Uptime monitoring for SMBs: don't hear about it from your customers

Your website is down and you find out from a customer. There's a better way. Here's how to set up uptime monitoring as an SMB without an IT department — without drowning in false alerts.

Your website goes down. Not because it's been hacked, but because your host is doing maintenance, your payment provider has an outage, or your newsletter was just a little too successful. The question isn't if it happens, but when. And more importantly: how quickly do you find out?

For SMBs without an in-house IT department, this is a blind spot. Large companies have monitoring dashboards and a NOC that pages people in the middle of the night. You have... a customer calling on Monday morning saying "your website isn't working." There's a better way, and it doesn't have to cost much.

So what exactly is uptime monitoring?

Simply put: a service that opens your website every few minutes and checks whether it responds. If the page doesn't load, returns an error code, or is too slow to respond, you get an email or push notification. No rocket science — but incredibly useful.

Most tools do three things:

• Check availability — is your site online?
• Measure response time — how fast does it respond?
• Monitor your SSL certificate — is it about to expire?

That last one is worth its weight in gold. An expired SSL certificate triggers that dreaded "your connection is not private" browser warning, and you'll lose visitors within seconds.

Why you shouldn't leave it to your hosting provider

"Our hosting provider keeps an eye on things." We hear this a lot. And sometimes it's true — most hosts do monitor their servers. But that's not the same as monitoring your website.

A few examples of what a hosting provider won't catch:

• Your contact form is throwing a PHP error, but the page still loads
• Your payment module is broken, but the homepage works fine
• Your DNS is accidentally pointing to the wrong IP address
• Your SSL certificate expires in three days
• A specific page returns a 500 error, while the rest of the site is fine

The server is running, so the hosting provider sees green. Meanwhile, you're losing revenue.

What you should monitor at a minimum

Not everything. That creates noise, and then you start ignoring alerts — which is the worst outcome possible. Start small:

1. The homepage — obviously.
2. A key conversion page — your contact form, webshop checkout, or quote request page.
3. The login screen — if customers or staff need to access it.
4. An API or integration — if you're connected to, say, an accounting package.

Four checks every 5 minutes is more than enough for most SMB websites. Checking every minute sounds thorough, but mainly results in middle-of-the-night alerts for brief hiccups that resolve themselves.

Thresholds: avoiding false alarms

Configure your monitoring so that you only receive an alert if your site fails two or three consecutive checks. A single failed check is usually a network glitch between the monitor and your server. Two or three in a row means: something is genuinely wrong.

The same logic applies to response time. A page that normally loads in 800 ms but suddenly takes 4 seconds is a warning sign. But a one-off spike? Ignore it. Only raise the alarm if the pattern persists.

Who receives the alert?

This is where things often go wrong in practice. The monitoring tool dutifully sends an email to info@ and nobody reads it until Tuesday. Agree on this upfront:

• Who is the first point of contact during office hours?
• What should that person do? (Usually: call the hosting provider — not start tinkering.)
• Who gets notified in the evening and at weekends — and does that person actually want that?
• How do you escalate if the first contact doesn't respond?

Write it down. One page, alongside your offboarding checklist and your incident procedure. Done.

Speed is availability too

A site that takes 12 seconds to load is technically online, but effectively offline as far as your visitor is concerned. Google knows this too, and penalises slow sites in search results. Monitoring load time is therefore just as important as monitoring availability.

Want a snapshot of how fast your site is right now? Run a one-off check with our speed test tool. In just a few seconds, you'll see how your site is performing and where the pain points are — useful before you set up monitoring at all.

What to do when an alert comes in

Panic is a poor adviser. A simple action plan prevents you from making changes at random:

1. Verify the outage — open the site yourself in an incognito window.
2. Check the scope — is it just your site, or are other sites on the same host affected too?
3. Check your hosting provider's status page — they may already have it listed.
4. Call or email your hosting provider or administrator — with a clear description, not just "it's not working."
5. Communicate to customers if the outage lasts more than half an hour — a tweet or LinkedIn post is better than silence.

What it costs

Free to a few euros per month. There are excellent free monitoring services that work perfectly well for one or two websites. Pay a little more and you get SMS alerts, monitoring from multiple countries, and better-looking reports. For the average SMB website that's overkill, but for a webshop with serious turnover it's worth considering.

In closing

Uptime monitoring is one of those things you don't know you're missing until you have it. An hour of downtime on a busy Tuesday morning costs more in missed leads than a full year of monitoring. And it can be set up in an afternoon.

Want help setting up monitoring, or a broader health check on your website? Take a look at our website speed optimisation service or request a website security check. We'll make sure you're not the last to know your site is down.