A customer fills in your contact form on Tuesday and hears nothing. On Friday they call a competitor. Nobody on your side knew the form had stopped delivering messages after a small update the week before. The website was up the whole time.
This is the usual pattern. Total outages are noticed quickly. Partial failures, such as a form that sends nothing, an expired certificate or a page that loads with an error, can run for days. A simple weekly routine and a few free or cheap alerts are enough to catch most of them.
The three tests
Keep the list short, so it actually gets done.
Test 1: Does the site load?
Open the home page and two or three pages that matter for business, such as the services page and the contact page. Check that they load, look right and show no error message. An external monitor that requests the home page every five minutes and sends an alert when it fails costs little or nothing, and it also runs at night.
Test 2: Does the form arrive?
Fill in the contact form with a test message, marked clearly as a test, and confirm that it reaches the inbox of the person who should answer. Check the spam folder too. Do this weekly and after every change to the site, the mail settings or the hosting. If you have more than one form, such as a quote request or a newsletter signup, rotate them so each one is tested at least once a month.
Test 3: Is the certificate and the domain valid?
Look at the padlock in the browser and note the certificate expiry date. Note also the renewal date of the domain. Both should be at least thirty days away. Put the dates in a shared calendar, so a renewal does not depend on one person's memory.
Who gets the alerts
An alert sent to a shared mailbox that nobody reads is the same as no alert. Set it up like this:
- One named recipient who looks at it the same day, with a phone notification if possible.
- One named deputy who gets it if the first person does not react within an hour during working hours.
- No distribution lists as the only recipient. Everybody assumes someone else is looking.
Write the two names on the one-page record described below, and update it when someone changes role or leaves.
What counts as an incident
Without a definition, every small hiccup becomes a debate. Agree on three levels:
| Level | Example | Reaction |
|---|---|---|
| Minor | One page is slow or shows a layout error | Note it and fix it in the weekly slot |
| Incident | The site is down for more than ten minutes, or the form does not deliver, or the certificate has expired | The alert recipient acts at once and informs the person who owns the site |
| Serious | Customers cannot order or reach you, or the site shows wrong or unexpected content | Owner decides on a customer notice and calls the hosting provider or developer |
Adjust the numbers to your business. A shop that sells at night has a shorter limit than an office that serves customers from nine to five.
The weekly check in fifteen minutes
Choose a fixed day and time, for example Monday at nine, and repeat the same steps:
- Open the home page and the key pages. Note anything strange.
- Send a test message through each form on the rotation and confirm that it arrived.
- Check the certificate and domain dates.
- Look at the monitor's report for the last seven days and read any alerts that fired.
- Write one line in the log: date, who checked, result.
If the log shows a month of "all fine", you have evidence that the routine works. If it shows gaps, the gaps are the real finding.
What to do when a test fails
Do not fix and forget. For each incident, write down when it started, how it was noticed, what was changed and what stopped it. If the cause was a change to the site, connect it to your process for website change requests, so that the form test becomes part of the release. For repeating or serious incidents, use a short post-incident review to agree on one fix and one owner.
Common mistakes
- Monitoring only the home page. It can load perfectly while the form is dead.
- Testing the form once, at launch. Mail settings and updates change things later.
- Alerts to a team address. Name a person.
- No test label. Mark test messages clearly so they are not answered as real enquiries.
- Forgetting the renewals. An expired domain or certificate is avoidable and embarrassing.
How to start
- Set up a free external monitor for the home page and one key page, with alerts to one named person.
- Send a test form message today and confirm where it lands.
- Put the certificate and domain dates in a shared calendar.
- Write the one-page record: monitored pages, forms, recipient, deputy, incident levels.
- Book the fifteen-minute weekly slot for the next four weeks and review the log at the end.
Summary
Website monitoring for a small company does not need special tools. Three tests, one named alert recipient and a clear line between a minor issue and an incident keep the site and the form working. A fifteen-minute weekly check finds problems before customers do, and the log shows whether the routine is really being followed.
Sources
- Our related articles linked above.
- The routine is common practice for small companies and is offered as a practical guide. It is not a service level agreement or a security audit. No client results or market figures are claimed in this article.
