Website Uptime Explained:What It Means and How to Monitor It
What uptime percentages mean in real hours of downtime, the most common causes of outages, and how to set up free monitoring that alerts you before customers do.
Table of Contents
Website uptime is the share of time a website is available and working. It is usually written as a percentage over a month or year: 99.9% uptime still allows about 8 hours 46 minutes of downtime per year, while 99.99% allows about 53 minutes. Monitoring is how you find out about downtime, ideally before your customers do.
This guide explains what the percentages mean, why sites go down, and how to set up monitoring that is actually useful.
What uptime percentages mean
| Uptime | Downtime per year | Downtime per month (30 days) |
|---|---|---|
| 99% | about 3 days 15 hours | about 7 hours 12 minutes |
| 99.5% | about 1 day 19 hours | about 3 hours 36 minutes |
| 99.9% | about 8 hours 46 minutes | about 43 minutes |
| 99.95% | about 4 hours 23 minutes | about 22 minutes |
| 99.99% | about 53 minutes | about 4 minutes 19 seconds |
When a provider quotes an uptime figure, read how it is defined: what counts as downtime, whether planned maintenance is excluded, how it is measured, and what compensation (if any) applies. A figure on its own says little about how a service behaves in practice.
Common causes of downtime
- Server or network problems at the hosting provider.
- Resource limits on shared hosting, which make a site return errors under load. See hosting resource limits explained.
- Expired domains or SSL certificates: a site can be "up" but unreachable or showing a security warning. See what happens when a domain expires.
- DNS mistakes, such as deleted records or wrong nameservers.
- Broken updates: a plugin, theme or code deploy that causes fatal errors.
- Database problems: crashed tables or a database server that is overloaded.
- Attacks and abuse: brute-force floods, DDoS or a compromised site suspended by the host.
Most of these are not the host's fault, which is why your own monitoring matters.
What to monitor
A basic setup covers more than "does the home page load":
- HTTP check of the home page every one to five minutes, alerting on non-200 responses or timeouts.
- Keyword check: confirm a word that only appears on a working page, because some failures return a 200 status with an error message.
- A key transaction page, such as checkout or the contact form page.
- SSL certificate expiry, with a warning a few weeks ahead.
- Domain expiry, with warnings well ahead of the renewal date.
- DNS resolution of your main records, for larger setups.
Setting up monitoring
Many external monitoring services offer free plans with checks every few minutes. Whichever you choose:
- Monitor from outside your hosting, so the monitor does not go down with the site.
- Use more than one check location to avoid false alarms caused by one network path.
- Send alerts where someone will see them: email plus a phone notification or chat message.
- Set a sensible threshold, for example two failed checks in a row, to avoid alert fatigue.
- Keep a log of incidents with start time, end time and cause.
What to do when the site goes down
- Confirm it is really down from another network or a monitoring tool.
- Check the obvious: domain and SSL expiry, DNS records, recent changes.
- Check the hosting account: error logs, resource usage, disk space.
- Roll back the last change if the outage started right after an update.
- Contact your host with the exact time, the URL and the error message.
The errors you are most likely to see are explained in common hosting errors: 500, 503, 508 and 403.
Example monitoring setup for a small business
| Check | Interval | Alert after | Sent to |
|---|---|---|---|
| Home page returns 200 and contains the business name | 1 minute | 2 failures | Owner's phone + developer's email |
| Contact page loads | 5 minutes | 2 failures | Developer's email |
| Checkout page loads (for shops) | 2 minutes | 2 failures | Owner + developer |
| SSL certificate expiry | Daily | Under 14 days | Developer |
| Domain expiry | Daily | Under 30 days | Owner |
| Monthly uptime report | Monthly | — | Owner |
The keyword check (business name on the page) catches cases where the server returns a 200 status but shows an error, a blank page or a hacked page.
Calculating your own uptime
If your monitor recorded 52 minutes of downtime in a 30-day month (43,200 minutes):
uptime = (43,200 − 52) ÷ 43,200 × 100 ≈ 99.88%
Tracking this monthly, along with the cause of each outage, shows whether problems come from hosting, your own changes or external services, which is far more useful than any headline number.
Status pages
If customers depend on your website or app, a public status page (hosted separately from the site) lets you tell them what is happening during an incident. It reduces repeated support questions and builds trust when something goes wrong.
Frequently Asked Questions
How often should uptime be checked?
Every one to five minutes is typical. More frequent checks catch short outages but produce more noise.
Why does my monitor say the site is down when it loads for me?
The monitor may check from a different region, hit a firewall rule, or detect a problem your browser hides because of caching. Check the monitor's error details and test from another network.
Does 100% uptime exist?
In practice, no service can promise zero downtime. Planned maintenance, network events and software faults all happen; what matters is how quickly problems are detected and fixed.
Related reading
Uptime is one part of a wider plan for keeping a business online; see business continuity for websites. For hosting fundamentals, read what is web hosting. If you need help investigating an outage on ServerNeed hosting, contact support.
Sources
Last updated 7 October 2026
