PostHog · Schema

ExperimentParameters

A/B TestingAnalyticsFeature FlagsOpen SourceProduct AnalyticsSession Recording

Properties

Name Type Description
feature_flag_variants array Experiment variants. If not specified, defaults to a 50/50 control/test split.
minimum_detectable_effect number Minimum detectable effect as a percentage. Lower values need more users but catch smaller changes. Suggest 20–30% for most experiments.
rollout_percentage number Overall rollout percentage (0-100). Controls what fraction of all users enter the experiment. Users outside the rollout never see any variant and are excluded from analysis. Default: 100.
View JSON Schema on GitHub

JSON Schema

posthog-experimentparameters-schema.json Raw ↑ 
{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "#/components/schemas/ExperimentParameters",
  "title": "ExperimentParameters",
  "additionalProperties": false,
  "properties": {
    "feature_flag_variants": {
      "default": null,
      "description": "Experiment variants. If not specified, defaults to a 50/50 control/test split.",
      "title": "Feature Flag Variants",
      "items": {
        "$ref": "#/components/schemas/ExperimentVariant"
      },
      "type": "array",
      "nullable": true
    },
    "minimum_detectable_effect": {
      "default": null,
      "description": "Minimum detectable effect as a percentage. Lower values need more users but catch smaller changes. Suggest 20\u201330% for most experiments.",
      "title": "Minimum Detectable Effect",
      "type": "number",
      "nullable": true
    },
    "rollout_percentage": {
      "default": null,
      "description": "Overall rollout percentage (0-100). Controls what fraction of all users enter the experiment. Users outside the rollout never see any variant and are excluded from analysis. Default: 100.",
      "title": "Rollout Percentage",
      "type": "number",
      "nullable": true
    }
  },
  "type": "object"
}