Metrics

One metric definition — for readouts, guardrails, and dashboards.

Send counters, gauges, and distributions through the same pipeline your flags already feed. Every metric is scoped by service and environment, discoverable with autocomplete, and usable as an experiment metric or a guardrail without redefining it.

FreeIncluded from the Free tier.

checkout_conversioncounter · prod
24h184,204 events
p95 latency242 ms
scopesservice · env · variation

Define once. Reuse across the platform.

What you get

Why Foredeck for metrics.

Counters, gauges, distributions

Every metric type you need, including latency distributions with p50/p90/p95/p99 from mergeable sketches.

Scoped by service & env

Reserved service and env dimensions plus arbitrary tags, surfaced uniformly for filtering and discovery.

Discovery & autocomplete

A live catalog of metric names and tag values so the right metric is one keystroke away.

One definition, many uses

The same metric powers an experiment readout, a guardrail, and a chart — no duplicate definitions.

How it works

Three steps, one pipeline.

01Emit

Send metric events via the SDK or the events API.

02Discover

Find it in the catalog, scoped by service and env.

03Use it

Attach it to a flag, a guardrail, or a dashboard.

Get in

Try Metrics free.