I'm unhappy with Puppeteer for recurring website monitoring

Got it — if you’re unhappy with Puppeteer for recurring website monitoring, there are usually a few common pain points:

• Brittleness: small UI changes break selectors
• Maintenance overhead: lots of retry/debug logic
• Performance: browser automation is heavier than needed
• Scaling: running many monitors gets expensive/fragile
• Anti-bot issues: sites detect or block headless browsers

If your goal is monitoring websites regularly, you may not need full browser automation at all. Better options depend on what you’re monitoring:

If you’re checking for page changes

• Playwright: similar to Puppeteer, but generally more robust and better maintained
• HTTP fetch + diffing: if the content is server-rendered, use fetch/axios and compare HTML/text
• RSS/API/webhooks: best when available

If you need rendered content but want less pain

• Playwright with:
  • resilient selectors
  • trace/screenshot on failure
  • built-in retries
• Browserless / hosted headless browsers to offload infrastructure

If you only care about uptime or keyword presence

• Uptime Kuma
• Changedetection.io
• Visualping
• Distill.io

If you’re monitoring at scale

• Use a headless browser only as a fallback
• Prefer:
  • HTML fetch
  • DOM parsing
  • conditional browser rendering
  • queue-based workers
  • alert deduplication

A practical pattern

1. Try fetch the page
2. Parse content and compare to previous snapshot
3. Only launch a browser if the page is JS-rendered or fetch fails
4. Store normalized text, not raw HTML, to reduce noisy diffs

If you want, I can help you with one of these next:

1. Suggest a better alternative to Puppeteer for your exact use case
2. Design a reliable monitoring architecture
3. Migrate a Puppeteer script to Playwright
4. Replace browser automation with fetch + parsing

If you share what you’re monitoring and what’s going wrong, I can be specific.