TL;DR: Building a LinkedIn analytics scraper using Playwright/Puppeteer but hitting multiple roadblocks. Need advice from developers who've tackled similar challenges.
What I'm Building
A service that automatically fetches LinkedIn profile analytics (post impressions, followers, profile views, etc.) for users. Think of it as a dashboard that pulls your LinkedIn metrics without manual logging in each time.
Technical Stack
- Browser Automation: Playwright with Chromium
- Deployment: AWS Lambda
- Authentication: Stored cookies + user agents
- Session Management: Persistent browser profiles
Problems I've Hit (and semi-solved)
Cookie Authentication Issues
Stored LinkedIn cookies but still getting login screens
Tried: Session cookies, proper domain/path settings, complete storage state
Partial Fix: Detecting and clicking saved login profilesLinkedIn's Bot Detection
Getting "Try again later" pages even with valid credentials
Tried: Stealth mode, realistic user agents, proper headers, timing delays
Partial Fix: Anti-detection measures + automatic "try again" button clickingPage Loading Issues
Profile pages showing skeleton/placeholder content instead of actual data
Analytics buttons not appearing or clickable
Tried: Various wait strategies, network idle detection, content-specific selectors
Current: 45+ second waits + content validationNavigation Reliability
Me menu → View Profile → Analytics flow was unreliable
Solution: Direct URL navigation to specific profileRate Limiting & Session Persistence
Frequent blocks when running multiple attempts
Tried: Persistent browser directories, proper session state saving
Current Status
Current Status
✅ Login automation working (saved profiles)
✅ Profile navigation working (direct URLs)
❌ Still struggling: Page content loading + Analytics button detection
❌ Main blocker: LinkedIn's aggressive bot detection
Questions for the Community
- Has anyone successfully automated LinkedIn analytics collection? What approach worked?
- Better alternatives to cookie-based auth? OAuth feels overkill for personal analytics.
- LinkedIn-specific anti-detection techniques? Beyond standard stealth measures.
- Page loading detection strategies? How do you know when LinkedIn's dynamic content is actually ready?
Top comments (0)