Introduction to latency

The early-warning system for everything your product depends on.

Your product runs on things you don't control: a CDN, a payments API, a database host, a managed cloud. When one of them breaks, you usually hear it from angry customers long before the provider's status page ever turns red. latency is the solution.

We fuse crowd reports, official status feeds, and probes from every corner of the internet into a single early warning. Not just up or down: the whole truth about your presence on the internet, before anyone else has it.

Start here

• Quick Start: watch your first site in minutes
• Grab the iOS app or see how it works
• Platform primitives: the public APIs
• Interested in what's next: Roadmap

Notice

Some platform features are in the final stages of beta testing. Any pages with Beta are still a work in progress and subject to change.

Who it's for

• You ship something people rely on. Get informed about issues first, reconciled from everywhere at once, when it degrades, and again when it recovers.
• You use a 3rd party service. Stripe/PayPal, a CDN, your database host. We watch them next to you, so their bad day doesn't blindside yours.
• You're just curious. Watch anything on the internet or just poke at any host on the public inspector and see what we see.

Quick questions

What makes latency different, wont you go down with the cloud providers?

Yes, but to an extent, our cloud services are not immune to outages. The difference is we run a copy of Latency in many different datacenters in small units we call Cells to avoid any single point of failure. We've also designed our services to be robust with outages in mind.

Do I need an account?

Not to explore. The public platform primitives are open, no sign-up: DNS, geo, or inspect. An account is where your own monitors, alerts, and incidents live.

Is it open source?

The backend isn't (yet), but our CLI is, under MIT.

Why not just use Datadog or Grafana?

Those are big observability suites for instrumenting your own code. Latency is laser-focused on the network-and-dependency question: is it reachable, is it fast, and is the thing it runs on having a bad day? Need deep app tracing? Reach for a full platform. Need to know your stack is up, and hear about it first? That's us.

Ping and traceroute already exist.

And we use them! The difference is we remember the results, connect the dots across providers, and actually tell you when something changes. That's the part all the other monitors miss.