Back to Details

email-response-simulation

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Copy Feedback

文案反馈

Simulate a real prospect reading your cold email. Build their world, then roast the email through their eyes.
模拟真实潜在客户阅读你的冷邮件。构建他们的认知世界，然后站在他们的视角对邮件进行严苛评审。

Related Skills

相关技能

email-generation → email-response-simulation → campaign-sending
Run this for Tier 1 prospects only — the ones worth individual attention. For Tier 2, use templates from 
email-generation
directly.
email-generation → email-response-simulation → campaign-sending
仅针对一级潜在客户使用此功能——这些客户值得投入单独关注。对于二级潜在客户，直接使用
email-generation
生成的模板即可。

Scraping Tell Detection

抓取痕迹检测

During the simulation, check the email for signals that reveal the data was scraped from LinkedIn or a database. These are instant credibility killers — the prospect recognizes them and mentally files the email as automated spam.
Common tells to flag:
  • Titles left in names — "Hey Dr. Martinez" or "Hi Prof. Chen" screams LinkedIn scrape. Real humans say "Hey Carlos"
  • Over-formal name usage — using full name ("Dear Jonathan Smith") instead of first name
  • LinkedIn headline as job title — if the email references a title that's clearly copy-pasted from LinkedIn (e.g. "As a Serial Entrepreneur | Angel Investor | Speaker")
  • Stale data — referencing a role or company the prospect has already left
Flag these as High severity risk flags in Phase 4. Any scraping tell detected should be called out in the Emotional Reaction (Phase 3, Pass 1) as an immediate red flag.
在模拟过程中，检查邮件是否存在暴露数据来自LinkedIn或数据库的痕迹。这些是即时破坏可信度的因素——潜在客户一眼就能识别，会下意识将邮件归类为自动垃圾邮件。
需要标记的常见痕迹：
  • 姓名中保留头衔 ——“Hey Dr. Martinez”或“Hi Prof. Chen”这种表述明显是从LinkedIn抓取的。真实的人会说“Hey Carlos”
  • 过度正式的姓名用法 ——使用全名（“Dear Jonathan Smith”）而非名字
  • 将LinkedIn头衔作为职位称呼 ——如果邮件中引用的头衔明显是直接复制自LinkedIn（例如“作为连续创业者 | 天使投资人 | 演讲者”）
  • 过时数据 ——提及潜在客户已经离开的职位或公司
在第4阶段将这些标记为高风险标识。一旦检测到任何抓取痕迹，需在第3阶段第1轮的情绪反应环节中立即指出。

When to Use

使用场景

  • After 
    email-generation
    produces a draft for a Tier 1 company
  • When you want to validate whether a specific person would reply
  • When you're training your instinct for a new audience before automating
  • When the user says "would [name] reply to this?" or "roast this email"
  • 在
    email-generation
    为一级企业生成邮件草稿后
  • 当你想验证特定收件人是否会回复时
  • 在自动化之前，训练你对新受众的判断直觉时
  • 当用户询问“[姓名]会回复这封邮件吗？”或“roast this email”时

Environment

运行环境

Provider selection and credentials are handled in Phase 1 of the workflow. If LinkedIn data is not available from the Extruct people table, a deep research provider is needed — the skill will ask which one to use.
服务商选择和凭证处理在工作流的第1阶段完成。如果Extruct人员表中没有LinkedIn数据，则需要使用深度调研服务商——该功能会询问你要使用哪一个。

Inputs

输入项

InputRequiredSource
Draft email textyesUser pastes or from 
email-generation
output
Prospect nameyesFrom contacts CSV or user input
Prospect companyyesFrom contacts CSV or user input
Prospect job titleyesFrom contacts CSV or user input
Prospect LinkedIn URLstrongly recommendedFrom 
people-search
output
Hypothesis matchedrecommendedFrom 
list-segmentation
output
Extruct people table dataoptionalLinkedIn Data column from 
people-search
输入项是否必填来源
邮件草稿文本是用户粘贴或来自
email-generation
的输出
潜在客户姓名是来自联系人CSV或用户输入
潜在客户企业是来自联系人CSV或用户输入
潜在客户职位是来自联系人CSV或用户输入
潜在客户LinkedIn链接强烈推荐来自
people-search
的输出
匹配的假设推荐来自
list-segmentation
的输出
Extruct人员表数据可选来自
people-search
的LinkedIn Data列

Workflow

工作流

Phase 1: Data Enrichment and World Building

阶段1：数据丰富与世界构建

Gather as much public data as possible, then simulate the prospect's full world.
Step 1a: Scrape public data
If LinkedIn data is available from the Extruct people table (linkedin_data column), use it directly. Otherwise, run research queries via the chosen deep research provider (from Environment).
Query 1 — Professional identity:
"Research [Name], [Title] at [Company]. Find their LinkedIn About section, career history, previous companies and roles, education, and any professional accomplishments. Include their full career arc — where they started, how they got to this role. Look at LinkedIn, company bios, conference speaker pages, and press mentions."
Query 2 — Public voice and positions:
"Has [Name] from [Company] written or said anything publicly? Look for LinkedIn posts, blog posts, podcast appearances, conference talks, interviews, or published articles. What topics do they engage with? What language do they use? Include direct quotes if available."
Query 3 — Company and role context:
"What is [Company] doing right now? Recent news, strategic initiatives, funding, product launches, or hiring signals from the last 12 months. What challenges would a [Title] at this type of company be dealing with? What does their team likely look like?"
Step 1b: Simulate their world
Using the research, build a vivid simulation of this person's professional reality. Go deep — the goal is to predict how they think, not just what they do.
markdown
undefined
收集尽可能多的公开数据，然后模拟潜在客户的完整认知世界。
步骤1a：抓取公开数据
如果Extruct人员表（linkedin_data列）中有LinkedIn数据，直接使用。否则，通过选定的深度调研服务商（来自环境设置）运行调研查询。
查询1——职业身份：
“调研[姓名]，[职位]，就职于[企业]。查找他们的LinkedIn简介、职业经历、过往公司与职位、教育背景以及任何职业成就。涵盖完整的职业发展轨迹——从何处起步，如何达到当前职位。查看LinkedIn、企业官网介绍、会议演讲者页面和新闻提及。”
查询2——公开言论与立场：
“[企业]的[姓名]是否有过公开的文字或言论？查找LinkedIn帖子、博客文章、播客嘉宾、会议演讲、采访或已发表的文章。他们关注哪些话题？使用何种语言风格？如有可用的直接引语，请包含在内。”
查询3——企业与职位背景：
“[企业]当前的动态如何？过去12个月内的最新新闻、战略举措、融资、产品发布或招聘信号。这类企业的[职位]会面临哪些挑战？他们的团队可能是什么样的？”
步骤1b：模拟他们的认知世界
利用调研结果，生动模拟该用户的职业现状。深入挖掘——目标是预测他们的思维方式，而非仅仅了解他们的行为。
markdown
undefined

World Simulation: [Name]

世界模拟：[姓名]

Daily Reality

日常现状

  • What their day looks like: [Meetings, priorities, where they spend time]
  • Tools they live in: [CRM, Slack, email volume, dashboards]
  • What keeps them up at night: [The 1-2 problems they can't solve yet]
  • 一天的工作内容： [会议、优先级、时间分配]
  • 常用工具： [CRM、Slack、邮件数量、数据面板]
  • 最头疼的问题： [1-2个尚未解决的核心难题]

Psychology and Decision-Making

心理与决策模式

  • Decision style: [Data-driven / gut / consensus-seeker / authority-based]
  • Communication preference: [Terse / detailed / formal / casual / story-driven]
  • Risk appetite: [Early adopter / wait for proof / only buys market leaders]
  • Trust signals: [What makes them trust a vendor — referrals, data, case studies, free trial]
  • 决策风格： [数据驱动/直觉/寻求共识/权威主导]
  • 沟通偏好： [简洁/详细/正式/随意/故事导向]
  • 风险偏好： [早期采用者/等待验证/只买市场领导者产品]
  • 信任信号： [什么因素会让他们信任供应商——推荐、数据、案例研究、免费试用]

Inbox Behavior

收件箱行为

  • Email volume: [Estimate — 50/day? 200/day?]
  • Cold email tolerance: [Opens most? Deletes by subject? Has assistant filter?]
  • Reply-blocking mindset: [What makes them NOT reply even if interested]
  • What triggers a reply: [Specific enough to feel researched, not so specific it feels creepy]
undefined
  • 邮件数量： [估算——每天50封？200封？]
  • 冷邮件容忍度： [大部分都会打开？看主题就删除？由助理过滤？]
  • 拒绝回复的心态： [即使感兴趣也不会回复的原因]
  • 回复触发点： [具体到足够体现调研，但又不会让人觉得突兀]
undefined

Phase 2: Professional Reality

阶段2：职业现状定义

Define the parameters that determine whether your email hits or misses.
markdown
undefined
定义决定邮件成败的关键参数。
markdown
undefined

Professional Reality

职业现状

KPIs and Motivators

KPI与动机

  • Measured on: [The 2-3 KPIs their boss actually evaluates them on]
  • Career motivator: [What they're trying to achieve — promotion, build something, stability]
  • Internal politics: [Who do they need buy-in from? Budget authority?]
  • 考核指标： [老板实际评估他们的2-3个KPI]
  • 职业动机： [他们想要达成的目标——晋升、打造新事物、稳定]
  • 内部关系： [需要获得谁的支持？预算决策权在谁手中？]

Pain Points (ranked by severity)

痛点（按严重程度排序）

  1. [Most acute pain — the thing that wastes their time or blocks their goals]
  2. [Second pain]
  3. [Third pain]
  1. [最迫切的痛点——浪费时间或阻碍目标的核心问题]
  2. [次要痛点]
  3. [第三痛点]

Relationship to Your Solution

与解决方案的关系

  • Awareness level: [Never heard of tools like this / aware of category / tried competitors]
  • Current workaround: [How they solve this problem today without you]
  • Switching cost: [What would they have to give up or change to use you]
undefined
  • 认知程度： [从未听说过此类工具/了解品类/试过竞品]
  • 当前替代方案： [没有你的产品时，他们当前如何解决该问题]
  • 转换成本： [使用你的产品需要放弃或改变什么]
undefined

Phase 3: Skeptical Buyer Roast

阶段3：挑剔买家式评审

Now put on the prospect's hat and read the email cold. Two distinct passes:
Pass 1: Emotional Reaction (2 seconds)
This is the gut response — before any rational evaluation. Read the email as if you're this person, scanning your inbox between meetings.
markdown
undefined
现在代入潜在客户的角色，客观阅读邮件。分为两轮独立评审：
第一轮：情绪反应（2秒内）
这是直觉反应——在任何理性评估之前。模拟该用户在会议间隙浏览收件箱时阅读邮件的状态。
markdown
undefined

Emotional Reaction

情绪反应

Subject line gut feel: [What they think in 0.5 seconds] First sentence gut feel: [Do they keep reading or move on?] Overall vibe: [Feels like spam / feels like a human / feels like someone who gets my world] Immediate red flags: [Anything that triggers "delete" instinct]

**Pass 2: Business Evaluation (10 seconds)**

If they made it past the emotional filter, they now evaluate the substance. This is the more calculated read.

```markdown
主题栏直觉感受： [0.5秒内的第一想法] 第一句话直觉感受： [会继续阅读还是直接跳过？] 整体氛围： [感觉像垃圾邮件/像真人发送/像懂我的人发送] 即时红色预警： [任何触发“删除”本能的因素]

**第二轮：业务评估（10秒内）**

如果邮件通过了情绪筛选，他们会开始评估内容实质。这是更理性的阅读过程。

```markdown

Business Evaluation

业务评估

Does this hit my KPIs? [Yes/no — and which one specifically] Priority level: [Would I deal with this today, this week, this quarter, or never?] Bridge quality: [How strong is the connection between my pain and their solution?] Credibility check: [Do I believe this person/company can deliver?] Effort-to-value ratio: [Is the CTA worth my time? What am I risking by replying?]
undefined
是否契合我的KPI？ [是/否——具体说明哪一个KPI] 优先级： [我会今天处理、本周处理、本季度处理，还是永远不处理？] 关联度： [我的痛点与你的解决方案之间的关联强度如何？] 可信度验证： [我是否相信这个人/公司能兑现承诺？] 投入产出比： [CTA是否值得我花时间？回复会有什么风险？]
undefined

Phase 4: Risk Flags and Refinement

阶段4：风险标识与优化方向

Identify specific issues in the copy that reduce reply probability.
markdown
undefined
识别邮件文案中降低回复率的具体问题。
markdown
undefined

Risk Flags

风险标识

FlagLocationSeverityIssue
[flag type][P1/P2/P3/P4/subject][High/Med/Low][what's wrong]

**Flag types to check:**
- **Spam trigger** — phrasing that sounds like mass email, not personal
- **Wrong pain** — opener references a problem this person doesn't actually have
- **Weak bridge** — gap between their pain and your solution is too big or too vague
- **Bad personalization** — mentions something generic disguised as personal
- **CTA mismatch** — ask is too big (meeting) or too small (nothing) for this prospect
- **Tone clash** — email tone doesn't match how this person communicates
- **Credibility gap** — proof point doesn't resonate with this type of buyer
- **Length violation** — too long for their inbox behavior
- **Scraping tell** — name, title, or data reveals it was pulled from LinkedIn or a database (titles left in names, LinkedIn headline as job title, stale role data)
标识类型位置严重程度问题说明
[标识类型][P1/P2/P3/P4/主题栏][高/中/低][具体问题]

**需要检查的标识类型：**
- **垃圾邮件触发词** ——听起来像批量发送而非个性化的表述
- **痛点匹配错误** ——开头提及的问题并非该用户的实际痛点
- **关联度弱** ——用户痛点与你的解决方案之间的差距过大或模糊
- **个性化不足** ——将通用内容伪装成个性化内容
- **CTA不匹配** ——请求对于该潜在客户来说过于宏大（如要求开会）或过于琐碎
- **语气冲突** ——邮件语气与该用户的沟通风格不符
- **可信度缺失** ——证明点无法引起该类买家的共鸣
- **长度超标** ——超出该用户的收件箱阅读习惯
- **抓取痕迹** ——姓名、头衔或数据暴露其来自LinkedIn或数据库（姓名中保留头衔、LinkedIn头衔作为职位称呼、过时职位数据）

Phase 5: Output

阶段5：输出结果

Present the full analysis and rewrite.
markdown
undefined
呈现完整的分析与重写版本。
markdown
undefined

Copy Feedback: [Name] at [Company]

文案反馈：[企业]的[姓名]

Verdict

结论

Would they reply? [Yes / Maybe / No] Reasoning: [1-2 sentences — the real reason, not a polite version]
他们会回复吗？ [是/可能/否] 理由： [1-2句话——真实原因，而非委婉表述]

Emotional Reaction Summary

情绪反应总结

[2-3 sentences — what they feel in the first 2 seconds]
[2-3句话——他们在最初2秒内的感受]

Business Evaluation Summary

业务评估总结

[2-3 sentences — does it hit their KPIs, is the bridge strong]
[2-3句话——是否契合他们的KPI，关联度是否足够]

Risk Flags

风险标识

[Table from Phase 4]
[来自第4阶段的表格]

Ranked Changes

优先级优化建议

  1. [Highest impact]
    • Current: "[exact text]"
    • Problem: [why it fails for THIS person]
    • Rewrite: "[new text]"
  2. [Second]
    • Current: ...
    • Problem: ...
    • Rewrite: ...
  3. [Third] ...
  1. [影响最高]
    • 当前内容：“[原文]”
    • 问题：[为何对该用户无效]
    • 重写：“[新内容]”
  2. [次要影响]
    • 当前内容：...
    • 问题：...
    • 重写：...
  3. [第三影响] ...

Rewritten Email

重写后的邮件

[Full email with all changes applied]
undefined
[应用所有优化后的完整邮件]
undefined

Phase 6: Iterate

阶段6：迭代优化

After presenting, ask:
  • "Want to adjust any of the changes?"
  • "Want to run another prospect through the same email template?"
  • "Ready to finalize?"
Repeat Phases 3-5 until the user is satisfied.
输出结果后，询问用户：
  • “是否需要调整任何优化内容？”
  • “是否需要用同一邮件模板测试另一位潜在客户？”
  • “是否准备定稿？”
重复第3-5阶段，直到用户满意为止。

Running One by One vs. Template Training

单次处理与模板训练的区别

One by one (primary use): Run this for each Tier 1 prospect individually. The value prop, opener, and CTA should be tailored to their specific role and world.
Template training (secondary use): Even running 3-5 prospects through this process teaches you patterns about the audience. After a few roasts, you'll notice:
  • Which pain points consistently hit
  • Which CTAs work for which seniority
  • Which proof points resonate with which role type
  • What inbox behavior looks like for this ICP
Use these patterns to improve the base template in 
email-generation
for Tier 2 prospects.
单次处理（主要用途）： 针对每位一级潜在客户单独处理。价值主张、开头和CTA应根据他们的具体职位和认知世界量身定制。
模板训练（次要用途）： 即使仅对3-5位潜在客户进行此流程，也能让你了解受众的规律。经过几次评审后，你会发现：
  • 哪些痛点能持续引发共鸣
  • 哪些CTA适合不同层级的用户
  • 哪些证明点能引起不同职位用户的共鸣
  • 该理想客户画像（ICP）的收件箱行为特征
利用这些规律优化
email-generation
中的基础模板，供二级潜在客户使用。

Guidelines

准则

  • Never invent persona details. If research doesn't find something, say "unknown" and reason from the role/title instead.
  • Be brutally honest. If the email will be deleted, say so. Sugarcoating wastes the user's time.
  • The emotional reaction is primary. Most cold emails are killed in 2 seconds. The business case only matters if they survive the gut check.
  • Respect the voice rules. All rewrites must stay within the voice constraints from 
    email-generation
    / 
    email-prompt-building
    .
  • One email, one prospect. Never run in bulk. The whole point is depth.
  • The bridge is everything. The gap between "their pain" and "your solution" is where most emails fail. If the bridge is weak, no amount of copywriting fixes it — the hypothesis match might be wrong.
  • 切勿虚构角色细节。 如果调研未找到相关信息，请标注“未知”，并根据职位/头衔进行合理推断。
  • 保持绝对坦诚。 如果邮件会被直接删除，如实说明。粉饰太平只会浪费用户的时间。
  • 情绪反应是核心。 大多数冷邮件在2秒内就会被删除。只有通过直觉筛选，业务价值才会被考虑。
  • 遵循语气规则。 所有重写内容必须符合
    email-generation
    /
    email-prompt-building
    中的语气约束。
  • 一封邮件对应一位客户。 切勿批量运行。深度定制是此功能的核心价值。
  • 关联度是关键。 用户痛点与你的解决方案之间的差距是大多数邮件失败的原因。如果关联度弱，再多的文案优化也无济于事——可能是假设匹配错误。

Reference

参考资料

See references/persona-review-template.md for the full output template and red flag checklist.
完整的输出模板和红色预警清单请查看references/persona-review-template.md。