Setting up monitoring feels simple until you realize how many things can break independently. Your homepage is up, but your checkout is down. Your server responds, but your SSL expired. Your site loads, but your cron jobs stopped.
This checklist covers everything worth monitoring. You don't need to do all of it on day one—start with the essentials, then work your way down. Each item links to a detailed guide if you want to go deeper.
New to monitoring? Read What is uptime monitoring first for the fundamentals.
Essentials — Set Up Today
If you do nothing else, do these. They catch the most common and most damaging failures.
Basic HTTP check on your main URL. Verifies your site returns 200 OK. Set check interval to 5 minutes or less.
Setup guideMonitor your most important pages beyond the homepage: login page, checkout/pricing, dashboard, API documentation. Your homepage can be up while these are broken.
Set up alerts on channels you actually check. Email is baseline, but add Slack, Discord, or SMS for faster response. Make sure more than one person receives alerts.
Alert optionsTrack expiration dates for all your certificates. Set alerts at 30, 14, 7, and 3 days before expiry. An expired SSL means every visitor sees a security warning.
SSL monitoringHave a status page ready before your first incident. When something breaks, you need somewhere to communicate with users immediately.
Create free status pageImportant — Set Up This Week
These catch issues that basic monitoring misses. Implement them within your first week.
Check your site from multiple global locations. A CDN failure in Asia won't trigger a single-region monitor in the US.
Multi-region setupIf you have an API (public or internal), monitor your critical endpoints with response validation. A 200 status doesn't mean the response is correct.
API monitoring guideTrack when your domains expire. Auto-renewal fails more often than you think. Losing a domain to squatters can cost thousands to recover.
Domain monitoring guideCheck that specific content appears on your pages. Catches scenarios where your page loads but shows an error message, empty state, or database error instead of real content.
Set maximum acceptable response times. A site that takes 10 seconds to load is effectively down for most users. Alert when response times trend upward—it often precedes an outage.
Require 2-3 consecutive failures before alerting. Eliminates false positives from brief network blips and reduces alert fatigue.
Advanced — Set Up This Month
For production systems with real users and revenue. These add layers of protection that catch subtle failures.
Test complete user flows: login → action → verify. Catches auth failures, broken workflows, and state management issues that single-URL checks miss.
Process monitoringMonitor scheduled tasks: backups, email sending, data syncs. These fail silently without heartbeat monitoring. If your cron doesn't ping in, you get alerted.
Heartbeat setupScreenshot comparison detects layout breaks, missing images, defacements, and CSS failures that HTTP checks won't catch. Especially useful for marketing pages and e-commerce product pages.
Visual diffMonitor external services you depend on: payment processors (Stripe, PayPal), auth providers (Auth0, Clerk), email services (SendGrid, Postmark). Their outage becomes your outage.
Monitor non-HTTP services: database ports, mail servers, custom services. If you run services beyond web, verify they're accepting connections.
TCP monitoringDocument who gets alerted, who has access to fix issues, what the escalation path is, and how you communicate externally. Don't figure this out during an incident.
Incident communicationQuick Reference: What to Monitor
| What | Check Type | Interval | Priority |
|---|---|---|---|
| Homepage | HTTP | 1-5 min | 🔴 Essential |
| Login page | HTTP + keyword | 1-5 min | 🔴 Essential |
| Checkout/pricing | HTTP + keyword | 1-5 min | 🔴 Essential |
| API health endpoint | HTTP + JSON validation | 1-3 min | 🟡 Important |
| SSL certificates | SSL check | Daily | 🔴 Essential |
| Domain registration | WHOIS | Daily | 🟡 Important |
| Cron jobs / backups | Heartbeat | Per schedule | 🟢 Advanced |
| Full user flows | Multi-step | 5-15 min | 🟢 Advanced |
| Visual appearance | Visual diff | Hourly/daily | 🟢 Advanced |
| Third-party APIs | HTTP | 5 min | 🟢 Advanced |
| Database port | TCP | 1-5 min | 🟢 Advanced |
Minimum Monitoring by Business Type
Side Project / Blog
- ☐ Homepage HTTP check
- ☐ SSL monitoring
- ☐ Email alerts
3 monitors, 5-minute intervals. Free tier is enough.
SaaS Product
- ☐ Homepage + app + API monitoring
- ☐ Login flow monitoring
- ☐ SSL + domain monitoring
- ☐ Status page
- ☐ Multi-region checks
- ☐ Slack/email alerts
5-10 monitors, 1-3 minute intervals.
SaaS login monitoring guide →E-commerce
- ☐ Homepage + product pages + checkout
- ☐ Cart → checkout flow monitoring
- ☐ SSL monitoring
- ☐ Payment endpoint monitoring
- ☐ Multi-region checks
- ☐ Status page
5-15 monitors, 1-minute intervals on checkout.
Agency (Client Sites)
- ☐ All client homepages
- ☐ SSL monitoring per client
- ☐ Domain monitoring per client
- ☐ Client-facing status pages
- ☐ Separate alert channels per client
Scale with clients. Client views essential.
Client views for agencies →Frequently Asked Questions
How many monitors do I need?
Start with 3-5 covering your most critical pages. Most small-to-medium sites need 5-15 monitors. Enterprise sites may need 50+. Quality matters more than quantity—5 well-configured monitors with keyword validation and multi-region checks beat 50 basic ping monitors.
What check interval should I use?
5-minute intervals for most pages. 1-3 minutes for revenue-critical pages like checkout or API endpoints. Daily checks for SSL and domain expiry. The trade-off is faster detection vs. potential for more false positives.
Should I monitor staging/development environments?
At minimum, monitor SSL certificates on staging to prevent expired certs blocking developer access. Full uptime monitoring on staging is optional but helpful for teams that deploy frequently and want early warning of issues before they reach production.
How do I avoid alert fatigue?
Use confirmation checks (2-3 failures before alerting), multi-region validation, and reasonable thresholds. Reserve phone/SMS for critical issues only. Route informational alerts to email or dashboard. If you're getting more than 2-3 real alerts per week, something needs fixing—not just monitoring.
Start With the Essentials
You don't need to implement everything today. Start with the essentials: homepage monitoring, SSL checks, and alert notifications. Add layers as your site grows and your monitoring needs become clearer.
The worst monitoring setup is no monitoring setup. Even basic monitoring puts you ahead of most websites.