ktg-plugin-marketplace/plugins/linkedin-thought-leadership/skills/linkedin-analytics/SKILL.md

name

description

linkedin-analytics

Analyze LinkedIn content performance, import analytics data, generate reports, audit content strategy, and troubleshoot engagement issues. Covers the full analytics lifecycle from CSV import to actionable insights and recovery protocols. This skill should be used when the user wants to analyze post performance, import LinkedIn data exports, generate weekly reports, audit their content strategy, troubleshoot reach drops, or understand what's working and what isn't. Triggers on: "analyze my posts", "LinkedIn analytics", "weekly report", "import CSV", "what's working", "content audit", "why is my reach down", "LinkedIn troubleshooting", "performance report", "quarterly review", "engagement trends", "audit my linkedin", "low reach", "content not working", "reach dropped", "A/B test", "test different hooks", "compare post versions", "experiment with content".

Analytics Domain

This skill covers everything related to LinkedIn analytics, performance measurement, reporting, troubleshooting, and content strategy auditing.

---

Commands

Command	Purpose	When to Use
/linkedin:analyze	Content/performance analysis	When content isn't performing
/linkedin:audit	Periodic content strategy audit	Quarterly reviews
/linkedin:import	Import CSV with auto-detect, quick-import, feedback loop	After downloading LinkedIn data or to start fresh import
/linkedin:report	Generate weekly performance report	Weekly check-ins
/linkedin:ab-test	Design and manage content A/B tests	Testing content variations

Agents

Agent	Model	Responsibility
analytics-interpreter	Sonnet	Audience pattern analysis from structured data
performance-reporter	Sonnet	Weekly/monthly performance reports from JSON
trend-spotter	Sonnet	Trending topics + opportunity scores
content-tracker	Haiku	Plan vs. published tracking
post-feedback-monitor	Haiku	Post-publish 48h monitoring, anomaly detection

---

Analytics System

Architecture

Node.js CLI tool for parsing LinkedIn CSV exports into structured JSON.

Location: scripts/analytics/ Data: assets/analytics/ (gitignored -- personal performance data)

CLI usage:

ANALYTICS_ROOT="${CLAUDE_PLUGIN_ROOT}/assets/analytics" node --import tsx "${CLAUDE_PLUGIN_ROOT}/scripts/analytics/src/cli.ts" import <file>
ANALYTICS_ROOT="${CLAUDE_PLUGIN_ROOT}/assets/analytics" node --import tsx "${CLAUDE_PLUGIN_ROOT}/scripts/analytics/src/cli.ts" report --week <W>
ANALYTICS_ROOT="${CLAUDE_PLUGIN_ROOT}/assets/analytics" node --import tsx "${CLAUDE_PLUGIN_ROOT}/scripts/analytics/src/cli.ts" trends --period <P> --metric <M>

Storage structure:

assets/analytics/
+-- exports/           # Raw CSV from LinkedIn (drop files here)
+-- posts/             # Imported post data as JSON
+-- weekly-reports/    # Generated weekly reports
+-- README.md          # Import instructions

Import Workflow Enhancements

The /linkedin:import command reduces friction with:

1. Auto-detect from ~/Downloads — Scans for recent LinkedIn CSV files in ~/Downloads before asking users to manually move files
2. Quick-import helper — hooks/scripts/quick-import.mjs opens LinkedIn Analytics in the browser and watches ~/Downloads for new CSV files (auto-copies on detection)
3. Analytics-to-strategy feedback loop — After import, automatically runs content pillar performance, format analysis, posting time analysis, and generates 3 data-driven recommendations
4. State tracking — Updates last_import_date and last_import_week in state file. Session-start and idle-prompt hooks show import staleness warnings.

Data Freshness

The plugin tracks import freshness via last_import_date in the state file:

Staleness	Status line	Reminders
Never imported	Import: never	"No analytics data imported yet"
<7 days	Import: Xd ago	None
7-13 days	Import: Xd ago	"Consider /linkedin:import for fresh data"
14+ days	Import: Xd ago	"Analytics data is stale. Strategy recommendations may be inaccurate"

The strategy-advisor agent adds data confidence caveats based on import freshness.

---

Troubleshooting: Performance Issues

Quick Decision Tree

Is your reach down?
|
+-- Down <25%? --> Normal fluctuation, continue posting
+-- Down 25-50%? --> Review last week's posts, increase engagement
+-- Down 50-75%? --> Start 14-day recovery protocol
+-- Down 75%+? --> Full profile audit, check for violations

Common Causes

1. 360Brew profile mismatch -- Topic doesn't align with profile (see references/algorithm-signals-reference.md)

2. External link penalty -- Link in post body causes 40-50% reach suppression

3. Engagement drop -- Not doing pre-post engagement (5x5x5 method)

4. Topic inconsistency -- Posting outside your established expertise areas

5. Format fatigue -- Same format every post, audience stops engaging

6. Stale analytics data — Strategy advice based on old data; run /linkedin:import for fresh data

For complete troubleshooting and recovery protocols, see references/troubleshooting-guide.md.

---

Analytics Tools for Finding YOUR Edge

The mechanics in the main skill represent baseline knowledge. Your edge comes from discovering what works specifically for YOUR audience.

Essential free tools:

1. LinkedIn Native Analytics - Track weekly (15 min)
2. Google Trends + Exploding Topics - Catch emerging topics
3. Reddit + Niche Communities - Find real problems
4. Personal Knowledge System - Connect non-obvious dots
5. Structured Experimentation - Test hypotheses

For detailed tool guidance, see references/analytics-tools-guide.md.

---

Content Audit Framework

What to Review (Quarterly)

1. Top performers -- What topics, formats, hooks worked best?
2. Bottom performers -- What fell flat? Why?
3. Topic distribution -- Are you too narrow or too broad?
4. Format mix -- Matching 70/20/10 rule?
5. Engagement quality -- Comments vs. reactions ratio?
6. Profile optimization -- Still aligned with content themes?

Key Metrics

Metric	Good	Great	Exceptional
Impression rate (% of followers)	10-20%	20-40%	40%+
Engagement rate	2-4%	4-8%	8%+
Comment-to-reaction ratio	1:10	1:5	1:3
Profile views per post	5-15	15-30	30+

---

Common Patterns

User: "Why isn't my content getting engagement?"

1. Review recent posts for patterns
2. Check for algorithm penalty triggers
3. Assess first-hour engagement strategy
4. Evaluate topic consistency
5. Provide specific fixes from troubleshooting guide

User: "What's working for me?"

1. Import or review analytics data
2. Identify top-performing posts and patterns
3. Map patterns to content matrix
4. Recommend doubling down on winning formulas

---

Reference Files

File	When to Read
references/algorithm-signals-reference.md	Understanding algorithm behavior
references/troubleshooting-guide.md	When reach drops
references/analytics-tools-guide.md	Tool recommendations
references/ab-testing-framework.md	A/B testing methodology and tracking
references/linkedin-growth-playbook-2025-2026.md	Engagement benchmarks
assets/checklists/quality-scorecard.md	Quality standards reference