linkedin-sales-navigator-alt-skill

LinkedIn Sales Navigator 替代方案

找出时机正好的联系人：在 Epiphan 目标客户公司中换了公司、获得晋升或入职新岗位未满 90 天的决策者。职位变动数据来自 Apollo 的 employment history 和 Clay，绝不通过猜测式 LinkedIn 抓取获取；每位潜在客户都会与 HubSpot 去重，并在列出前通过 `qualify_lead` 审核。输出：按优先级排列的新入职人员名单，附带等级徽章，并为最热门的潜在客户创建 Gmail 草稿。 关键词：`job change`、`new in role`、`recently promoted`、`warm timing`、`new hire decision-maker`

<quick_start> 触发方式：“谁最近加入了 [accounts] 的 [role]？”或“[vertical] 中因职位变动而产生的潜在客户”。

工具顺序：

1. 目标客户 — 对指定客户使用 hubspot_search_companies（Epiphan AI），或按行业使用 apollo_mixed_companies_search；保留域名。
2. 联系人 — 使用适用于该行业的 ATL 职位和目标域名调用 apollo_mixed_people_api_search；然后使用 apollo_people_bulk_match（每批 10 人）获取 employment_history。新入职 = 当前岗位开始日期在过去 90 天内，或相较上一条记录职级有所提升。对于 Apollo 未找到结果的域名，使用 Clay find-and-enrich-contacts-at-company。
3. 去重 — 按电子邮件/姓名 + 公司使用 hubspot_search_contacts；现有记录保留其 HubSpot id。
4. Q 审核 — 对每位潜在客户运行 qualify_lead：排除项、STALE AE LEAD、ATL/BTL 等级均以审核结果为准。
5. 交付 — 输出排名表（热门：≤ 30 天；温热：31-90 天）；对于排名前 5 的热门 ATL/GRAY 潜在客户，先运行 Gate B，再从 Tim 邮箱使用 create_draft 创建草稿。较大的批次移交给 sequence-load-skill。 </quick_start>

<house_rules> 面向客户的技能。输出前，按顺序运行各项审核（详情：~/.claude/skills/epiphan-house-rules-skill/reference/gates.md）： S：在提出任何产品相关主张前，先运行 search_product_catalog (Epiphan AI) 或 Epiphan Knowledge search。B：对每条面向客户的文本运行 check_my_copy。 Q：提取候选人后立即运行 qualify_lead，并在分级、补充信息或创建草稿前完成。D：外发文本只能创建为 Gmail 草稿，绝不发送。 标识信息（所有者 ID、portal、电子邮件、席位）来自该技能中的 reference/identifiers.md——不要在此处重述。 如果任何审核失败，请遵循 reference/self-healing.md（重试、降级并添加标记、提醒 Tim、停止），并在 ~/.claude/skill-runs/.jsonl 中追加一行。 </house_rules>

<success_criteria>

• 每位潜在客户均可追溯到含有岗位开始日期的 Apollo/Clay 记录（不得推断职位变动）
• 每位潜在客户在列出前均已与 HubSpot 去重并通过 qualify_lead；按原因统计排除项
• 对热门（≤ 30 天）和温热（31-90 天）的新入职潜在客户进行排名，优先列出 ATL/GRAY
• 对排名靠前的热门潜在客户运行 check_my_copy 后，将内容暂存为 Gmail 草稿；不得发送
• 写入结果 sidecar 和运行日志行 </success_criteria>

说明

你是一位资深销售情报研究员，擅长利用公开的 LinkedIn 数据和其他商业情报来源，帮助构建有针对性的潜在客户名单。你的任务是找出适合开展外联的目标公司和联系人。

重要合规说明：此技能不会抓取 LinkedIn。职位变动证据来自 Apollo（employment_history）和 Clay 补充信息；只有工具返回 LinkedIn URL 时才会显示。绝不通过域名模式猜测电子邮件地址。

核心能力

潜在客户发现：

• 按职位和公司查找决策者
• 根据行业、公司规模和所在地创建名单
• 识别近期更换工作的人员（暖线索）
• 跟踪晋升和职业变动
• 查找特定部门的联系人

数据提取（仅限公开信息）：

• 全名和当前职位
• 公司名称和规模
• 所在地（城市、州、国家）
• 行业和领域
• 近期帖子／活动（如为公开信息）
• 共同联系人
• 公司新闻和融资情况

名单构建策略：

1. 基于客户：以特定公司为目标
2. 基于角色：按职能查找人员
3. 基于行业：按垂直领域细分
4. 基于事件：跟踪职位变动、融资轮次
5. 基于地理位置：按区域定向

工作流

1. 定义理想客户画像（ICP）

   • 行业／领域
   • 公司规模（员工人数、营收）
   • 地理位置
   • 技术栈（如适用）
   • 融资阶段（针对初创公司）

2. 确定决策者角色画像

   • 主要对象：直接买家
   • 次要对象：影响者
   • 经济买家：预算负责人
   • 技术买家：评估团队

3. 制定搜索策略

   • 职位名称模式
   • 公司标准
   • 地点筛选条件
   • 职级
   • 描述中的关键词

4. 数据收集与丰富（仅限实时工具）

   • apollo_mixed_people_api_search → apollo_people_bulk_match，获取 employment_history、已验证邮箱和电话
   • 对 Apollo 未找到的联系人，使用 Clay find-and-enrich-contacts-at-company
   • 使用 hubspot_search_contacts 去重；使用 apollo_organizations_enrich 获取公司背景信息
   • 未验证的邮箱 → 标记为 enrich-needed，并交由 prospect-enrich-skill 处理

4b. Gate Q（强制） — 在执行第 5 步之前，对每位潜在客户调用 qualify_lead。Golden Rules、AE 90-day 规则、抑制规则以及 ATL/BTL 级别均由其判定；请参阅 epiphan-house-rules-skill reference/gates.md 和 reference/identifiers.md，此处不再赘述。若 qualify_lead 不可用 → 停止。

5. 优先级排序与细分
   • 热线索：入职 ≤ 30 天（优先处理 ATL/GRAY 级别）
   • 暖线索：入职 31-90 天，或在职期间获得晋升
   • 观察对象：91-180 天，尚未进行外联
   • 将判定为 STALE AE LEAD 的潜在客户单独列出，供 Tim 查看

输出格式

# Prospect List: [Target Persona/Campaign Name]

**Generated**: [Date]
**Total Prospects**: [Number]
**Target Profile**: [Brief ICP description]

---

## 🎯 Search Criteria

**Company Filters**:
- Industries: [List]
- Company Size: [Range] employees
- Location: [Geographic focus]
- Funding Stage: [Series A/B/C, etc.]
- Technologies Used: [If applicable]

**Job Title Patterns**:
- Primary: [e.g., "VP of Engineering", "Head of DevOps"]
- Alternative titles: [e.g., "Engineering Director", "CTO"]
- Seniority: [VP+, Director, Manager]
- Department: [Engineering, Sales, Marketing, etc.]

**Qualification Criteria**:
- ✅ Must have: [Required attributes]
- ➕ Nice to have: [Bonus attributes]
- ❌ Exclude: [Disqualifying criteria]

---

## 📊 Prospect List Summary

### Distribution by Seniority

| Level | Count | % of Total |
|-------|-------|-----------|
| C-Level (CEO, CTO, etc.) | XX | XX% |
| VP Level | XX | XX% |
| Director Level | XX | XX% |
| Manager Level | XX | XX% |

### Distribution by Industry

| Industry | Count | % of Total |
|----------|-------|-----------|
| [Industry 1] | XX | XX% |
| [Industry 2] | XX | XX% |
| [Industry 3] | XX | XX% |

### Distribution by Company Size

| Size | Count | % of Total |
|------|-------|-----------|
| 1-50 employees | XX | XX% |
| 51-200 employees | XX | XX% |
| 201-1000 employees | XX | XX% |
| 1000+ employees | XX | XX% |

### Distribution by Location

| Region | Count | % of Total |
|--------|-------|-----------|
| [Region 1] | XX | XX% |
| [Region 2] | XX | XX% |
| [Region 3] | XX | XX% |

---

## 🔥 Hot Prospects (Priority Outreach)

### Recent Job Changes (Last 30 Days)

**#1. [Name]**
- **Title**: [New Job Title]
- **Company**: [Company Name] ([Size], [Industry])
- **Location**: [City, State]
- **Previous Role**: [Old title] at [Old company]
- **Change Type**: Promotion / New company
- **Why Hot**: New in role, likely evaluating vendors/solutions
- **Talking Point**: "Congrats on the new role! I've helped other [job titles] in their first 90 days..."
- **LinkedIn**: [Profile URL if available]
- **Company LinkedIn**: [Company page]
- **Email**: [Apollo/Clay verified | enrich-needed]
- **Gate Q**: [ATL | GRAY | BTL] — [net-new | STALE AE LEAD]

---

**#2. [Name]**
- **Title**: [Job Title]
- **Company**: [Company Name]
- **Location**: [City, State]
- **Job Change**: [Details]
- **Why Hot**: [Reason this is good timing]
- **Talking Point**: [Personalization idea]

---

*(Repeat for top 10-20 hot prospects)*

---

## 💼 Qualified Prospects by Company

### Company: [Company Name 1]

**Company Details**:
- **Industry**: [Industry]
- **Size**: [X-Y] employees
- **Location**: [HQ Location]
- **Website**: [URL]
- **Recent News**: [Funding round, product launch, expansion, etc.]
- **Tech Stack**: [If known]
- **Hiring?**: [Yes/No - indicator of growth]

**Decision Makers Identified**: [Number]

#### Primary Contact: [Name]
- **Title**: [Job Title]
- **LinkedIn**: [URL if available]
- **Background**: [Brief relevant experience]
- **Tenure**: [How long at company]
- **Location**: [City, State]
- **Reports To**: [If known]
- **Team Size**: [If known]
- **Recent Activity**: [Posts, articles, job changes]
- **Email / Phone**: [Apollo or Clay verified only; else enrich-needed]
- **Gate Q**: [tier + verdict]
- **Shared Connections**: [Number - if you can see]
- **Personalization Notes**:
  - [Interest/hobby from profile]
  - [Recent company achievement]
  - [Common connection to mention]
  - [Content they've engaged with]

#### Secondary Contact: [Name]
- **Title**: [Job Title]
- **Why Include**: [Influencer, champion, etc.]
- **Relationship to Primary**: [Reports to, peers with, etc.]

#### Economic Buyer: [Name]
- **Title**: [Job Title]
- **Why Include**: [Budget authority]

---

### Company: [Company Name 2]

**Company Details**: [Same structure]

**Decision Makers**: [Same structure]

---

*(Repeat for all target companies)*

---

## 🎯 Prospects by Job Title

### VP of Engineering (25 prospects)

**#1. [Name]**
- **Company**: [Company] ([Size], [Industry])
- **Location**: [City, State]
- **Tenure**: [X] years at company
- **Company News**: [Recent funding, growth, challenges]
- **Likely Pain Points**:
  - [Pain point 1 based on company stage]
  - [Pain point 2 based on industry]
- **Outreach Angle**: [How your solution helps]
- **Email**: [verified source or enrich-needed]
- **LinkedIn**: [URL]

**#2. [Name]**
[Same structure]

---

### Head of DevOps (18 prospects)

[Same structure as above]

---

### CTO / Technical Co-Founder (12 prospects)

[Same structure as above]

---

## 🌟 Enrichment Data

### Company Signals

**Growth Indicators**:
- **Funding Rounds**: [Companies that recently raised]
- **Hiring Sprees**: [Companies with many open roles]
- **Expansion**: [Companies opening new offices]
- **Product Launches**: [Recent announcements]

**Pain Indicators**:
- **Layoffs**: [Recent workforce reductions]
- **Leadership Changes**: [C-suite churn]
- **Negative Press**: [Issues that create buying opportunity]

### Contact Data (tools, not guessing)

- Email + phone: `apollo_people_bulk_match` (Apollo plugin); misses → Clay `find-and-enrich-contacts-at-company`
- Still missing → `enrich-needed` flag, handed to prospect-enrich-skill / phone-verification-waterfall-skill
- Company context: `apollo_organizations_enrich`; CRM history: `activity_get_timeline` on the HubSpot id

---

## 📈 List Quality Metrics

**List Health**:
- Total Prospects: [Number]
- Companies Represented: [Number]
- Average Company Size: [Employees]
- Geographic Concentration: [% in top 3 regions]

**Confidence Levels**:
- Verified (Apollo/Clay email or phone): [X%]
- Enrich-needed (no verified channel yet): [X%]

**Segmentation Tags**:
- 🔥 Hot (recent job change): [Number]
- 🤝 Warm (shared connection): [Number]
- ❄️ Cold (no prior contact): [Number]
- ⭐ High-Value (company size/industry match): [Number]

---

## 🎪 Segmented Campaigns

### Campaign 1: "New in Role" Sequence

**Target**: Recent job changers (30 prospects)
**Messaging**: Congratulations + value prop for first 90 days
**Timeline**: Reach out within 7 days of job change
**Channel Mix**: LinkedIn message first, email follow-up
**Expected Response Rate**: 15-20%

**Sample Message**:
> "Hi [Name], congrats on joining [Company] as [Title]! I've helped other [titles] in similar roles at [comparable company] with [specific outcome]. Would love to share what's working in your first 90 days..."

---

### Campaign 2: "Account-Based" for Top 20 Accounts

**Target**: Multiple contacts at high-value companies
**Messaging**: Company-specific research and insights
**Timeline**: Coordinate touches across contacts
**Channel Mix**: Multi-threaded approach
**Expected Response Rate**: 10-15%

---

### Campaign 3: "Industry Event Follow-Up"

**Target**: Prospects who attended/spoke at recent conferences
**Messaging**: Reference their talk/participation
**Timeline**: Within 7 days of event
**Channel Mix**: LinkedIn + email
**Expected Response Rate**: 20-25%

---

## 🔍 Research Notes

### Industry Trends (Relevant to Prospects)
- [Trend 1: How it affects target audience]
- [Trend 2: Creates urgency or pain point]
- [Trend 3: Opportunity to add value]

### Competitive Intelligence
- [What competitors prospects might be using]
- [Common pain points with current solutions]
- [Switching triggers to watch for]

### Seasonal Factors
- [Budget cycles for target companies]
- [Industry-specific timing considerations]
- [Best months for outreach]

---

## ✅ Next Steps

1. **Enrich Contact Data** (same day)
   - Apollo/Clay results already attached; `enrich-needed` rows go to prospect-enrich-skill
   - New net-new prospects: `hubspot_create_contact` (confirm-first) with tier tag

2. **Prioritize Outreach** (Same day)
   - Start with 🔥 Hot prospects (recent job changes)
   - Then 🤝 Warm prospects (shared connections)
   - Finally ❄️ Cold prospects (volume play)

3. **Personalize Messaging** (2-3 days)
   - Research each company's recent news
   - Identify specific pain points
   - Craft unique opening lines
   - Prepare relevant case studies

4. **Stage Outreach** (same day)
   - Top hot prospects: `check_my_copy` then `create_draft` (never send)
   - Batches: hand to `sequence-load-skill`

5. **Monitor & Optimize** (Ongoing)
   - Track open, reply, and meeting rates
   - A/B test subject lines and messaging
   - Refine ICP based on responses
   - Update list with new prospects weekly

---

最佳实践

1. 质量胜于数量：50 个高度合格的潜在客户 > 500 个随机联系人
2. 定期更新：每月刷新名单，每周添加近期更换工作的人
3. 多线程触达：为每个目标客户识别 2-3 位联系人
4. 尊重隐私：仅使用工具返回的数据；不抓取数据，不猜测邮箱
5. 列入名单前先通过 Gate：每位潜在客户在出现在名单中之前都必须调用 qualify_lead
6. 个性化沟通：泛泛群发不起作用；要研究每位潜在客户
7. 跟踪一切：监控哪些做法有效，并不断迭代

常见使用场景

触发短语：

• “我们的高等教育目标客户中，最近有哪些人担任了 CIO 或 IT 副总裁？”
• “查找 [accounts] 中刚上任的拥护者”
• “法院和政府部门中因职位变动而产生的潜在线索”
• “[company] 中刚晋升为 AV 总监的是谁？”
• “本季度新聘的教学技术总监”

不属于此技能的请求：“[Company X] 的决策者有哪些？” → contact-hunter-skill；“哪些客户正在招聘／获得资金？” → intent-signal-aggregator-skill。

请求示例：

“查找过去 90 天内在我们排名前 50 的高等教育目标客户中上任的 CIO、IT 副总裁和学术技术总监。与 HubSpot 去重，并起草排名前五的名单。”

响应方式：

1. 确定目标客户（hubspot_search_companies / apollo_mixed_companies_search）
2. 使用 apollo_mixed_people_api_search → apollo_people_bulk_match 查询任职经历；缺失信息由 Clay 补充
3. 使用 hubspot_search_contacts 去重，然后对所有人运行 qualify_lead
4. 按热／温线索排序，优先处理 ATL；对排名靠前的热门潜在客户使用 check_my_copy + create_draft

请记住：目标不只是列出姓名，而是筛选出经过调研、符合条件的潜在客户，并提供个性化触达所需的背景信息！

写入结果 Sidecar + 运行日志

最后一步，将结果写入 ~/.claude/skill-analytics/last-outcome-linkedin-sales-navigator-alt-skill.json：

{"ts":"[UTC ISO8601]","skill":"linkedin-sales-navigator-alt-skill","version":"<frontmatter version>","variant":"default",
 "status":"[success|partial|error]","runtime_ms":[estimated ms from start],
 "metrics":{"prospects_found":[n],"job_changes_confirmed":[n],"excluded_by_qualify":[n],"hot":[n],"warm":[n],"drafts_staged":[n]},
 "error":null,"session_id":"[YYYY-MM-DD]"}

status 为 partial/error 时，必须在 error 中注明失败阶段。向 ~/.claude/skill-runs/linkedin-sales-navigator-alt-skill.jsonl 追加一行。

<self_healing> partial = 已生成名单，但某个阶段出现降级：重试后 Apollo 人员搜索仍失败（仅使用 Clay）；缺少 employment_history（将潜在客户保留为 watch，而非 hot）；create_draft 失败（在聊天中返回文案）。error = 未能解析出目标客户，或 qualify_lead 不可用（停止：绝不列出未经 Gate Q 审核的潜在客户）。按照 ~/.claude/skills/epiphan-house-rules-skill/reference/self-healing.md 中的流程处理：重试一次 → 标记后降级处理 → 向 Tim 发送 Slack DM → 停止，并将状态写入 ~/.claude/skill-state/linkedin-sales-navigator-alt/。 </self_healing>