GrowthBook (open-source experiments)
GrowthBook is an open-source platform for feature flags and A/B experimentation that is warehouse-native: rather than collecting its own event stream, it queries metrics from your existing data warehouse to evaluate experiments. This page describes its data model and privacy posture even-handedly, without ranking it against other tools.
What this means
GrowthBook combines feature flagging with experimentation, and is open source with a self-hostable server plus client SDKs. The SDKs evaluate flags and assign variations, often locally, so decisioning does not require a network round-trip per visitor.
Its defining trait is warehouse-native analysis: experiment results come from querying metrics you already store (in a warehouse or analytics database), rather than from a proprietary event collector. The platform orchestrates flags and statistics; your data stays in your pipeline.
Data model and posture
The flag/experiment definitions live in GrowthBook, while the underlying metric events live in your warehouse. Variation assignment is deterministic from a hashing key, so the same unit maps to the same variation without persisting a separate vendor identifier server-side.
This model shifts the privacy and residency questions to your own infrastructure: how you collect and store the metric events governs the posture, and self-hosting keeps the control plane in your environment too.
- Open source, self-hostable, with client SDKs
- Deterministic hash-based variation assignment
- Results computed by querying your data warehouse
- Privacy/residency follow your own pipeline choices
How it appears in analytics and logs
GrowthBook in an app means an SDK is evaluating feature flags and assigning experiment variations locally; analysis happens later by querying your warehouse, not by a separate tracking beacon.
Diagnostic use case
Use GrowthBook to manage feature flags and run experiments whose results are computed from metrics already in your warehouse, keeping experiment data in infrastructure you control.
What WebmasterID can help detect
WebmasterID's first-party events can be one of the metric sources a warehouse-native tool like GrowthBook evaluates, keeping experiment inputs in data you control.
Common mistakes
- Assuming GrowthBook collects its own event stream like a SaaS beacon.
- Defining experiment metrics inconsistently with warehouse definitions.
- Ignoring sample-ratio checks before trusting a result.
Privacy and accuracy notes
Because analysis runs against your own warehouse, the data residency and identifier handling are largely determined by your pipeline, not a vendor beacon. This is educational, not legal advice.
Related pages
- Statsig experimentation and feature gates
Statsig is a commercial experimentation platform that combines feature gates and dynamic configs with built-in metric computation. It logs exposure events when a unit sees a gate or experiment and evaluates configured metrics against those exposures. This page describes its data model and privacy posture even-handedly, without ranking it.
- Eppo warehouse-native experimentation
Eppo is a commercial experimentation platform built to be warehouse-native: rather than collecting its own event stream, it runs experiment analysis directly against metrics in your data warehouse. It supports randomization and statistical methods including variance-reduction techniques. This page describes its model and privacy posture even-handedly.
- Warehouse-native analytics
Warehouse-native analytics is an approach where the data warehouse (BigQuery, Snowflake, Redshift, Databricks) is the source of truth, and analytics tools query that data in place rather than copying it into a separate vendor store. You own the schema and computation; tools sit on top. It trades plug-and-play convenience for control, joinability, and avoiding data duplication.
- Event Explorer
Inspect the first-party events feeding your metrics.
Sources and verification notes
- GrowthBook — DocumentationOpen-source experimentation and feature-flagging docs.
Last reviewed 2026-06-24. Facts are checked against primary/official sources where available; uncertain specifics are marked “Data not yet verified” rather than guessed.