Sv Analytics Review

SKILL.md

---
name: Sv Analytics Review
version: 1.0.0
tags: [short-video, content-analytics, performance-metrics, data-driven-creation]
---
# Short Video Performance Analytics & Review

Provides a structured framework for reviewing short video performance data — metrics interpretation, insight extraction, and iterative improvement planning.

## Target Users

- Content creators
- Social media managers
- Marketing analysts
- Agency strategists

## When to Use

- Weekly/monthly content performance review
- Diagnosing why a video underperformed
- Identifying patterns across top-performing content
- Building a data-informed content improvement loop

## Core Workflow

1. Metrics hierarchy definition
2. Performance dashboard template
3. Single-video deep-dive framework
4. Comparative analysis (video vs. video, series vs. series, platform vs. platform)
5. Insight-to-action translation
6. Iteration planning

## Inputs

- Performance data (manual entry)
- Video catalog
- Platform
- Review period
- Goals/KPIs

## Expected Outputs

- Review report template
- Single-video analysis framework
- Comparative analysis matrix
- Action plan for next content cycle

## Example Prompts

- "Help me create a weekly performance review framework for my 5 Douyin videos."
- "My latest video has high views but low completion rate — help me diagnose why."
- "Compare my top 3 and bottom 3 videos across the last month and suggest improvements."

## Trigger Keywords

video analytics, performance review, content review, video metrics, analytics framework, improve video


## Usage Scenarios

### Scenario 1

**User Input:** "Paste my last 10 TikTok post analytics. Which ones performed best and why?"

**Expected Output:** Performance ranking with correlation analysis: videos with hook in first 1.5s had 2.3× higher completion rate. Videos with text overlay outperformed by 40%. Top 3 videos dissected for replicable elements.

### Scenario 2

**User Input:** "Compare my Reels vs. TikTok performance for the same 5 videos cross-posted last month."

**Expected Output:** Side-by-side metrics: views, engagement rate, average watch time, follower conversion. Platform-specific audience behavior patterns highlighted.

### Scenario 3

**User Input:** "Based on my analytics, what's the optimal posting time, video length, and content type for my audience?"

**Expected Output:** Data-backed recommendation: Tues/Thurs 7 PM, 25-35 seconds, "how-to" format. Confidence scores and supporting evidence from the user's own data.


### Scenario 4: 视频发出去没人看
**User input:** "我在B站和抖音都发视频，播放量一直在500-1000徘徊。怎么分析数据找到问题出在哪？"
**Expected output:** 短视频数据诊断流程——第一步：看完播率（B站看平均播放时长/视频时长，抖音看2秒完播率，低于30%说明开头不够抓人）；第二步：看互动率（点赞/评论/转发/收藏各是多少，总互动/播放量<5%说明内容不够引发讨论）；第三步：看观众画像（在B站/抖音创作者中心看观众年龄段/性别/地域，和你的目标受众是否一致）；第四步：对比爆款（找到同类账号里播放量高的视频，对比你的标题/封面/开头/节奏/配乐差异）；第五步：做A/B测试（同选题做两种不同的开头/封面/标题，发在不同平台测试）。关键工具：B站创作中心+抖音创作者后台+火烧云数据。

## Safety & Limitations

Analytics framework is process guidance. Does not access, pull, or process platform analytics APIs. All data must be entered manually by the user. Insights are suggestions, not guaranteed improvements. Respect user data privacy.

---

*Generated for project short-video-skills-2026-04-27*