Google Data Manager API
Google Data Manager API receives conversion events and Customer Match audience members from Extract. Choose the destination for the data you want to send:
| Extract destination | Supported data |
|---|---|
| Google Ads Conversions API — Data Manager | Offline Conversions, including enhanced conversions for leads, and Enhanced Conversions for web |
| Google Ads Audiences API — Data Manager | Email, phone, mobile advertising ID, and third-party user ID audiences |
| Floodlight Conversions API — Data Manager | Floodlight conversions for Campaign Manager 360, Display & Video 360, and Search Ads 360 |
These are separate destinations from the original Google connectors. Google Analytics reporting and other Google reporting sources do not need to change for this migration.
Authorize a new destination
Create the appropriate Data Manager destination and click Connect to sign in with Google. Approve the Data Manager permission (https://www.googleapis.com/auth/datamanager) and any additional permissions requested for discovery. Use a Google account with access to the customer account or advertiser receiving the data. Google eligibility and account-access requirements still apply; see Google's access requirements.
A connection to an original Google destination does not grant this new permission. Authorize the new destination explicitly; credentials are not copied or migrated automatically. Creating a destination verifies access without uploading conversion events or audience members. Running a connection sends data to Google.
Migrate an existing connection
- Create and authorize a new Data Manager destination.
- Create a replacement connection with the same source and the new destination. Keep it disabled while configuring it.
- Select the corresponding data contract and recreate the field mappings. Follow the destination guide for changes to consent, identifiers, and transaction IDs. Matching contract names do not mean every legacy field is supported.
- Keep the same Ads customer ID and conversion-action assignments, audience names, or Floodlight profile and advertiser selections, where compatible.
- Disable the original connection before enabling the replacement. Check its cursor or source range so the first run sends the intended records.
- Review the run's ingestion responses, then check Google's delayed diagnostics.
The original destinations remain available with their existing behavior. Extract does not automatically switch versions or fall back to the original API if Data Manager rejects an upload.
Hashing and normalization
The destination guides explain normalization and hashing for their identifiers. Choose Apply hashing (SHA256) for raw email, phone, and supported name fields. Choose Source values are already hashed only when the source already contains correctly normalized SHA256 hexadecimal hashes. Mobile advertising IDs and third-party user IDs stay unhashed.
Check delivery and diagnostics
A successful upload with a Google requestId means accepted for processing. It does not confirm a matched user, attributed conversion, or usable audience size. Extract records request IDs and immediate field warnings in destination network request logs.
To inspect an accepted request:
- Copy its
requestIdfrom the network request log. - Wait at least 30 minutes.
- Open Google's request-status API Explorer, enter the request ID, and authorize with a Google account that has access to the receiving account.
- Check each destination's status and record counts.
PROCESSINGmeans the request is still pending. Final results areSUCCESS,PARTIAL_SUCCESS, orFAILURE; inspect warnings and errors even after success.
Processing can take up to 24 hours. Extract does not wait for these delayed results within a sync run. See Google's diagnostics guide.
For Google Ads conversions, open Goals > Conversions and inspect the relevant conversion action's diagnostics. The screens differ between offline imports and web enhancements; use the request-status API for a specific Data Manager request. For audiences, inspect the list in Google Ads Audience Manager. Synthetic identifiers can test request formatting but will not demonstrate real-user matching.
Immediate authorization, mapping, eligibility, and HTTP errors appear in Extract run logs. Continue on errors defaults to disabled. When enabled, processing continues after mapping errors or rejected uploads; failed records and batches remain failures. An invalid audience request rejects its entire batch. Unknown network outcomes are not automatically replayed, because Google might already have accepted the request.