Content Performance Analysis logo

Content Performance Analysis

Community
vivy-yi
content-performance-analysis

Use when evaluating individual Xiaohongshu post performance, identifying what makes content succeed or fail, extracting viral content patterns, recognizing underperforming content that needs optimization, or comparing performance across different content types and formats

Overview

Publishervivy-yi
Repositoryxiaohongshu-skills
Skill namecontent-performance-analysis
Stars
502
Forks
76
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by vivy-yi on GitHub. Read the source before you install it.

Installation

Install the Content Performance Analysis AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/vivy-yi/xiaohongshu-skills/tree/main/skills/03-数据分析/content-performance-analysis
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.

Use it in TypingMind

Enable Content Performance Analysis in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Content Performance Analysis on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Content Performance Analysis is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Content Performance Analysis (内容效果分析)

Overview

Content performance analysis is the systematic evaluation of individual posts and overall content portfolio to identify success patterns, understand what resonates with the audience, and make data-driven decisions about content strategy.

When to Use

Use when:

  • Post performance is inconsistent or unpredictable
  • Need to understand why certain content went viral
  • Identifying patterns in top-performing content
  • Recognizing underperforming content that needs improvement
  • Comparing different content formats (carousel vs video vs single image)
  • Deciding which content types to focus on
  • Planning content optimization based on past performance

Do NOT use when:

  • Account has fewer than 5 published posts (insufficient data)
  • Looking for real-time performance during first hours (wait 3-7 days)
  • Analyzing paid advertising performance (use ad analytics tools)

Core Pattern

Before (guessing what works):

❌ "This post should do well, I worked hard on it"
❌ "I don't know why this post went viral, lucky I guess"
❌ "All my content is pretty similar, performance is random"

After (data-driven content insights):

✅ "Top 5 posts all use carousel format with before/after structure"
✅ "Posts with question titles get 2.3x more comments than statement titles"
✅ "Video content underperforms images - shift strategy to graphic content"
✅ "Posts published on Tuesday outperform Sunday by 40%"

3 Analysis Dimensions Framework:

  1. Engagement Data - Likes, comments, shares, saves (audience response)
  2. Growth Data - New followers, profile visits (conversion impact)
  3. Viral Data - Exposure, discovery traffic (reach and algorithm favor)

Quick Reference

MetricWhat It RevealsGood BenchmarkAnalysis Method
Engagement RateContent resonance8-12% average(Likes+Comments+Shares+Saves)÷Views×100
Save RateContent value/reuse3-5% is goodSaves÷Views×100
Comment RateDiscussion spark2-4% averageComments÷Views×100
Follower ConversionContent converts to fans1-3%New Followers÷Views×100
Viral ScoreAlgorithm favorViews÷Followers>10 = viral hit

Implementation

Step 1: Collect Post Performance Data

From Xiaohongshu Creator Center:

  1. Open Creator Center → 内容数据
  2. Select time range (last 30 days recommended)
  3. Export or manually record data for each post:
    • Title
    • Content type (image/video/carousel)
    • Publish date/time
    • Views (浏览量)
    • Likes (点赞数)
    • Comments (评论数)
    • Shares (转发数)
    • Saves (收藏数)
    • New followers gained

From Qiangua Data (recommended for efficiency):

  1. Account analysis → Content performance
  2. Export all posts with metrics to Excel
  3. Sort by different metrics to identify patterns

Step 2: Calculate Key Performance Indicators

For each post, calculate:

Engagement Rate:

Engagement Rate = (Likes + Comments + Shares + Saves) ÷ Views × 100
  • Excellent: >15%
  • Good: 8-15%
  • Average: 5-8%
  • Below Average: <5%

Save Rate (content value):

Save Rate = Saves ÷ Views × 100
  • Excellent: >7%
  • Good: 4-7%
  • Average: 2-4%
  • Low: <2%

Comment Rate (engagement depth):

Comment Rate = Comments ÷ Views × 100
  • Excellent: >5%
  • Good: 3-5%
  • Average: 1-3%
  • Low: <1%

Viral Score:

Viral Score = Views ÷ Follower Count
  • Viral Hit: >10 (reached 10x beyond existing audience)
  • Strong Performance: 5-10
  • Expected Performance: 1-5
  • Underperforming: <1

Step 3: Identify Top and Bottom Performing Content

Top Performers (analyze last 10-20 posts):

  1. Sort by Engagement Rate - Find top 5
  2. Sort by Viral Score - Find top 5
  3. Sort by Save Rate - Find top 5

Bottom Performers:

  1. Sort by Engagement Rate - Find bottom 5
  2. Identify posts with Viral Score <1 (underperformed existing audience)

Step 4: Extract Success Patterns from Top Content

Analyze top 5 posts for common patterns:

Content Format Patterns:

  • Single image vs carousel vs video
  • Carousel slide count (3-5 slides optimal)
  • Video length (under 60 seconds optimal)

Content Structure Patterns:

  • Hook/intro style (question, statement, before/after)
  • Main content organization (list, tutorial, story, comparison)
  • Call-to-action presence and type

Title Patterns:

  • Question vs statement
  • Length (short vs long)
  • Keyword usage
  • Emotional triggers (curiosity, urgency, benefit)

Visual Patterns:

  • Cover design style
  • Color scheme
  • Text overlay presence
  • Face presence vs product-only

Topic Patterns:

  • Content category (educational, entertainment, inspiration)
  • Specific subtopics
  • Target audience segment

Timing Patterns:

  • Day of week
  • Time of day
  • Seasonal relevance

Document findings:

Pattern: Carousel format
- Frequency in top 5: 4/5 posts (80%)
- Average engagement: 14.2%
- Common structure: Before/after transformation

Pattern: Question-based titles
- Frequency in top 5: 3/5 posts (60%)
- Average engagement: 13.8%
- Comment rate: 4.1% (above average)

Step 5: Diagnose Underperforming Content

For bottom 5 posts, analyze:

Content Quality Issues:

  • Low production value (poor images, bad lighting)
  • Unclear value proposition
  • Weak or confusing message
  • Inadequate detail or depth

Content Format Issues:

  • Suboptimal format for topic (e.g., complex topic in single image)
  • Wrong carousel length (too short or too long)
  • Video length issues (too long, boring intro)

Title and Cover Issues:

  • Uncompelling cover (low CTR)
  • Confusing or unappealing title
  • Mismatch between title and content
  • Lack of keywords for search

Targeting Issues:

  • Content too broad/narrow for audience
  • Off-brand or irrelevant to niche
  • Missing audience pain point

Timing Issues:

  • Posted at low-traffic times
  • Competing with major events/holidays
  • Seasonal mismatch

Create improvement plan for each underperforming post type:

Issue: Low CTR on single-image posts
Diagnosis: Covers lack visual hook, text-only, no face
Fix: Add face, use before/after format, bold overlay text
Test: Create 3 new posts with improved covers, measure CTR improvement

Step 6: Build Content Performance Matrix

Create a matrix to visualize performance by content type:

Content Type | Avg Engagement | Avg Viral Score | Post Count | Verdict
-------------|----------------|-----------------|------------|--------
Carousel     | 12.4%          | 6.2             | 8          | ⭐⭐⭐ Primary
Video        | 7.8%           | 3.1             | 5          | ⭐⭐ Secondary
Single Image | 9.2%           | 4.5             | 7          | ⭐⭐ Secondary

Strategy decisions:

  • Primary format (⭐⭐⭐): Use for 60-70% of content
  • Secondary format (⭐⭐): Use for 20-30% of content
  • Deprecated format: Phase out or significantly improve

Step 7: Develop Replicable Content Formulas

Based on top performers, create 3-5 proven content formulas:

Formula 1: Problem-Solution Tutorial

  • Format: 5-slide carousel
  • Title: Question-based ("Struggling with X?")
  • Structure: Problem → Solution → Step-by-step → Result → CTA
  • Performance: 14.2% engagement, 8.3 viral score
  • Best for: Educational content, how-to topics

Formula 2: Before-After Transformation

  • Format: 3-slide carousel
  • Title: Benefit-driven ("How I achieved X in Y days")
  • Structure: Before → Process → After
  • Performance: 15.8% engagement, 12.1 viral score
  • Best for: Transformation topics, results showcase

Formula 3: List-Based Tips

  • Format: 4-6 slide carousel
  • Title: Number-based ("7 tips for X")
  • Structure: Hook → 7 tips (1 per slide) → Summary → CTA
  • Performance: 11.3% engagement, 5.7 viral score
  • Best for: Tips sharing, quick value content

Use formulas to:

  • Plan content calendar
  • Maintain consistency
  • Scale content production
  • Predict performance

Step 8: Track Performance Trends Over Time

Build weekly content performance log:

Week | Posts | Avg Engagement | Top Format | Viral Hits | Insights
-----|-------|----------------|------------|------------|----------
W1   | 4     | 9.2%           | Carousel   | 1          | Carousels outperforming
W2   | 5     | 10.5%          | Video      | 2          | Video improvement working
W3   | 3     | 12.8%          | Carousel   | 1          | Question titles boosting comments

Trend analysis:

  • Engagement rate rising? Content strategy is working
  • Certain formats consistently winning? Double down
  • Viral hits increasing? Algorithm favor improving
  • Specific topics always perform well? Create series

Common Mistakes

MistakeWhy HappensFix
Judging content by views onlyViews are vanity metricFocus on engagement rate and saves - they indicate true resonance
Analyzing posts too soon (first 24 hours)Early data is misleadingWait 3-7 days for post to fully perform before analyzing
Comparing posts with different audience sizesUnfair comparisonUse engagement rate % not raw numbers for fair comparison
Ignoring outlier posts (viral hits)Assume they're flukesDeep analyze viral hits - they contain valuable insights
Changing strategy too frequentlyImpatience with slow growthCollect data from 10+ posts before changing strategy
Focusing on averages onlyAverages hide patternsLook at top/bottom 20% to extract winning/losing patterns
Not tracking by content formatMiss format-specific insightsAlways analyze by format (carousel vs video vs single)
Ignoring save rateSaves predict long-term successHigh save rate = evergreen content that drives ongoing traffic
Over-optimizing for one metricUnbalanced strategyBalance engagement (likes) with value (saves) and growth (followers)
Not documenting success formulasReinventing the wheelCreate 3-5 proven formulas, reuse consistently

Real-World Impact

Case Study: Fashion Account Turnaround

  • Before: 6.8% average engagement, unpredictable performance, no clear content direction
  • Analysis: Discovered carousels with before/after structure averaged 14.5% engagement vs 5.2% for single images
  • Action: Shifted 80% of content to carousel format using proven formulas
  • After 30 days: 11.2% average engagement (+65%), follower growth rate doubled, 3 viral hits
  • Key insight: Content format matters more than production quality

Data-Backed Insights:

  • Accounts that document and reuse content formulas grow 3x faster
  • Posts following proven formats outperform experimental content 70% of the time
  • High-save content (>7% save rate) continues getting views for 30+ days
  • Top-performing content often follows 1 of 5 predictable structures
  • Analyzing bottom performers reveals as much as analyzing top performers

Related Skills

REQUIRED: Use data-analytics (overall data analysis framework) REQUIRED: Use data-metrics-understanding (understand metrics)

Recommended for deeper analysis:

  • traffic-analysis - Analyze which traffic sources drive best content performance
  • content-planning - Apply insights to plan future content
  • viral-strategy - Optimize content for viral potential

Use content-performance-analysis BEFORE:

  • content-planning (align plan with proven content formulas)
  • cover-design (design covers based on top-performing patterns)
  • title-writing (write titles using successful patterns)
  • graphic-content-creation / short-video-production (create content in winning formats)

Skills that provide context:

  • account-positioning (ensure content aligns with positioning)
  • persona-building (maintain consistent voice across content)

Frequently asked questions

What does the Content Performance Analysis AI skill do?

Use when evaluating individual Xiaohongshu post performance, identifying what makes content succeed or fail, extracting viral content patterns, recognizing underperforming content that needs optimization, or comparing performance across different content types and formats

Why use Content Performance Analysis on TypingMind?

Because you install it once and use it with any model. Content Performance Analysis is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Content Performance Analysis in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/vivy-yi/xiaohongshu-skills/tree/main/skills/03-数据分析/content-performance-analysis. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Content Performance Analysis?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Content Performance Analysis?

As many as you like. As long as a model supports skills, you can use Content Performance Analysis with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Content Performance Analysis AI skill free?

It is published on GitHub by vivy-yi. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇