Skip to main content

Configure and manage Identities

Manage your Identities deployment and configure identifiers using Snowplow Console.

Navigate to Identities in the Console sidebar. If Identities is not yet deployed for your pipeline, you'll see a setup page. Click Start configuration to begin.

If Identities is already deployed, you'll see the Identities overview page with your current configuration. Click Edit configuration to make changes.

Configure identifiers

Configure which fields from your events should be used as identifiers.

Choose stable identifiers

The fields you choose affect both resolution quality and the cost and latency of running Identities. Before adding an identifier, read choosing identifiers and validate its cardinality against your warehouse.

Each identifier type has:

FieldDescriptionRequired?
NameA unique name for this identifier type, e.g. acme_user_idYes
PropertyThe event property to extract the identifier value fromYes
UniqueWhether this identifier should prevent merges between Snowplow IDs with different valuesNo
PriorityThe priority used when generating fallback IDs; higher is preferredYes

Only one identifier can be marked as unique.

A single event field can only be mapped to one identifier type. Identifier names must be unique.

Add an identifier

Follow the steps in Console to create a new identifier.

Removing identifiers

Removing an identifier affects identity resolution for all future events. Snowplow IDs that would have been connected through this identifier will appear as separate Snowplow IDs going forward. Historical identities won't change. This can't be undone.

Enable cross-domain tracking aliases

If you use cross-domain tracking, check the Enable cross-domain tracking aliases box. This automatically extracts refr_domain_userid as an identifier and maps it to domain_userid and client_session_user_id, so the same user is resolved across sites with different cookie domains.

Mapping requirements

For cross-domain tracking aliases to work, you must have identifiers named domain_userid or client_session_user_id (or both) defined in the identifiers section above.

Set event filters

By default, Identities processes all events that flow through your pipeline. Use event filters to limit identity resolution to a subset of events.

Event filters are optional. Common reasons to filter events include:

  • Processing only events from specific applications, using app_id
  • Excluding server-side events that don't represent user activity
  • Limiting to events from specific platforms

You can add multiple filters and combine them with AND (all conditions must match) or ANY (at least one condition must match) logic.

To add a filter, click Add filter and configure:

FieldDescription
PropertyThe event property to filter on
Operatoris one of (include) or is not one of (exclude)
ValueA list of values to match

Supported scale

Identities is self-serviceable from Console. The default deployment is sized for typical production event volumes and reasonable identifier configurations, where each configured identifier is stable for a user. Following the guidance in choosing identifiers keeps your deployment within these defaults.

High-volume deployments

High-volume pipelines aren't covered by the self-serve defaults and shouldn't be deployed without guidance. If you expect a large event volume, or your warehouse checks show high-cardinality identifiers you can't avoid, contact Snowplow Support before deploying Identities.