Evaluate the account quality and influence of social media influencers, calculate comprehensive scores, and assist brands in screening KOL cooperation targets.
Trigger Conditions
Start this skill when the user says the following:
- "Evaluate this influencer"
- "Analyze this KOL"
- "influencer evaluation"
- "Influencer data analysis"
- "Help me take a look at this blogger"
- "Is this creator worth collaborating with?"
Workflow
┌─────────────────────┐
│ 输入网红信息 │
│ (用户名/链接) │
└──────────┬──────────┘
│
▼
┌─────────────────────┐
│ 识别平台类型 │
│ IG/YouTube/TikTok │
└──────────┬──────────┘
│
▼
┌─────────────────────┐
│ 搜索公开数据 │
│ 粉丝/互动/内容 │
└──────────┬──────────┘
│
▼
┌─────────────────────┐
│ 计算综合评分 │
│ 多维度加权 │
└──────────┬──────────┘
│
▼
┌─────────────────────┐
│ 生成评估报告 │
│ 合作建议 │
└─────────────────────┘
Execution Steps
Step 1: Obtain Influencer Information
Input Method：
用户: 帮我评估这个网红 @username
用户: 分析一下 https://instagram.com/username
用户: 这个YouTube频道值得合作吗 https://youtube.com/@channel
Parse Information：
- Platform type (Instagram/YouTube/TikTok/Xiaohongshu)
- Username or channel ID
- Profile link
Step 2: Search Public Data
Use WebSearch to search for public information about the influencer:
Search Strategy：
"{用户名} {平台} followers stats"
"{用户名} social blade statistics"
"{用户名} engagement rate"
"{用户名} influencer marketing"
site:socialblade.com "{用户名}"
Target Data：
| Dimension | YouTube | TikTok | |
|---|---|---|---|
| Number of followers | Followers | Subscribers | Followers |
| Content volume | Posts | Videos | Videos |
| Number of interactions | Likes/Comments | Views/Likes | Views/Likes |
| Update frequency | Posts/week | Videos/month | Videos/week |
| Account age | Account age | Channel age | Account age |
Step 3: Calculate Key Metrics
Engagement Rate：
Instagram: (平均点赞 + 平均评论) / 粉丝数 × 100%
YouTube: 平均观看量 / 订阅数 × 100%
TikTok: (平均点赞 + 评论 + 分享) / 粉丝数 × 100%
Engagement Rate Rating：
| Rating | YouTube | TikTok | |
|---|---|---|---|
| Excellent | >6% | >10% | >10% |
| Good | 3-6% | 5-10% | 5-10% |
| Average | 1-3% | 2-5% | 2-5% |
| Poor | <1% | <2% | <2% |
Follower Growth Trend：
- Search for follower changes over the last 30 days/90 days
- Determine whether it is rising, stable, or declining
Content Quality Metrics：
- Posting frequency
- Content verticality
- Comment quality (real interactions vs bots)
Step 4: Comprehensive Scoring Model
Scoring Dimensions and Weights：
| Dimension | Weight | Description |
|---|---|---|
| Follower scale | 15% | Foundation of influence |
| Engagement rate | 30% | Follower activity |
| Content quality | 20% | Verticality and professionalism |
| Growth trend | 15% | Development potential |
| Commercial value | 20% | Historical cooperation cases |
Score Calculation：
总分 = Σ (各维度得分 × 权重)
各维度得分 (1-10分):
- 粉丝规模: 根据量级评分
- 互动率: 根据行业基准评分
- 内容质量: AI 分析评估
- 增长趋势: 根据数据变化评分
- 商业价值: 根据合作案例评分
Grade Classification：
| Total Score | Grade | Recommendation |
|---|---|---|
| 8.5-10 | S-level | Strongly recommended, prioritize cooperation |
| 7.0-8.4 | A-level | Recommend cooperation |
| 5.5-6.9 | B-level | Can be considered |
| 4.0-5.4 | C-level | Needs careful evaluation |
| <4.0 | D-level | Cooperation not recommended |
Step 5: Generate Evaluation Report
Report Format：
# KOL评估报告
## 📊 基础信息
| 项目 | 数据 |
|------|------|
| 账号名称 | @username |
| 平台 | Instagram |
| 粉丝数 | 125K |
| 内容数 | 856 posts |
| 账号年龄 | 3年 |
| 主要领域 | 美妆/护肤 |
## 📈 核心指标
### 粉丝规模
- 当前粉丝: 125,000
- 30天增长: +2,300 (+1.9%)
- 趋势: 📈 稳定上升
### 互动率
- 平均点赞: 3,200
- 平均评论: 85
- 互动率: **2.6%** (行业平均: 2.1%)
- 评级: ⭐ 良好
### 内容分析
- 发布频率: 4-5篇/周
- 内容垂直度: 高 (90%美妆相关)
- 内容质量: 专业，有原创性
## 🎯 综合评分
| 维度 | 得分 | 权重 | 加权分 |
|------|------|------|--------|
| 粉丝规模 | 7/10 | 15% | 1.05 |
| 互动率 | 8/10 | 30% | 2.40 |
| 内容质量 | 8/10 | 20% | 1.60 |
| 增长趋势 | 7/10 | 15% | 1.05 |
| 商业价值 | 7/10 | 20% | 1.40 |
**总分: 7.5/10 - A级** ⭐⭐⭐⭐
## 💰 商业价值评估
### 预估报价范围
- 单条帖子: ¥8,000 - ¥15,000
- 系列合作(3条): ¥20,000 - ¥35,000
- 长期合作(月度): ¥30,000 - ¥50,000
### 适合品类
- ✅ 护肤品
- ✅ 彩妆
- ✅ 美容仪器
- ⚠️ 服装（相关度一般）
- ❌ 3C数码（不匹配）
### 历史合作案例
- 品牌A: 2024年10月，单条合作
- 品牌B: 2024年8月，系列合作
- 品牌C: 长期合作中
## ✅ 合作建议
### 优势
1. 互动率高于行业平均
2. 内容垂直度高，粉丝精准
3. 发布频率稳定，账号活跃
### 风险点
1. 粉丝量级中等，覆盖面有限
2. 近期有竞品合作，需确认排他期
### 建议
- **合作意向**: 推荐 ✅
- **合作形式**: 单条测试后考虑长期
- **谈判重点**: 争取排他期和二次使用权
---
*报告由 Claude Code influencer-evaluator 技能生成*
*数据来源: 公开信息，仅供参考*
Usage Examples
Example 1: Single Influencer Evaluation
用户: 帮我评估一下 @xiaohongshu_user 这个小红书博主
Claude:
1. 搜索该博主的公开数据
2. 分析粉丝量、互动率、内容质量
3. 计算综合评分
4. 输出评估报告和合作建议
Example 2: Batch Screening
用户: 这是5个候选网红的列表，帮我筛选出最适合美妆品牌的
Claude:
1. 逐一评估5个网红
2. 按评分排序
3. 输出对比表格
4. 推荐Top 3
Example 3: Competitor Influencer Analysis
用户: 分析一下竞品最近合作的几个网红
Claude:
1. 搜索竞品最近的KOL合作案例
2. 分析这些KOL的数据和效果
3. 评估是否值得我们也合作
Evaluation Metrics Reference
Classification by Follower Scale
| Type | Follower Count | Characteristics | Suitable Scenarios |
|---|---|---|---|
| Top-tier KOL | >1 million | Broad coverage, high price | Brand Exposure |
| Mid-tier KOL | 100,000-1,000,000 | High cost-effectiveness | Seeding conversion |
| Tail KOL | 10,000-100,000 | Vertical precision | Word-of-mouth accumulation |
| KOC | <10,000 | Strong authenticity | UGC materials |
Industry engagement rate benchmark
| Industry | YouTube | Xiaohongshu | |
|---|---|---|---|
| Beauty | 2.5% | 4% | 3% |
| Fashion | 2.0% | 3% | 2.5% |
| Food | 3.0% | 5% | 4% |
| Technology | 1.5% | 3% | 2% |
| Maternity and baby | 2.8% | 4% | 3.5% |
Risk identification
Fake follower identification signals
- Abnormal follower growth curve (short-term surge)
- Engagement rate does not match follower count
- Low comment quality (repeated emojis/phrases)
- Abnormal follower profile (geographically concentrated)
Collaboration risk signals
- Frequent category switching
- History of negative public opinion
- Overly dense collaborations with competitors
- Unstable content style
Data storage
- Evaluation report:
~/.claude/cache/influencer/ - File naming:
{YYYYMMDD}-{platform}-{username}.md
Dependent tools
- WebSearch: Search influencer public data
- WebFetch: Retrieve detailed page information
- Write: Save evaluation report
Limitations
- Data comes from public information and may be incomplete
- Unable to obtain precise data from platform backends
- Scores are for reference only and should be judged in combination with actual circumstances
- Data acquisition is limited on some platforms (such as Xiaohongshu)