What Does 99.9% Uptime Actually Mean? Downtime Per Month, Explained
Uptime is usually sold in "nines" — 99.9%, 99.99%, and so on. The numbers look almost identical, but the downtime they allow is wildly different. Here's what they actually mean.
Uptime is a percentage of time
Uptime is the share of a period during which your service was available. The gap between that percentage and 100% is your allowed downtime. Because it compounds over a month or a year, small-looking differences matter a lot.
How much downtime each "nine" allows
Per month (about 30 days), roughly:
- 99% ("two nines") — about 7 hours 12 minutes of downtime.
- 99.9% ("three nines") — about 43 minutes.
- 99.95% — about 21 minutes.
- 99.99% ("four nines") — about 4 minutes 20 seconds.
- 99.999% ("five nines") — about 26 seconds.
Per year, 99.9% works out to roughly 8 hours 45 minutes of downtime — a full workday offline. Five nines is about 5 minutes a year, which is why it's so hard and expensive to reach.
Why the difference is bigger than it looks
Going from 99.9% to 99.99% isn't a rounding tweak — it's a 10x reduction in allowed downtime, from 43 minutes a month to about 4. Each extra nine cuts your downtime budget by ten. That's why hosting and SLA pricing jumps sharply with every nine: the engineering needed to prevent those minutes gets exponentially harder.
What counts as "down"?
This is where SLAs get slippery. A server that returns a 500 error is technically responding — is that up or down? A page that takes 40 seconds to load is "available" but useless. When you measure your own uptime, decide what down means:
- Non-2xx/3xx status codes count as down.
- A missing keyword on the page counts as down.
- Response time over a threshold counts as degraded.
Planned maintenance is usually excluded — as long as it's announced and suppressed in your monitoring so it doesn't distort the number.
How to measure your real uptime
You can't improve what you don't measure. To know your true figure:
- Check the site on a short interval (every 1–5 minutes) from outside your network.
- Record every result, not just the failures.
- Calculate uptime as successful checks divided by total checks over the period.
- Exclude announced maintenance windows.
SiteIsOnline does this automatically — it stores every check, shows your uptime percentage per day and month, and alerts you when you're burning through your downtime budget.
Start measuring your uptime free, or read the complete guide to uptime monitoring to see how the checks work under the hood.