
- What was the decision timeline?
Example insight:
"They google '[category] vs [category]' then ask for recommendations in a Slack community. Reviews on G2 are the final checkpoint before talking to sales."
第 3 步:加入待完成任务层
## JTBD Analysis Per Persona
### The Job Statement
"When I [situation], I want to [motivation], so I can [expected outcome]."
### Three Types of Jobs
**Functional Job:** What they need to accomplish
- "I need to get this report done faster"
- "I need to track our campaigns across channels"
**Emotional Job:** How they want to feel
- "I want to feel confident presenting to the board"
- "I don't want to worry about this anymore"
**Social Job:** How they want to be perceived
- "I want my team to see me as innovative"
- "I don't want to look like I made a bad decision"
### Forces That Drive Switching
**Push (away from current):**
- Pain points with current solution
- Frustrations with status quo
- External pressures (boss, market, competition)
**Pull (toward new):**
- Desired outcomes
- Attractive features
- Vision of better future
**Anxiety (about switching):**
- Fear of failure
- Implementation concerns
- Uncertainty about claims
**Habit (keeping current):**
- Familiarity with status quo
- Sunk costs
- "Good enough" mentality
For a persona to buy: Push + Pull > Anxiety + Habit
第 4 步:角色画像模板
## PERSONA: [Name Based on Behavior, Not Demographics]
### Snapshot
| Aspect | Detail |
|--------|--------|
| **Role** | [Job title / responsibility] |
| **Environment** | [Company size, industry, team structure] |
| **Primary JTBD** | [Job statement] |
| **Current Solution** | [What they use today] |
---
### Ring 1: Priority Initiative (Trigger)
**What triggers the search:**
- [Specific event 1]
- [Specific event 2]
- [Specific event 3]
**Why now (urgency drivers):**
- [What makes them act now vs. later]
**Quotes:**
> "[Verbatim from research]"
---
### Ring 2: Success Factors
**Desired outcomes:**
1. [Functional outcome]
2. [Emotional outcome]
3. [Social outcome]
**How they'll measure success:**
- [Metric or indicator]
**What makes them a hero:**
- [Internal win they're seeking]
**Quotes:**
> "[Verbatim from research]"
---
### Ring 3: Perceived Barriers
**Concerns/anxieties:**
1. [Concern about product/vendor]
2. [Concern about implementation]
3. [Concern about results]
**What almost stops them:**
- [Top barrier]
**Risk they fear most:**
- [What failure looks like]
**Quotes:**
> "[Verbatim from research]"
---
### Ring 4: Decision Criteria
**Must-haves (non-negotiable):**
1. [Feature/capability]
2. [Feature/capability]
3. [Feature/capability]
**Nice-to-haves:**
- [Feature/capability]
**Deal-breakers:**
- [What would eliminate you]
**How they compare:**
- [Their evaluation process]
**Quotes:**
> "[Verbatim from research]"
---
### Ring 5: Buyer's Journey
**Stage 1 - Trigger:** [What happens]
**Stage 2 - Research:** [Where they look]
**Stage 3 - Evaluate:** [How they compare]
**Stage 4 - Decide:** [Who's involved]
**Stage 5 - Buy:** [Process/procurement]
**Information sources:**
- [Channel 1]
- [Channel 2]
- [Influencers]
**Timeline:** [Typical decision timeframe]
---
### Jobs-to-be-Done
**Primary Job:**
"When I [situation], I want to [motivation], so I can [outcome]."
**Functional Job:** [What to accomplish]
**Emotional Job:** [How to feel]
**Social Job:** [How to be seen]
**Forces Analysis:**
| Force | Strength | Specifics |
|-------|----------|-----------|
| Push | High/Med/Low | [Details] |
| Pull | High/Med/Low | [Details] |
| Anxiety | High/Med/Low | [Details] |
| Habit | High/Med/Low | [Details] |
---
### Marketing Implications
**Messaging that resonates:**
- [Message/hook]
- [Message/hook]
**Content they'd engage with:**
- [Content type]
- [Content type]
**Where to reach them:**
- [Channel]
- [Channel]
**What NOT to say:**
- [Anti-message]
---
### Sales Implications
**Objections to expect:**
- [Objection]
- [Objection]
**Questions they'll ask:**
- [Question]
- [Question]
**Proof points they need:**
- [Case study/demo]
**Decision process:**
- [How deals close with this persona]
第 5 步:验证并更新
## Persona Validation Checklist
### Is this persona based on evidence?
- [ ] Minimum 5 interviews with this type
- [ ] Includes verbatim quotes
- [ ] Describes behavior, not just demographics
- [ ] Explains WHY they buy
### Does this persona enable decisions?
- [ ] Marketing team can use it to target campaigns
- [ ] Product team can use it to prioritize features
- [ ] Sales team can use it to tailor conversations
- [ ] Content team can use it to create relevant assets
### Is this persona distinct?
- [ ] Clearly different from other personas
- [ ] Different triggers
- [ ] Different decision criteria
- [ ] Different objections/anxieties
### Red Flags (Persona Needs Work)
- [ ] Based on assumptions, not research
- [ ] All demographics, no behaviors
- [ ] No quotes from real customers
- [ ] Too vague to drive decisions
- [ ] Overlaps significantly with another persona
示例
示例 1:B2B SaaS 角色画像
输入:
“为我们的营销分析仪表盘生成角色画像。目标对象:员工人数为 100–500 人的 B2B 公司中的营销经理。价格:每月 500 美元。”
输出:
角色画像 1:证明投资回报率的经理
概览
| 方面 | 详情 |
|---|---|
| 职位 | 营销经理,向 VP/CMO 汇报 |
| 工作环境 | B2B SaaS 公司,员工人数为 100–300 人,营销团队规模较小(3–5 人) |
| 主要 JTBD | 证明营销的投资回报率,以争取预算和人手 |
| 当前解决方案 | 电子表格 + GA + Hubspot 报告(手动整理) |
方面 1:优先事项(触发因素)
什么会触发搜索:
- CFO/CEO 问“营销支出的投资回报率是多少?”
- 即将进行预算审核,需要为数据辩护
- 新任 CMO 上任,希望了解业绩情况
- 竞争对手的营销似乎更有效,董事会开始提问
为什么是现在:
- 即将进入年度规划周期
- 刚刚未能达成目标,需要解释原因
- 面临越来越大的压力,必须做到“数据驱动”
引述:
“我的 CFO 上周问我,各渠道的 CAC 是多少。我花了一整天整理电子表格,才给出一个粗略估算。”
Ring 2:成功因素
期望成果:
- 一个展示营销投资回报率的仪表板(功能层面)
- 有信心向领导层汇报(情感层面)
- 被视为以数据为依据、具有战略眼光的人(社交层面)
衡量成功的方式:
- 节省的报告时间
- 能否立即回答 CFO 的问题
- 更清晰的归因,以便为支出辩护
成就感来源:
- “这个仪表板证明我们应该将 Google Ads 预算翻倍。”
Ring 3:感知障碍
顾虑:
- 我的团队真的会用这个吗?(采用情况)
- 数据准确吗?(信任)
- 实施需要多长时间?(时间)
几乎让他们放弃的因素:
- 与现有技术栈集成过于复杂
- 担心看到他们不想看到的数据
他们担心的风险:
- “我买了这个,结果数据是错的,现在我还给了 CFO 错误的数据。”
引述:
“我们以前试过一个仪表板,但数字总是和 Hubspot 对不上。简直是一场灾难。”
Ring 4:决策标准
必备条件:
- 集成 Hubspot(或他们的 CRM)
- 连接 Google Ads 和 LinkedIn
- 归因建模(而不只是末次点击)
加分项:
- 自动生成报告
- 与行业基准对比
无法接受的条件:
- 需要手动录入数据
- 不支持 Hubspot 集成
- 实施时间超过 2 周
对比方式:
- 创建对比电子表格
- 查看 G2 评论
- 需要来自类似公司的用户口碑
Ring 5:买家的旅程
- 触发: CFO 提出尖锐问题 → 意识到有此需求
- 调研: 在 Google 搜索“营销归因软件”,并在 Slack 上咨询同行
- 评估: 将 3 个选项列入候选名单,逐一查看演示,检查集成情况
- 决策: 向老板推荐,制定商业论证,申请预算批准
- 购买: IT 安全审查、采购流程、签署合同
时间线: 从触发到购买需要 4-8 周
营销启示
引人共鸣的信息:
- “实时回答 CFO 的问题”
- “无需电子表格,证明营销投资回报率”
- “值得信赖的归因”
触达渠道:
- LinkedIn(定向触达营销经理)
- RevOps/营销 Slack 社区
- G2/Capterra(销售漏斗底部)
不要说的话:
- “虚荣指标”(他们关注的是收入,而不是展示次数)
- “AI 驱动”(他们对炒作持怀疑态度)
Persona 2:意外成为分析师
概览
| 方面 | 详情 |
|---|---|
| 职位 | 营销协调员或专员,职业生涯初期 |
| 工作环境 | B2B,150-400 名员工,向营销经理汇报 |
| 主要待办任务(JTBD) | 不用整周埋头做电子表格,也能把数据交给老板 |
| 当前解决方案 | 从 5 个工具导出数据,在 Excel 中合并,在 Slides 中排版 |
Ring 1:优先事项
促使他们开始寻找的因素:
- 每周报告需要 4 小时以上才能完成
- 老板要求提供他们无法轻松整理的新数据切分
- 上一份报告出错,受到批评
引述:
“每到周一,我就害怕做报告。我半天都在 Excel 里忙着把数字凑到一起。”
Ring 2:成功因素
期望成果:
- 报告耗时从 4 小时缩短到 30 分钟(功能层面)
- 不再害怕周一(情感层面)
- 在经理眼中显得高效且有能力(社交层面)
Ring 3:感知障碍
顾虑:
- “我不懂技术——我能设置好吗?”
- “这在我的预算权限范围内吗?”(需要获得批准)
- “如果数字和我一直在报告的数据对不上怎么办?”
Ring 4:决策标准
必备条件:
- 无代码设置
- 预制报告模板
- 易于学习(不复杂)
对比方式:
- 在 YouTube 上搜索教程
- 在评论中寻找“易于使用”的评价
- 必须提供免费试用
营销启示
宣传信息:
- “5 分钟完成每周报告”
- “无需 Excel”
- “简单到一个下午就能设置好”
触达渠道:
- YouTube 教程
- “适合初学者的最佳 [category]”内容
- 营销职业通讯
他们能影响的人:
- 他们不会购买,但会向经理(Persona 1)推荐
示例 2:消费产品用户画像
输入:
“我们即将推出高端备餐服务(150 美元/周)。请基于这 8 次客户访谈创建用户画像。”
访谈摘要:
- 4 位在职专业人士表示“没时间做饭”
- 2 位新手父母表示“宝宝出生后一直在吃垃圾食品”
- 2 位注重健身的人表示“需要宏量营养素比例精准的餐食”
输出:
用户画像 1:精疲力竭的专业人士
概览
| 方面 | 详情 |
|---|---|
| 角色 | 专业人士,28-40 岁,工作要求高(咨询、金融、科技) |
| 环境 | 城市生活、工作时间长、收入高、个人时间有限 |
| 主要待完成任务(JTBD) | 在不影响职业表现的情况下健康饮食 |
| 当前解决方案 | 外卖,偶尔订购 HelloFresh(但不会自己做) |
第一层:优先事项
触发因素:
- 医生说胆固醇偏高
- 开始新工作后体重增加了 10 磅
- 意识到自己连续 5 天都在吃外卖
- 朋友看起来状态很好,问他们最近做了什么
为什么是现在:
- 健康警讯或重要生日
- 新年计划
- 伴侣对他们的饮食发表了看法
引述:
“我知道自己应该吃得更健康。我试过周日做饭,但周末时我已经累得不想动了。”
第二层:成功因素
期望的结果:
- 毫不费力地吃上健康餐食(功能层面)
- 不再因自己的饮食而感到内疚(情感层面)
- 看起来像是把生活安排得井井有条(社交层面)
成功的表现:
- “我不用多想就能吃上真正的食物”
- “下午精力充沛”
第三层:感知障碍
顾虑:
- “150 美元/周感觉很贵”(但他们花在外卖上的钱也有这么多)
- “要是我不喜欢这些食物怎么办?”
- “这真的适合我的日程安排吗?”
担忧:
- “又一项我一个月后就会取消的订阅”
- “外送食品的品质如何——能保持新鲜吗?”
引述:
“我试过两次 Blue Apron。两次最后冰箱里都堆满了腐烂的食材。”
第四层:决策标准
必备条件:
- 完全无需烹饪(关键!)
- 健康(真正的食材、营养均衡)
- 种类丰富(不会吃腻)
加分项:
- 可根据偏好定制
- 配送时间灵活
不可接受项:
- 需要任何烹饪或准备工作
- 只有健康的“兔子食物”
- 长期订阅承诺
第五层:买家的旅程
- 在 Instagram 上看到朋友晒美味餐食
- 询问朋友,或搜索“无需烹饪的送餐服务”
- 在网站上比较 2-3 个选项并查看评价
- 试用一周
- 试用效果满意后订阅
时间线: 当周决定是否试用
营销启示
宣传语:
- “吃得像有私人厨师为你服务”
- “零准备,零清理,真正的食物。”
- “适合忙得没空做饭、又聪明到不会吃垃圾食品的人”
触达渠道:
- Instagram(生活方式/美食内容)
- 播客(商业、效率)
- LinkedIn(按行业定向)
不要说:
- “料理包”(听起来像是还要做饭)
- “减肥餐”(听起来限制很多)
用户画像 2:追求极致的运动员
(因篇幅所限,内容从略)
与用户画像 1 的主要区别
| 方面 | 精疲力竭的专业人士 | 追求极致的运动员 |
|---|---|---|
| 主要待完成任务(JTBD) | 毫不费力地健康饮食 | 精准达到我的宏量营养素目标 |
| 触发因素 | 健康顾虑、内疚 | 比赛临近、训练遇到瓶颈 |
| 成功因素 | 便利 | 表现/结果 |
| 决策标准 | 无需烹饪 | 宏量营养素信息透明 |
| 障碍 | “又一项订阅” | “宏量营养素数据准确吗?” |
| 宣传语 | “零费力” | “为你的表现补充能量” |
检查清单与模板
用户画像访谈问题
## Customer Interview Guide for Persona Development
### Trigger (Priority Initiative)
- "What was happening when you first started looking for [solution]?"
- "Why now? Why not 6 months ago?"
- "What was the trigger event?"
- "What finally pushed you to do something?"
### Success Factors
- "What did you hope would change after buying?"
- "How would you know if this worked?"
- "If this was perfect, what would that look like?"
- "What would success mean for you personally?"
### Barriers (Perceived Risks)
- "What concerns did you have before buying?"
- "What almost made you NOT buy?"
- "What worries did you have about making a change?"
- "What would failure look like?"
### Decision Criteria
- "What was most important to you in evaluating options?"
- "What features were must-haves?"
- "What made you choose us over alternatives?"
- "What would have been a deal-breaker?"
### Buyer's Journey
- "How did you first hear about this category?"
- "What did you research? Where?"
- "Who else was involved in the decision?"
- "How long from first search to purchase?"
用户画像验证评分
## Persona Quality Score
**Persona:** _______________
| Criteria | Score (1-5) | Notes |
|----------|-------------|-------|
| Based on real interviews (not assumptions) | | |
| Includes verbatim quotes | | |
| Explains triggers (why now) | | |
| Explains decision criteria | | |
| Surfaces real anxieties | | |
| Maps buyer's journey | | |
| Includes JTBD statement | | |
| Distinct from other personas | | |
| Marketing team can use it | | |
| Sales team can use it | | |
**Total:** __/50
- 40+: Ready to use
- 30-39: Needs more research
- <30: Start over with customer interviews
快速用户画像卡片
## [Persona Name]
**One-liner:** [Job title] who [key behavior] because [primary motivation].
**Trigger:** They start looking when ________________________________
**Job-to-be-Done:** "When I [situation], I want to [motivation], so I can [outcome]."
**Top 3 Decision Criteria:**
1.
2.
3.
**Biggest Anxiety:** ________________________________
**We win by:** ________________________________
技能边界
此技能擅长的方面
- 构建音频制作工作流
- 提供技术指导
- 创建质量检查清单
- 提出创意方案
这项技能无法做到的事
- 取代音频工程方面的专业知识
- 做出主观的创意决策
- 直接访问或编辑音频文件
- 保证商业成功
参考资料
- Revella,Adele。《Buyer Personas》(2015)——Five Rings 方法论
- Christensen,Clayton。《Competing Against Luck》(2016)——Jobs-to-be-Done
- Ulwick,Tony。《What Customers Want》(2005)——Outcome-Driven Innovation
- Blank,Steve。《The Four Steps to the Epiphany》(2005)——客户原型
- Klement,Alan。《When Coffee and Kale Compete》(2016)——产品领域的 JTBD
相关技能
- jobs-to-be-done - 深入了解 JTBD 理论
- buyer-personas - Adele Revella 的原始框架
- mom-test - 如何通过访谈开展客户画像研究
- customer-discovery - 系统化验证方法
- audience-research - 更全面地了解受众
技能元数据(内部使用)
name: persona-generator
category: validation
subcategory: customer-research
version: 1.0
author: MKTG Skills
source_expert: Adele Revella, Clayton Christensen
source_work: Buyer Personas, Competing Against Luck
difficulty: intermediate
estimated_value: $3,000 persona research project
tags: [personas, customer-research, JTBD, buyer-personas, segmentation, YC]
created: 2026-01-25
updated: 2026-01-25