Datadog has launched a Feature Flags product that links feature releases to observability data across its application performance monitoring and real user monitoring tools.
The company positions the release as a way for engineering teams to connect feature management with production telemetry. It says teams can identify which feature change correlates with a spike in errors or latency and act on it during a rollout.
Feature flags have become a common tool for software teams that release code frequently. Teams use them to gate functionality, run canary releases, and control exposure by user segment or environment. Datadog said many organisations still manage flags separately from monitoring, which leaves teams to correlate changes and incidents by hand.
Datadog said this separation creates blind spots during deployments. It also said organisations often carry stale flags in codebases and configuration, which can add technical debt and complicate future releases.
Product Scope
Datadog said Feature Flags is generally available. The company said it integrates natively across Datadog APM and Datadog RUM.
The product associates each flag with telemetry collected in Datadog. Datadog said teams can view the impact of a feature on performance and reliability in a single place, rather than switching between tools.
Datadog also said the product can automate rollouts and rollbacks. It cited canary releases, circuit breakers, and instant rollbacks based on service health signals. Datadog said teams can run these actions without manual intervention or custom scripts.
The company described the product as a way to adjust system behaviour without a redeploy. It said teams can apply guardrails across environments while they run experiments.
Incident Response
Datadog framed the release as a response to the operational risk that comes with frequent software delivery. It said engineers often watch dashboards during a rollout and ramp traffic slowly. It also said teams can struggle to determine whether an incident stems from a feature change, a configuration change, or unrelated service behaviour.
The company said these conditions can extend release cycles or increase risk when teams push changes without full confidence. It also said coordination of rollouts and rollbacks across distributed systems remains difficult, particularly when teams rely on scripts and manual oversight.
Datadog said Feature Flags connects feature changes to production signals in real time. It said teams can trace reliability issues back to the specific feature or configuration responsible.
Toolchain Links
Datadog also positioned the product alongside its existing software delivery tools. It said Feature Flags complements Datadog's CI/CD visibility and test optimisation products and extends observability into release management.
Datadog referenced its acquisition of Eppo in connection with the product's development.
"Releasing new features is one of the riskiest parts of modern software delivery, and releasing frequently is even more important in today's AI-driven development age," said Yanbing Li, Chief Product Officer, Datadog. "Datadog Feature Flags, created with a head start after our acquisition of Eppo, allows development teams to automatically detect regressions, enforce reliability guardrails, and ship updates faster and more safely by tying every flag to real-time telemetry."
Flag Cleanup
Datadog also highlighted flag lifecycle management. The company said Feature Flags can identify unused flags and assist with removing "dead paths" from codebases.
It pointed to integrations with Bits AI and MCP. Datadog said these integrations can identify unused flags and generate pull requests for removal.
The company said it views stale flags as a source of technical debt. It linked this to the operational complexity that emerges when teams manage multiple environments and distributed services.
Datadog said Feature Flags is available now as part of its product portfolio for cloud application observability and security.