A site can stop opening at any moment: the hosting bill went unpaid, the certificate expired, an update broke the form. The worst case is learning about it from a client, or not learning at all while requests quietly stop coming in.
Monitoring fixes that: a service checks the site for you and messages you as soon as something goes wrong. Below is what to check, where to send alerts, and how not to drown in false alarms.
Why not to wait for complaints
Most visitors never tell you the site is down. They close the tab and go to whoever did load. Meanwhile ads keep spending the budget, and a search crawler may notice the outage before you do. The sooner you learn about a failure, the less it costs.
What to check first
A basic set fits almost any small site:
- The home page and a couple of key pages respond and do not return an error
- Expected text is on the page, to catch a blank or broken page that technically loads
- The SSL certificate expiry date, so browsers do not start scaring visitors with a warning
- Domain and hosting renewal dates
- The form works: a test request reaches where it should
Forms have their own write-up: why site form requests go missing.
Where alerts go
An alert only helps if someone sees it. An email to an inbox opened once a week will not. A channel you already read all day works better: Telegram, a work chat, or a push on your phone. It is good when two people get it: you and a developer who can sort things out quickly.
The second message matters too, the one saying the site is back. Without it you cannot tell whether the outage is over.
How not to drown in false alarms
If monitoring fires on every short network hiccup, people stop reading it within a week. A few simple settings help:
- Alert only after several failed checks in a row, not after one
- Check from different locations, so one provider's problem does not look like your site is down
- Different urgency: an outage right away, a certificate or domain reminder calmly and in advance
- Planned work marked ahead of time, so an update does not raise an alarm
A ready service or your own script
A ready uptime service is usually enough for a small site: many have a free plan, and setup is simple. Your own script or bot makes sense when you need to check something specific, such as the catalog syncing from your accounting system or a Telegram bot answering commands.
Monitoring does not replace backups and protection. More on those in the post on small site security.
Who acts when an alert comes
The most common mistake here is not technical. The alert arrives, but nobody knows what to do with it. Agree in advance who watches alerts, how fast they respond, and who has access to hosting and the domain. This is usually part of support after launch, and it is better discussed before the site breaks.
If you want someone to watch your site or bot and report outages in Telegram, get in touch via the contact page.