Most monitoring tells you something broke. That's the easy half.

The hard half is knowing whether it's your server or the internet between you and it — at 3am, before your customers start writing in.

Why this exists

UptimeGlobe started with a bad night and a useless alert. A site was down, the notification said so, and that was the entire contents of the message. What followed was forty minutes of checking a server that turned out to be fine — the problem was a route between one probe and the machine, and nothing in the alert could have told us that.

Every monitoring tool we tried had the same two gaps.

The alerts said what, never why. "Your site is down" is a fact you can act on only after you have worked out whose fault it is. So every UptimeGlobe check runs from several places at once and the results get compared before you are paged: unreachable from everywhere points at your server, unreachable from some places points at the route in between. The verdict arrives in the alert, not after twenty minutes in a terminal.

The pricing punished being small. Free tiers that forbid commercial use. Per-responder fees that grow the bill each time someone joins the rota. Multi-region checks parked behind an enterprise call. Our free plan allows commercial use, every plan checks from more than one place, and what you pay never depends on how many people get the message.

We use it on ourselves

The status card on our homepage is not a screenshot. It is UptimeGlobe watching UptimeGlobe, drawn from the same public feed any customer gets, with the time of the reading printed beside it. When we are having a bad day the front page says so. A monitoring company that hides its own outages is telling you something about the product.

The promises

Say hello

Questions, feedback, or a story about what your last monitoring tool got wrong? Get in touch — every message is read by someone who works on the product.