Experimentation Events Export
Securely export your Optimizely Experimentation event data so you can analyze experiment results with the tooling of your choice.
Optimizely Experimentation Events Export delivers daily Parquet-formatted files with one row per decision or conversion event your visitors generate. Each row includes the event name, visitor ID, and any metadata you sent to Optimizely Experimentation. Use the export when you need to analyze your experiments outside Optimizely. See the Data specification for the complete field reference.
ImportantYou can download your Experimentation Events Export from two storage locations: Amazon S3 and Google Cloud Storage (GCS). GCS is available to all customers today. Choose whichever fits your workflow.
Customers on the EU data servers can use GCS only.
Service overview
ImportantAmazon S3 export is being decommissioned.
Starting November 30, 2026, Optimizely will stop delivering Experimentation Events Export data through Amazon S3. After this date, use Google Cloud Storage (GCS) to download your data.
GCS is available to all customers today and contains data from June 2026 onward. To migrate, use the Authentication API to exchange your Personal Access Token for GCP credentials, then read your data with the Google Cloud CLI
gcloud storage).Data from before June 2026 is available only in Amazon S3.
Optimizely exports your event data continuously to Amazon S3 and Google Cloud Storage (GCS). Each daily export is available the next day by 13:00 UTC through the Authentication API.
- Read S3 data with the AWS API, CLI, or SDK.
- Read GCS data with the Google Cloud CLI or a Google Cloud Storage client library.
The format and schema are identical in both locations, only the storage location and credentials differ. See the Authentication API to get credentials for each.
Use cases
The Optimizely Experimentation Results page displays how your experiments perform against your configured metrics. Export the underlying event data when you need to go further than the Results page allows, like combining Optimizely Experimentation data with data from external sources.
Audit or troubleshoot results
- Verify event counts and confirm data accuracy.
- Inspect the events you sent to Optimizely, including impressions.
Calculate additional metrics
- Analyze experiment data using SQL or other data querying methods.
- Create experiment data dashboards using third-party visualization tools.
- Compute compound metrics ("this event and that event").
Combine experiment data with external sources
- Centralize data management by copying source-of-truth experiment data to a third-party data warehouse.
- Join experiment data with other data sources to measure impact on a broader range of external metrics.
Build applications with experiment data
- Create experiment notifications.
- Build internal dashboards that alert on guardrail-metric violations.
Retention policy
Optimizely retains files for one year, then deletes them. Optimizely follows a data-deletion workflow that complies with the General Data Protection Regulation (GDPR). For details, see the Optimizely Trust Center.
Updated 5 days ago
