Monitor Website Performance

Get alerts when your site goes down or slows down, before customers complain.

Performance monitoring watches your site 24/7 and alerts you when something’s wrong. Without it, you find out about downtime from angry customers—or worse, you don’t find out at all.

This guide covers: Uptime monitoring, speed monitoring, and real user metrics.

The Three Types of Monitoring

Start with uptime monitoring—it’s free and takes 5 minutes.

Uptime Monitoring (Free)

These services ping your site every few minutes and alert you if it doesn’t respond.

Free Options

• UptimeRobot – Free for 50 monitors, 5-minute intervals
• Freshping – Free for 50 monitors, 1-minute intervals
• Better Uptime – Free tier available

Setup (UptimeRobot Example)

1. Sign up at uptimerobot.com
2. Click “Add New Monitor”
3. Type: HTTP(s)
4. URL: Your homepage URL
5. Monitoring interval: 5 minutes (free) or 1 minute (paid)
6. Add alert contacts (email, Slack, SMS)

Takes 2 minutes. Now you’ll know within 5 minutes if your site goes down.

Speed Monitoring

Uptime tells you if the site is accessible. Speed monitoring tells you if it’s usably fast.

Free Scheduled Testing

• Pingdom – Free basic plan, tests from one location
• GTmetrix – Free daily monitoring with alerts

Setup (GTmetrix Example)

1. Create free account at gtmetrix.com
2. Add your URL to monitoring
3. Set alert threshold (e.g., alert if page load > 5 seconds)
4. Configure email alerts

What to Monitor

Don’t just monitor your homepage. Add monitors for:

• Homepage – Basic availability
• Key landing page – Where paid traffic goes
• Checkout/cart – Most important for e-commerce
• Contact/lead form – Where conversions happen

What Metrics Matter

• Uptime % – Target 99.9% or higher (8.7 hours downtime/year max)
• Response time – Time to first byte, target under 200ms
• Page load time – Full page load, target under 3 seconds
• Core Web Vitals – LCP, CLS, INP (Google ranking factors)

What to Do When Alerts Fire

1. Verify it’s real – Check from your phone/another network
2. Check your host – Is there a known outage?
3. Check recent changes – Did you deploy something?
4. Contact support – If it’s not on your end

See Site Down? Diagnose Why for detailed troubleshooting.

Tuning Alert Thresholds

Default alert settings create noise. A single failed ping does not mean your site is down—it means one monitoring server could not reach you for one check. Configure alerts to require 2-3 consecutive failures before firing. For speed monitoring, set thresholds based on your baseline, not arbitrary numbers. If your homepage normally loads in 1.8 seconds, alert at 4 seconds, not 3. Too-tight thresholds train you to ignore alerts, which defeats the purpose.

When monitoring becomes noise: If you are getting more than 2-3 alerts per week on a stable site, your thresholds are wrong. Every alert should require action. If you find yourself dismissing alerts without investigating, either raise the threshold or remove that monitor. Alert fatigue is the #1 reason monitoring fails in practice—teams stop checking because the signal-to-noise ratio is too low.

Escalation Paths

An alert is useless if no one acts on it. Define who gets notified and when:

• First alert: Email to the primary site owner or developer
• 5 minutes still down: SMS or Slack notification to the same person
• 15 minutes still down: Notify your hosting provider support and a backup contact
• 30+ minutes: Escalate to management; post a status update if you have a status page

Write this escalation path down and share it with your team. During an actual outage, people panic and forget the process. A documented checklist eliminates that failure mode.

Sources

• UptimeRobot
• Freshping
• Better Uptime
• Pingdom
• GTmetrix