CONNECTIONS / Databricks / Databricks to Meta Custom Audience
Your Databricks lakehouse, wired to Meta lookalike and suppression seeds.
Signals reads the Delta table through a Unity Catalog grant and keeps Meta audiences synced as adds and removals so lists never drift, hashed on the way out and governed like any other lakehouse job.
- ● databricks → meta.audience :: live
- > read main.audiences.segment_tier rows=17,346
- hash sha256(email,phone) consent=filtered
- route meta deliver
- ✓ delivered · 14,744 matched
WHAT THIS ENABLES
Databricks to Meta Custom Audience
Databricks to Meta Custom Audience: A notebook job reads the customer segments in your Databricks Delta tables, and Signals keeps Meta audiences synced as adds and removals so lists never drift, hashed and consent-checked each time the model rebuilds.
- Signals keeps Meta audiences synced as adds and removals so lists never drift, reading your Databricks Delta tables each time the model rebuilds.
- Lookalike seeds and suppression lists that track your live customer base, fed by data engineers modeling in Databricks notebooks.
WHAT FLOWS WHERE
Databricks to Custom Audience, mapped.
The notebook job streams customer segments into Signals for hashing and consent screening, then Meta's Custom Audiences API records the match-ready audiences with audience refresh cadence accounted for.
The notebook job streams customer segments into Signals for hashing and consent screening, then Meta's Custom Audiences API records the match-ready audiences with audience refresh cadence accounted for.
Built for the teams that own the number.
Data engineers modeling in Databricks notebooks who need lookalike seeds and suppression lists that track your live customer base.
From kickoff to verified events.
-
Connect
A Unity Catalog-scoped service principal on the single Delta table in play to feed Meta's Custom Audiences API, sized for audience refresh cadence.
-
Map
A notebook job reads your Databricks Delta tables, and its columns map to Meta's Custom Audiences API in the visual mapper, hashed and checked for audience refresh cadence.
-
Deliver
Each time the model rebuilds, Signals keeps Meta audiences synced as adds and removals so lists never drift, reads only the rows the model rebuilt, with audience refresh cadence watched in the debugger.
What ships with this use case.
Lookalike seeds and suppression lists that track your live customer base on every run, since Signals reads only the rows the model rebuilt.
Always-current lookalike seeds and exclusions, delivered each time the model rebuilds without job-cluster time climbing.
Each run returns a match report on audience refresh cadence from Meta's Custom Audiences API, read from your Databricks Delta tables each time the model rebuilds, at flat job-cluster time.
What Custom Audience actually receives.
- ● signals :: audience upload
- > POST /customaudiences source=databricks.main.audiences.segment_tier
- schema ["EMAIL_SHA256","PHONE_SHA256"]
- num_received 17,346 · num_invalid_entries -319
- audience_id "1691364" · session "upsert"
- ✓ accepted match=85%
Adjacent moves on the same stack.
from Databricks
Databricks to Google Customer Match
Databricks to Google Customer Match: A notebook job reads the customer segments in your…
Learn more
from Databricks
Databricks to TikTok Custom Audience
Databricks to TikTok Custom Audience: A notebook job reads the customer segments in your…
Learn more
same use case
ActiveCampaign to Meta Custom Audience
ActiveCampaign to Meta Custom Audience: Signals reads your ActiveCampaign contacts and deal…
Learn moregeneric
Meta Custom Audience
Build and refresh Meta Custom Audiences from your CRM, warehouse, or SingleView, with…
Learn moreAsked on almost every call.
How often does the Databricks Meta Custom Audience sync run?
Cadence follows the job that rebuilds the table: a nightly model gives Meta a nightly feed, an hourly one an hourly feed. Because a run picks up only the rows the model rebuilt, lookalike seeds and suppression lists that track your live customer base keeps pace on audience refresh cadence and job-cluster time stays modest.
What match rate should a Databricks-sourced Meta Custom Audience batch expect?
Identifier quality in the Delta table drives it, not the cluster size. Rows carrying a live hashed email or phone land; rows with neither will not, and no rebuild changes that. The opening run's feedback from Meta's Custom Audiences API flags the gap for performance teams running first-party suppression and lookalike seeds tracking audience refresh cadence.
NEXT STEP