Back to Product Updates
Integrations

Sync dbt semantic layer metrics to GrowthBook

November 3, 2026

Use dbt Sync to bring your semantic layer definitions into GrowthBook. The new @growthbook/dbt CLI turns dbt semantic models into GrowthBook fact tables and dbt metrics into fact metrics. Each run of the CLI updates them.

How the sync works

The CLI reads target/manifest.json, which dbt writes after dbt parse. It needs no warehouse connection and no dbt runtime of its own. It uses the same bulk import API as the GitHub metrics workflow, but it reads dbt's definitions instead of a GrowthBook-format YAML file.

  • Each semantic model becomes a fact table, with its agg_time_dimension as the timestamp column.
  • Most simple metrics become Mean metrics, ratio metrics become Ratio metrics, and conversion metrics become two-step Funnel metrics. Median and percentile aggregations become Quantile metrics.
  • Synced fact tables and fact metrics are managed by dbt and read-only in the GrowthBook UI. You change them in dbt and sync again.
  • Definitions the CLI can't map, such as min aggregations, are skipped with a message.

Settings with no dbt equivalent, such as conversion windows and capping, go under config.meta.growthbook. The CLI upserts by ID, so renaming a dbt metric creates a new GrowthBook metric. Run with --prune to archive the old one.

Check with a dry run, then automate

Start with --dry-run. It validates the sync against the GrowthBook API without writing anything and lists what it would skip. This excerpt from the docs assumes ds_abc123 is your Data Source ID and maps a dbt customer entity to user_id:

export GROWTHBOOK_API_KEY=secret_admin_...
npx @growthbook/dbt sync --datasource ds_abc123 --entities customer=user_id --dry-run

Remove --dry-run to apply the changes. If your API key requires a person to confirm writes, the CLI prints a confirmation link and waits. You can run the same command by hand or as a CI step after dbt parse. Add --strict in CI so the job exits with an error when anything is skipped.

Getting started

  • Open Metrics and Data → Data Sources, open the Data Source your dbt project writes to, and copy its ID from the URL.
  • Go to Settings → API Keys and create a Secret Key with the Admin role.
  • Synced Quantile and Funnel metrics need a plan that includes those metric types. If you self-host, use GrowthBook 5.2 or later for this workflow.

Follow the dbt semantic layer sync guide for the full mapping, selectors, and overrides.

Similar updates

See All Updates
Key Feature
Feature Flags

Target lists with millions of IDs using Remote Saved Groups

November 3, 2026
Key Feature
AI Platform

Confirm risky agent changes with Human in the Loop

November 3, 2026
Key Feature
Experimentation

Compare ranking algorithms with interleaving experiments (private beta)

November 3, 2026
Integrations

Event Forwarder adds a free tier on Starter plans

November 3, 2026
Feature Flags
Integrations

Roll back ramp-ups from Datadog alerts

November 3, 2026
Integrations

Expanded REST API coverage in GrowthBook 5.2

November 3, 2026
Experimentation

Personalize AI Visual Editor variations with contextual bandits

November 3, 2026
Integrations

Redesigned data source setup for connecting your warehouse

November 3, 2026
Experimentation

Create threshold and active days metrics and preview the results

November 3, 2026
Experimentation

Redesigned experiment page with inline editing and a To Do tab

November 3, 2026
Experimentation

See what drives warehouse usage on each data source

November 3, 2026
Experimentation

Inspect experiment exposure logs on the Health tab

November 3, 2026
Key Feature
SDKs

Evaluate multiple Java feature flags in one call

October 15, 2026
Key Feature
Product Analytics

Inspect warehouse query statistics and audit records

October 12, 2026
Key Feature
Product Analytics

See how analysis adjustments affect a metric's results

October 12, 2026
Key Feature
Feature Flags

Reuse Saved Groups and find broken targeting references

October 9, 2026
Key Feature
Product Analytics

Connect LLM traces to experiment metrics

October 5, 2026
Key Feature
AI Platform

Set up GrowthBook SDKs with your coding agent

September 29, 2026
Key Feature
Product Analytics

Explore user journeys in Product Analytics

September 22, 2026
Key Feature
Product Analytics

Explore SQL results and turn them into charts

September 22, 2026
Key Feature
Integrations

New Slack integration

September 22, 2026
Key Feature
AI Platform

Connect agents to hosted MCP with OAuth

September 22, 2026
Key Feature
Experimentation

Measure multi-step funnels inside experiments

September 22, 2026
Key Feature
Feature Flags

Personalize variations with contextual bandits (beta)

September 22, 2026
Key Feature
Product Analytics

Keep Managed Warehouse event data in the EU

September 11, 2026
Key Feature
Experimentation

Capture reusable Learnings across experiments

August 27, 2026
Key Feature
Integrations

Analyze Adobe Experience Platform data in GrowthBook

August 18, 2026
Key Feature
SDKs

Use typed feature values in Swift and Kotlin

August 15, 2026
Key Feature
Feature Flags

Share feature flags across projects

July 24, 2026
Key Feature
Feature Flags

Validate string and number flags with schemas and scoped hooks

July 23, 2026
Key Feature
Integrations

Connect Databricks with service principal OAuth

July 22, 2026
Key Feature
Product Analytics

Product Analytics is generally available

July 20, 2026
Key Feature
AI Platform

Build website variations with the AI Visual Editor

July 20, 2026
Key Feature
AI Platform

Use the in-app AI Assistant across GrowthBook

July 20, 2026
Key Feature
AI Platform

Use agent skills and the rebuilt GrowthBook CLI

July 20, 2026
Key Feature
Experimentation

Create experiment drafts before finishing analysis setup

July 20, 2026
Key Feature
Experimentation

Apply custom validation before experiment changes

July 16, 2026
Key Feature
Feature Flags

Scheduled analysis refreshes for monitored ramps

July 14, 2026
Key Feature
SDKs

Remote evaluation in Python, PHP, and Java

June 30, 2026
Key Feature
Feature Flags

Review and schedule changes to Saved Groups

June 29, 2026
Key Feature
Feature Flags

Override selected fields in JSON flag values

June 19, 2026
Key Feature
Product Analytics

Use incremental refresh with Snowflake

May 28, 2026
Key Feature
Feature Flags

Run bandits without sticky bucketing

May 27, 2026
Key Feature
Product Analytics

Run Product Analytics explorations through MCP and REST

May 27, 2026
Key Feature
AI Platform

Ask the AI Data Analyst about your product metrics

May 27, 2026
Key Feature
Feature Flags

Plan gradual releases with ramp schedules

May 27, 2026
Key Feature
Experimentation

Manage the experiment lifecycle through the REST API

May 27, 2026