Developer guide
The Post-Deployment Monitoring Checklist
You shipped. CI passed. Deployment succeeded. But "deployed" doesn't mean "working." Most user-facing failures happen in the first 30 minutes after a deploy — broken bundles, stale CDN assets, missing environment variables, failed database migrations. This checklist covers what to verify after every deploy, whether you do it manually or automate it.
- Step-by-step checklist for every deploy
- Covers: immediate, 30-minute, and ongoing checks
- How to automate the entire checklist
Build artifacts verified
Asset validation
main.js → 404
API health check
Alert sent
Rollback initiated
First 5 minutes
Immediate post-deploy checks
These checks should happen within 5 minutes of every deploy. Automate them if possible.
1. Verify the deploy actually completed
Check your deployment platform's dashboard or CLI output. A "successful" deploy in CI doesn't always mean the deployment platform received and published the build. Verify the deploy ID or commit hash matches what you expect.
2. Hit your critical pages
Load your most important pages in an incognito browser window (bypasses your local cache):
- Homepage — first impression, highest traffic
- Signup/login — conversion-critical
- Pricing — revenue page
- Checkout/cart — if e-commerce
- API docs — if developer product
3. Check the browser console
Open DevTools → Console. Look for red errors — especially 404 errors on JS/CSS files, CORS errors on API calls, and TypeError from broken imports. A clean console means the page is likely working.
4. Verify API health endpoints
If you have health check endpoints (/api/health, /api/status), hit them and verify the response. Check that database connections, cache connections, and external service integrations all report healthy.
5. Check for asset hash mismatches
View page source. Are the JS/CSS filenames what you expect? If your HTML references main-abc123.js but that file doesn't exist on the CDN, you have a deploy/CDN cache mismatch. This is the #1 cause of "up but broken" failures.
11
Detection rules
5–30 min
Check intervals
23+
Stack playbooks
First 30 minutes
30-minute post-deploy checks
6. Check multi-region availability
Your site works from your location. Does it work from other regions? CDN propagation isn't instant. Use an external checker or multi-region monitoring to verify your deploy reached all edge locations.
7. Monitor error rates
Check your error tracking tool (Sentry, LogRocket, Datadog). Compare error rates in the 30 minutes post-deploy vs. the 30 minutes before. A spike in new error types is a regression signal.
8. Check analytics for anomalies
Compare real-time analytics to the same time yesterday. A sudden drop in pageviews, signups, or conversions within 30 minutes of a deploy is a strong signal that something broke — even if monitoring shows green.
9. Verify third-party integrations
Test any third-party integration that touches the frontend: payment forms (submit a test transaction), auth flows (sign in with Google/GitHub), analytics (verify events fire in GA4/Mixpanel), chat widgets (verify they load).
Ongoing
Ongoing post-deploy monitoring
10. Set up continuous website monitoring
Manual checks don't scale. You won't check at 2 AM. You won't check from Europe. You won't check every page. Automated website monitoring runs these checks on a schedule and after every deploy:
- Every linked JS, CSS, image, and font asset validated
- MIME types checked (a CSS file served as text/plain breaks silently)
- Redirect chains followed and loops detected
- Content fingerprinted for unexpected changes
- Multi-region checks from EU and US
11. Configure deploy hooks
Connect your CI/CD pipeline to your monitoring tool. Deploy hooks trigger instant checks after every deploy — catching regressions in minutes instead of waiting for the next scheduled check.
12. Set up smart alerting
Route alerts to Slack, Discord, or PagerDuty — wherever your on-call team actually looks. Include root cause information in alerts (which asset broke, what the error was) so the responder can act immediately without debugging from scratch.
Automate the checklist
How Sitewatch automates this checklist
Connect your pipeline
Add a Sitewatch deploy hook to your Vercel, Netlify, or GitHub Actions pipeline. One webhook URL, 30 seconds to set up.
Every deploy triggers a full check
Sitewatch fetches your critical pages, validates every asset, checks MIME types, follows redirect chains, and fingerprints content — automatically, after every deploy.
Get alerted on regressions
If anything broke — a 404'd JS bundle, a MIME mismatch, a redirect loop — you get a Slack/email alert with the exact issue, root cause, and stack-specific fix steps.
Deployment monitoring FAQ
Immediately. Most deploy-related failures are visible on the first page load after the deploy publishes — broken asset references, missing environment variables, failed migrations. The issue is detection time: without monitoring, these failures go unnoticed for hours.
Items 1-5 and 10-12 can be fully automated with website monitoring + deploy hooks. Items 6-9 (multi-region, error rates, analytics, third-party integrations) require additional tools — but Sitewatch covers the most critical items: asset validation, MIME checks, and multi-region verification.
The automated checks (deploy hooks) should run on every deploy. The manual checks (browser console, analytics review) are most important for large deploys, breaking changes, or deploys to production for the first time. Use your judgment based on the deploy's risk level.
Automate your post-deploy checks
Free plan available. Deploy hooks included on all plans.