Available connectors
Advertising platforms
The Trade Desk
Deliver audiences to TTD’s first-party or third-party marketplace with automatic UID2 generation
Meta Custom Audiences
Push custom audiences to Facebook and Instagram advertising
Meta Conversions API
Send conversion events for ad attribution and campaign optimization
Google Ads
Deliver Customer Match audiences to Google Ads
Google DV360
Activate audiences in Google Display & Video 360
TikTok
Deliver custom audiences to TikTok Ads Manager
TikTok Conversions API
Send offline conversion events to TikTok for ad attribution
Snapchat
Deliver custom audiences to Snapchat Ads Manager
Yahoo DSP
Activate audiences in Yahoo’s demand-side platform
Yahoo Conversions API
Send conversion events to Yahoo DSP for ad attribution
Deliver custom audiences to Pinterest for ad targeting
Pinterest Conversions API
Send conversion events to Pinterest for ad attribution and optimization
Deliver custom audiences to Reddit Ads for ad targeting
Magnite
Deliver audience segments to Magnite’s SSP
PubMatic
Deliver audience segments to PubMatic for programmatic advertising
PubMatic Conversions API
Send conversion events to PubMatic for ad attribution
AI/ML
Hugging Face
Export trained AI models to Hugging Face Hub
Cloud storage
Amazon S3
Deliver data files to your Amazon S3 bucket
Connector interfaces
A connector is installed as an app (identified by anapp_id). Each app exposes one or more interfaces (each identified by an interface_id); a dataset connects to a specific (app_id, interface_id) pair, not to the connector as a whole. Some connectors appear more than once in Available connectors above because a single app exposes several interfaces — Meta Custom Audiences and Meta Conversions API are two interfaces of one installed Meta app, each with its own collaboration policy.
Each interface defines:
Before creating a connection, check which interfaces accept a dataset’s schema; an accepted
(app_id, interface_id) pair is one a connection will succeed against. See Connector Interfaces for the concept and the connection creation flow for the API calls.
Connector requirements
Most connectors match users on identifiers, enforced by the interface’s collaboration policy. The requirements are listed on each connector’s own page under Supported identifiers — open a connector from Available connectors above for its exact table. A few connectors work differently:For identifier formatting requirements, see Hashing PII for Upload.
Audience membership duration
Membership duration is how long a delivered identifier remains active in the destination audience before the destination platform expires it. It is distinct from the dataset’s retention policy (which governs data stored in Narrative) and the interface’s refresh window (which governs how stale the source dataset may get). Each audience destination handles membership duration differently. Per audience means the duration is set once when the audience is created and never updated afterward; per record means the duration is sent with every delivered identifier, so each delivery renews the clock. See each connector’s Audience membership duration section for details.
Conversions API interfaces deliver events rather than audience memberships, so membership duration doesn’t apply to them; see each Conversions API page for its event age limits. Amazon S3 and Hugging Face deliver files and models, so nothing expires.
Common workflows
Delivering first-party data
Push your own customer data to platforms for targeting:- Prepare your data with hashed identifiers
- Upload to a dataset
- Map to Rosetta Stone attributes
- Configure and deliver through the connector
Delivering acquired data
Activate third-party data purchased through Narrative:- Query data from data collaborators using NQL
- Materialize results into a dataset
- Deliver through the appropriate connector
Related content
Data Activation Overview
Understand how connectors fit into the activation workflow
Structuring Audiences
Organize your data for efficient delivery
Rosetta Stone Mapping
Map your schema for connector compatibility
Schema Incompatibility Errors
Resolve HTTP 400 errors when a dataset schema doesn’t match a connector

