Documented steps

Use a Console API Key with at least read access to the project.
Identify the target experiment's id, e.g. via GET https://statsigapi.net/console/v1/experiments or the Console UI.
Fetch the experiment's group IDs via GET https://statsigapi.net/console/v1/experiments/{id} and note the control group's and desired test group's IDs from the groups array.
Build request: GET https://statsigapi.net/console/v1/experiments/{id}/pulse_results
Header: STATSIG-API-KEY: <key>.
Required query params: control=<controlGroupID>, test=<testGroupID>.
Optional query params: date (YYYY-MM-DD snapshot), confidence (0-100, overrides experiment default), cuped (true/false), applyBonferroniPerVariant, applyBonferroniPerMetric, bonferroniPrimaryMetricWeight, applyBenjaminiHochbergPerMetric, applyBenjaminiHochbergPerVariant, benjaminiPrimaryMetricsOnly.
Send the request; a 200 returns { message, data: { ds (data snapshot date), primaryMetrics, secondaryMetrics, otherMetrics } }, where each metric entry includes metricID, metricName, directionality, absoluteChange, percentChange, confidenceInterval {lower, upper}, pValue, testMean/controlMean, testUnits/controlUnits, adjustedAlpha, absoluteToplineImpact/relativeToplineImpact.
Use pValue and confidenceInterval per metric to judge statistical significance, and the topline impact fields for projected business impact.
For feature-gate rollout rules (rather than experiments) the analogous path is GET /console/v1/gates/{id}/rules/{ruleID}/pulse_results.
Reference: https://docs.statsig.com/api-reference/experiments/retrieve-pulse-results-beta

Known gotchas

This endpoint is explicitly marked "(Beta)" - its response schema and availability are subject to change without GA stability guarantees.
control and test query params are required and must be actual group IDs pulled from the experiment's groups array, not free-text group names.
Results are tied to a data snapshot (ds); requesting a date outside the experiment's actual run window can return stale or empty results.
Multiple correction methods (Bonferroni and Benjamini-Hochberg, per-variant and per-metric) are independently toggled via query params and default off - results may not match the Console UI if the experiment was configured with sequentialTesting/bonferroniCorrection at creation and those flags aren't replicated here.
Console API keys can be personal and role-scoped; a key without visibility into a given experiment or team can receive a 403 even though it works for other experiments.
The older get_daily_reports Console API endpoint is deprecated in favor of the current metrics/pulse endpoints.

docs.statsig.com · 11 steps · unrated

Create a Statsig A/B test experiment with variant groups and metrics, then start it live via the Console API

docs.statsig.com · 10 steps · unrated

Read a Statsig experiment in the Node.js server-core SDK and get the user's assigned group and parameter values

docs.statsig.com · 12 steps · unrated

Give your agent this knowledge — and 15,700+ more routes

One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus: claude mcp add --transport http waymark https://mcp.waymark.network/mcp

Retrieve Statsig Pulse experiment results (lift, p-value, confidence intervals) for a test vs control group via the Console API

Documented steps

Known gotchas

Related routes

Give your agent this knowledge — and 15,700+ more routes

Need this verified for your stack — or a route we don't have yet?