Feature Experimentation

Fresh

by Amplitude

Feature flag evaluation, variant assignment at the SDK level, frequentist and sequential stat-sig calculationEngineers and growth teams who need to safely roll out, gate, and measure the impact of new features on product metricsFree tier

Server-side feature flag management and A/B/n experiment framework with built-in statistical significance engine, guardrail metrics, and mutual exclusion groups.

Features

Behavioral funnelsall plansRetention analysisall plansBehavioral cohort segmentationpaid onlyUnlimited feature flagsall plansWeb experimentation (visual editor A/B testing)all plansFeature experimentation (server-side A/B with stat sig)paid onlyMulti-armed bandit experimentsenterprise onlySession replayall plansAI Feedback (unstructured text analysis)all plansCausal insightspaid onlyPredictive audiencespaid onlyAutomated insights (AI-driven analysis)paid onlyMCP / AI client integrationspaid onlyReal-time data streamingpaid onlyUnlimited sources and destinationsall plansCentralized RBACenterprise onlyCustom audiences (CDP activation sync)paid onlyIn-app guides and surveyspaid only

More from Amplitude