Metrics

Metrics are the eyes and ears of your rollout. They allow Flagship to determine if a new feature is behaving correctly or causing issues.

Connecting Metrics

Flagship connects to your existing monitoring providers (e.g., Datadog, Prometheus, New Relic) to read metric data. You don't need to send data to Flagship; we query it where it lives.

Types of Metrics

Guardrail Metrics

These are general health indicators that shouldn't degrade during any rollout.

• Error Rate: Percentage of HTTP 500 responses.
• Latency: P99 response time.
• Crash Rate: Application crash frequency.

Feature Metrics

Specific business KPIs related to the feature being rolled out.

• Conversion Rate: Did the new checkout flow improve sales?
• Engagement: Are users clicking the new button?

Thresholds

You define thresholds for each metric.

• Warning: Something looks off. Halt the rollout and notify the owner.
• Critical: Something is broken. Automatically rollback immediately.