ENHEIT

payless.host

How to know your site is down before your customers do

Updated Nov 22, 2026

Pick an uptime monitor

You need a service that checks your site from outside your hosting provider's network. Start with a free tier: UptimeRobot, Freshping, or StatusCake all offer monitoring every five minutes at no cost. The interval matters less than having something running.

Sign up, confirm your email, and add your site's URL. Use the full address including https:// — monitors default to checking the homepage unless told otherwise.

What can go wrong: Adding only the domain without the protocol can confuse some monitors into checking http when your site runs on https, or vice versa. Test by visiting the URL exactly as entered.

Set where alerts go

Configure notifications to reach you instantly. Email works, but most monitors also support SMS, Slack, or webhook integrations. Add at least two channels in case one fails.

Test notifications by using the monitor's manual test feature, or deliberately enter a broken URL as a second monitor and watch for the alert. If nothing arrives within the stated check interval plus a few minutes, check spam folders and notification settings.

What can go wrong: Mobile apps often require separate notification permissions. Email alerts can land in promotions tabs. If the first real outage produces no alert, you have defeated the purpose.

Decide what counts as down

Most monitors let you define failure: a timeout, a non-200 status code, or missing page content. The default — any response other than HTTP 200 — catches most problems but can false-alarm on maintenance pages.

For a WordPress site, set the monitor to look for specific text that appears on every page, such as your site name in the footer. For a static site, any unique string from the homepage works. This confirms the page loads and renders.

What can go wrong: Matching text that appears in error pages triggers false positives. Matching text from a dynamic element that sometimes hides — a promotion banner, a login state — causes the same problem. Pick something structural.

Check more than the homepage

Add monitors for your contact form, checkout page, or any URL that generates revenue. A homepage that loads while the payment processor is unreachable is functionally down for anyone trying to buy.

If you are on shared hosting, this also catches the provider's server becoming overloaded in a way that affects some scripts but not static pages. HostGator's Hatchling plan at 3.75 USD per month or Namecheap's shared tiers sit on servers with hundreds of accounts; a neighbour's traffic spike can slow dynamic requests even when the domain resolves.

What can go wrong: Monitoring too many URLs on a free tier hits rate limits. Start with three: homepage, most important transactional page, and one mid-depth content page.

Understand what the monitor cannot see

Uptime monitoring checks whether the server responds. It does not check whether DNS is resolving, whether your CDN is serving stale content, or whether a plugin has broken the page layout while leaving the HTML intact.

For DNS monitoring, add your domain to a separate DNS-specific monitor — many uptime tools include this. For layout and functionality, browser-based monitors exist but cost more and belong in a second phase.

Confirm it caught a real outage

After setup, create a deliberate failure. Rename your .htaccess file if on Apache, or temporarily block port 80 in your firewall if you control the server. Wait for the alert, then restore.

If you are on managed WordPress — DreamHost's DreamPress 1 at 14.99 EUR per month, for example — you may not have access to create a convincing outage. In that case, set up a second monitor pointing at a nonexistent subdomain on your own domain and confirm alerts arrive.

What can go wrong: Skipping this test means discovering on the first real outage that notifications are silenced, or routed to an old email address, or blocked by your provider's firewall.

Set the check interval you can act on

A five-minute check interval means you learn about downtime within five minutes, sometimes ten if the monitor retries before alerting. A one-minute interval on a paid plan means faster alerts but also faster alert fatigue if your host has brief hiccups.

Match the interval to how quickly you can respond. If you check email once an hour, a one-minute monitor changes nothing. If you are on call, a shorter interval justifies the cost.

What can go wrong: Shortening the interval on a flaky host produces constant alerts, all of which resolve before you act. You start ignoring them. At that point the monitor is running but you are not.

The providers in this article