Uptime Monitoring for Small Sites: Know Before Your Customers Do

The problem: you find out last

Every website goes down eventually. For a small business the damage is rarely the outage itself; it is the days of outage nobody noticed. The site falls over on Friday evening, the owner hears about it from a customer on Tuesday, and every visitor in between met an error page and drew their own conclusions.

The causes are rarely dramatic. Shared hosting maintenance overruns, a plugin update takes the site down at 2am, an SSL certificate quietly expires and browsers start showing security warnings, a domain renewal card fails, or the contact form stops delivering while the site itself looks perfectly healthy. None of these announce themselves. Monitoring exists to make them announce themselves, to your phone, within minutes.

What to monitor (it is more than up or down)

A simple ping to your homepage misses most real-world failures. A useful monitoring stack for a small site watches six things:

• HTTP uptime on the homepage and one deep page, such as your contact or checkout page
• A keyword check: the monitor confirms a string that only appears when the page renders properly, like your footer phone number, so a database error page does not count as "up"
• SSL certificate expiry, with warnings at 14 and 7 days
• Domain expiry, because a lapsed domain is the slowest disaster to reverse
• Form delivery, via a scheduled test submission each month that you confirm arrives
• Background jobs such as backups, using a dead-man's-switch service that alerts when a job fails to check in

Free and cheap tools that do the job

You do not need enterprise tooling. Four services cover almost every small-site need, and each has a genuinely usable free tier:

• UptimeRobot: the default choice. The free plan checks sites at five-minute intervals, supports keyword monitors, and covers more monitors than most small businesses will ever need.
• StatusCake: a strong free alternative with uptime, SSL and domain-expiry checks in one place.
• Better Stack: polished incident views and public status pages; paid tiers add phone-call escalation if you want to be woken up properly.
• Healthchecks.io: free dead-man's-switch monitoring for scheduled jobs like nightly backups; if the job does not ping, you get told.

Your host's own status page is worth bookmarking too, but never rely on it alone: it reports their infrastructure, not your site, and certainly not your forms.

The 30-minute setup

Here is the whole job, start to finish:

• 1. Create a free UptimeRobot or StatusCake account (5 minutes)
• 2. Add an HTTPS monitor for your homepage with a keyword check on text from your footer (5 minutes)
• 3. Add a second monitor for your contact, booking or checkout page (3 minutes)
• 4. Add SSL-expiry alerts at 14 and 7 days, and a domain-expiry monitor (5 minutes)
• 5. Install the provider's mobile app and enable push notifications (5 minutes)
• 6. Add email as a second alert channel and invite a colleague as backup contact (4 minutes)
• 7. Test it: temporarily set a keyword that does not exist on the page and confirm the alert reaches your phone, then correct it (3 minutes)

That final step matters. An alert path you have never tested is a rumour, not a system.

Tune alerts so you actually act on them

The enemy of monitoring is alert fatigue: after the third false alarm, notifications get muted and the whole setup silently dies. A few settings prevent that.

• Require confirmation from a second location or a repeat failure before alerting, so a routing blip does not page you
• Alert on two consecutive failed checks rather than one
• Escalate sensibly: push notification first, then SMS or a phone call if the incident is still open after 15 minutes
• Set maintenance windows when you are deliberately updating the site
• Avoid ultra-aggressive check intervals on shared hosting, where some hosts rate-limit repeated automated requests

Once a month, glance at the response-time graph. A site that has drifted from half a second to three seconds has not failed yet, but it is telling you something about your hosting.

When the alert fires: a five-minute triage

An alert at 9pm is only useful if you know what to do with it. Keep this sequence somewhere findable:

• 1. Confirm it is real: load the site on your phone using mobile data, not your office wifi
• 2. Check your host's status page for a wider incident
• 3. Check the obvious: SSL expiry date, domain status, anything updated in the last 24 hours
• 4. If it is host-side, raise a ticket quoting the exact times your monitor logged
• 5. Note the incident, cause and duration in a simple log; patterns across months justify a hosting move

Half an hour of setup buys you the quiet confidence that if the site breaks at the weekend, your phone knows before your customers do. If you would rather someone else watched the dashboards and handled the fixes, our hosting and maintenance team does exactly that.