--- title: "Get experiment results" method: GET path: "/api/v2/experiments/{experiment_id}/results" tags: ["Experiments"] --- # Get experiment results `GET /api/v2/experiments/{experiment_id}/results` ## Request - Base URL: `https://api.datadoghq.com` - URL: `https://api.datadoghq.com/api/v2/experiments/{experiment_id}/results` - Auth: one of: - API key in header `DD-API-KEY` and API key in header `DD-APPLICATION-KEY` - OAuth 2 - Authorization code: `/oauth2/v1/authorize` - Token URL: `/oauth2/v1/token` scopes: `product_analytics_experiments_read` - TypeScript: (a function with typed inputs; its response is checked against the schema) Get an experiment's computed results: per-variant statistical analysis for the latest successful run. A historical run cannot be selected. Unavailable analysis statistics, including `p_value` and `confidence_interval`, are omitted. The `numerator`, `denominator`, and `variant_metric_value` fields can be null when their values are unavailable. Do not treat an omitted or null value as zero. ## Path parameters - `experiment_id` string, uuid, required: The UUID of the experiment. ## Response `200` OK - ExperimentsVariantResultsV2DTOArray: List of variant result resources. - `data` ExperimentsVariantResultsV2DTOData[], required: Resources returned in this response. - `attributes` ExperimentsVariantResultsV2DTODataAttributes: Details of the variant result. - `assignment_count` integer: Number of subjects assigned to this variant. - `experiment_id` string: ID of the experiment associated with this result. - `is_control` boolean: Whether this variant is the experiment's control. - `metrics` ExperimentsVariantResultsV2DTODataAttributesMetricsItems[]: Metrics reported for this variant. - `analyses` ExperimentsVariantResultsV2DTODataAttributesMetricsItemsAnalysesItems[]: Statistical analyses calculated for this metric and variant. - `confidence_interval` ExperimentsConfidenceInterval: Lower and upper bounds of the reported statistical interval. - `lower` number, double: Lower bound of the reported interval. - `upper` number, double: Upper bound of the reported interval. - `confidence_level` number, double: Configured nominal confidence level. Interval bounds can use an adjusted level for multiple testing or hybrid methods. - `evsi` number, double: Expected reduction in minimum expected regret from collecting more sample data. - `expectation_above_zero` number, double: Expected effect when the effect is positive. - `expectation_below_zero` number, double: Expected effect when the effect is negative. - `global_lift` number, double: Estimated lift across the population. Calculated as metric coverage multiplied by the experiment lift. - `global_lift_lower_bound` number, double: Lower bound of the estimated lift across the population. - `global_lift_upper_bound` number, double: Upper bound of the estimated lift across the population. - `is_cuped_adjusted` boolean: Whether CUPED used pre-experiment data to reduce variance in this result. - `is_unreliable` boolean: Whether the statistical result is marked as unreliable. - `lift_type` 'RELATIVE' | 'ABSOLUTE' | 'UNKNOWN': Whether the reported lift is relative or absolute. - `method` 'FIXED_SAMPLE' | 'BAYESIAN' | 'SEQUENTIAL' | 'SEQUENTIAL_FIXED_HYBRID' | 'UNKNOWN': Statistical method used to calculate this result. - `minimum_expected_regret` number, double: The smaller expected opportunity cost of choosing treatment or control. - `p_value` number, double: Probability, under the no-effect hypothesis, of a result at least as extreme as the observed result. - `point_estimate` number, double: Estimated difference between the variant and control for this metric. - `probability_above_zero` number, double: Estimated probability that the effect is greater than zero. - `probability_below_zero` number, double: Estimated probability that the effect is less than zero. - `standard_error` number, double: Estimated uncertainty in the effect estimate. - `unreliable_reason` 'CONTROL_DENOMINATOR_NEAR_ZERO' | 'TREATMENT_DENOMINATOR_NEAR_ZERO' | 'CONTROL_AND_TREATMENT_DENOMINATORS_NEAR_ZERO' | 'CONTROL_MEAN_NEAR_ZERO' | 'ZERO_VARIANCE' | 'UNKNOWN': Reason that the statistical result is marked as unreliable. - `variant_metric_value` number, double, nullable: Metric value calculated for this variant. - `z_score` number, double: Standardized statistic used to compare the observed effect with zero. - `assignment_count` integer: Number of subjects assigned to this variant. - `coverage` number, double: Estimated share of the global metric total from the eligible population if that population received control. The estimate can exceed 1. - `coverage_summary` ExperimentsVariantResultsV2DTODataAttributesMetricsItemsCoverageSummary: Population totals and allocation used to calculate metric coverage. - `control_total` number, double: Total metric value for the control population in the coverage calculation. - `coverage` number, double: Estimated share of the global metric total from the eligible population if that population received control. The estimate can exceed 1. - `coverage_unavailable_reason` string: Reason that metric coverage could not be calculated. - `eligible_population_total` number, double: Estimated metric total for the eligible population if that population received control. - `experiment_total` number, double: Total metric value for the experiment population in the coverage calculation. - `global_metric_total` number, double: Observed metric total across subjects inside and outside the experiment. - `traffic_allocation` number, double: Fraction of eligible traffic allocated to the experiment, weighted by time. - `coverage_unavailable_reason` string: Reason that metric coverage could not be calculated. - `denominator` number, double, nullable: Aggregated denominator value for this metric and variant. - `desired_change` 'METRIC_INCREASES' | 'METRIC_DECREASES' | 'UNKNOWN': Direction of metric change considered desirable. - `metric_id` string: ID of the metric represented by this entry. - `metric_name` string: Display name of the metric represented by this entry. - `numerator` number, double, nullable: Aggregated numerator value for this metric and variant. - `sub_metric_property_name` string: Name of the property used to split this metric result. - `sub_metric_property_value` string: Property value represented by this split metric result. - `variant_key` string: Key that identifies the experiment variant. - `variant_name` string: Display name of the experiment variant. - `id` string, required: ID of the variant result. - `type` 'experiment-variant-results', required: Experiment variant results resource type. - `meta` ExperimentsExperimentResultsV2MetaDTO: Information about when experiment results were updated and whether they are stale. - `is_stale` boolean: Whether the saved results require a refresh or their freshness cannot be confirmed. See stale_reasons for details. - `results_last_updated` string, date-time, nullable: Time when the experiment results were last updated. - `stale_reasons` string[]: Reasons the saved experiment results are stale. ## Other responses - `400`: Malformed experiment ID (not a valid UUID), or the experiment is not configured for analysis. - `401`: Missing or invalid authentication. - `403`: Authenticated caller lacks the product_analytics_experiments_read permission. - `404`: No experiment with this ID exists, or results are not available for it yet. - `409`: The experiment's metrics are incompatible with analysis. - `429`: Too many requests --- [API](https://skmtc.dev/datadog/apis/datadog-api-v2-collection.md) - [All operations](https://skmtc.dev/datadog/apis/datadog-api-v2-collection/llms.txt) - [OpenAPI document](https://skmtc.dev/datadog/apis/datadog-api-v2-collection/revisions/3f70db624329?raw)