# Experiments (/docs/experiments) An experiment tests whether a change moves a metric. Describe what you want to test, review the plan Datahop drafts, and put the variant in your code. Once it's live, results update continuously and tell you: - the **chance each variant is best**; - the **lift** over control, with a range it very likely falls in; - a **verdict**: winner, keep control, inconclusive or needs more data; - whether the **split looks healthy**, **how much longer** to run, and a one-line **recommendation**; - the same numbers **per cohort**, such as new users vs power users. You can check results as often as you like. Looking early doesn't invalidate them. ```text title="What you and your agent read" Experiment "Annual plan discount banner" (running). Exposures: control 1475, test 1525. Primary: Upgraded — control: 30.8% (n=1475); test: 39.1% (n=1525) lift +26.9% [15.0%, 39.8%], P(beats control) 100.0%. Verdict: winner (test). Segment "New users": Upgraded — … lift +61.2% [27.5%, 102.8%], P(beats control) 100.0%. Verdict: winner (test). Recommendation: Ship "test": it is the best variant with 100.0% probability. ``` ## Lifecycle [#lifecycle] | Status | What happens | | --- | --- | | **Draft** | The experiment and its flag exist, but the flag is off, so nothing changes for users. | | **Running** | Launch turns the flag on. Exposures and metrics count from here. | | **Stopped** | No new people enter. People already in keep counting until their metric window closes. | | **Concluded** | The decision is recorded: the winner, the reason, and who decided. Optionally the winner is rolled out to everyone. | Every step shows up in the activity log with the person or agent who took it. Experiments need the [Feature flags](/docs/feature-flags) product. Describe it, review the plan, put the variant in your code. Verdicts, lift, segments and concluding.