Data 360 Activation
Getting Data 360 audiences and signals into the systems that act on them, and proving that what arrived matches what you built.
Activation is where Data 360 earns its keep: unified profiles and segments only matter once another system uses them. It is also where most silent failures happen. An audience can publish on time and still arrive incomplete, late or with the wrong attributes.
This page covers activation across destinations: which targets Salesforce documents, how Braze fits through an integration layer, how fresh the data really is, and how to detect and recover from failures. For Marketing Cloud journeys specifically, see activating Data 360 segments in journeys.
The business problem
Teams build a segment in Data 360 and assume it is live everywhere. In practice each destination has its own identifiers, refresh behaviour and limits. Without a design, marketing sends to last week's audience, ad platforms receive stale suppressions, and nobody can say which version of a segment a campaign used.
Proposed architecture
This is a reference design. We validate the supported targets, editions and permissions for each implementation.
Data 360 activation reference architecture
Reference architecture: a reusable pattern, not a live client integration
Select a component to see what it is responsible for and how data moves to and from it.
Components and data flows (text version)
| Component | Type | Responsibility |
|---|---|---|
| Salesforce CRMSales and Service Cloud | Salesforce product | Accounts, contacts, leads and cases, ingested through Salesforce's zero cost connectors. |
| Web and mobile SDKsSalesforce Interactions SDK | Channel or touchpoint | Behavioural events sent only after the visitor consents to tracking. |
| Warehouse and filesSnowflake, BigQuery, files | Data and analytics | External data ingested by file or API, or read through zero copy federation. |
| Data 360 unificationDMOs and unified profiles | Data and analytics | Maps DLOs to DMOs and runs identity resolution to produce unified profiles. |
| Segments and insightspublish on schedule or now | Data and analytics | Defines audiences on the Unified Individual and computes calculated insights. |
| Activation targetssegment plus target | Salesforce product | Store authentication for each destination and link segments with chosen attributes; results land in an Audience DMO. |
| Data actionsPlatform Events, webhook | Salesforce product | Send insight or record signals to Platform Events, Marketing Cloud or a webhook. |
| Marketing CloudEngagement or Next | Salesforce product | Receives activated audiences for journeys and sends. |
| Ad platformscheck supported list | Channel or touchpoint | Receive audiences for paid media targeting where Salesforce supports the platform. |
| Cloud storagee.g. S3, GCS, Azure Blob | External or custom system | Receives activated files that downstream tools can pick up. |
| CRM automationFlow or Apex | Salesforce product | Flow or Apex subscribers act on Platform Events from data actions. |
| Integration layermiddleware or iPaaS | Integration layer | Middleware or iPaaS that receives webhooks, transforms payloads and calls APIs with retries. |
| Brazevia CDI or REST API | Engagement platform | Receives profile attributes and audience flags through Cloud Data Ingestion or the REST API. Not a native Data 360 target. |
Data flows
- Salesforce CRM → Data 360 unification: CRM records. Zero cost connector streams. Mapping failures stop records reaching DMOs, so watch stream status.
- Web and mobile SDKs → Data 360 unification: consented events. SDK events, standard or real-time path. Standard engagement data lands every 15 minutes.
- Warehouse and files → Data 360 unification: ingest or federate. Files, Ingestion API or zero copy queries. Ingestion API data can take up to an hour to show.
- Data 360 unification → Segments and insights: unified profiles. Segments built on the Unified Individual. A failed identity run leaves profiles stale.
- Segments and insights → Activation targets: publish segment. Scheduled or Publish Now. Activations support incremental refresh.
- Segments and insights → Data actions: insight thresholds. Calculated insights trigger data actions when conditions are met.
- Activation targets → Marketing Cloud: audience. Members and attributes to Marketing Cloud. Reconcile counts before journeys use them.
- Activation targets → Ad platforms: audience. Audience identifiers in the form the platform accepts. Match rates vary by platform; check each one.
- Activation targets → Cloud storage: audience files. Files written to your bucket. Failed writes must be detected and re-published.
- Data actions → CRM automation: platform events. High-volume events are retained 72 hours, so subscribers can replay after short outages.
- Data actions → Integration layer: webhook. Webhook payloads to middleware, which must acknowledge, validate and queue them.
- Cloud storage → Braze: CDI file sync. Braze CDI reads files from S3, Azure Blob or GCS. File-storage sources are event-driven.
- Integration layer → Braze: users/track. Batches of up to 75 objects per request; back off on 429 using the rate-limit headers.
- Braze → Integration layer: engagement events. Optional. Currents via the Custom HTTP Connector or storage, summarised back to Salesforce by middleware.
Component responsibilities
| Component | Responsibility |
|---|---|
| Segment | Defines the audience "Segment On" an entity, usually the Unified Individual |
| Activation target | Stores the authentication and authorisation details for one destination platform |
| Activation | Links a segment and a target, with direct and related attributes; results land in an Audience DMO |
| Data action | Sends a signal from a calculated insight or record change to Platform Events, Marketing Cloud or a webhook |
| Integration layer | Carries data to destinations with no native target, with retries, logging and rate-limit handling |
Activation targets
Activation targets typically include Marketing Cloud Engagement, Marketing Cloud Next, cloud storage and ad platforms, but that list is not confirmed on the Salesforce pages we checked and it changes between releases, so we check Salesforce Help for the current targets before designing around one. For CRM, the usual path is a data action to Platform Events that Flow or Apex subscribes to.
Braze through an integration layer
We found no official Salesforce or Braze documentation of Braze as a native Data 360 activation target. Two technically sound routes exist:
- Storage hand-off: activate to a cloud storage bucket, and let Braze Cloud Data Ingestion (CDI) read it. CDI supports Amazon S3, Azure Blob Storage and Google Cloud Storage, and file-storage sources are event-driven.
- API hand-off: send data actions to a webhook in middleware, which calls the Braze
/users/trackendpoint. Each request carries up to 75 objects, and the burst limit is 3,000 requests per three seconds for customers with data points in their pricing, otherwise set by contract.
See Data 360 and Braze and Braze data activation for the full patterns.
Freshness
An activated audience is only as fresh as its slowest upstream step: source ingestion (standard web and mobile streams every 15 minutes for engagement, hourly for profiles), identity resolution runs, the segment publish schedule, and the destination's own import. Activations support incremental refresh. Write down the end-to-end freshness each use case needs, then check every step against it.
Security boundaries
Every target is an exit from your Salesforce trust boundary. Limit attributes to what the destination needs, keep credentials in the target configuration or middleware secrets store, and review who can create targets.
Failure handling
| Failure | Detection | Response |
|---|---|---|
| Segment publish fails or runs late | Publish status and timestamps | Re-publish; hold dependent journeys |
| Partial audience at destination | Count reconciliation against the Audience DMO | Re-activate; investigate identifier gaps |
| Webhook not acknowledged | Middleware logs and retry queue | Retry with backoff; dead-letter after a set limit |
| Braze returns 429 | X-RateLimit headers | Back off and resume from the last confirmed batch |
| Attribute mapping change | Schema diff in release review | Update targets before publishing |
Reconciliation
Reconcile at three points: segment count in Data 360, records in the Audience DMO, and records accepted by the destination. Differences are normal where a destination needs an identifier some profiles lack, but they should be explained and stable. Log each run so you can prove which audience version a campaign used.
Deployment model
Activation targets live in your org. Middleware runs in your cloud or iPaaS tenant. GeekMidst works remotely from India and builds in sandboxes first.
Implementation stages
- 01
Inventory destinations
List each destination, its identifier and required freshness.
- 02
Confirm targets
Check Salesforce Help for native support; design an integration path where none exists.
- 03
Build and map
Create targets, activations and attribute mappings.
- 04
Reconcile
Compare counts and sample records end to end.
- 05
Monitor
Alert on publish failures, count swings and API errors.
Testing strategy
Use a seed segment of known test profiles. Confirm each profile arrives with the right attributes, then remove one and confirm the removal propagates. Test failure paths deliberately: expired credentials, a rejected record and a rate-limit response.
Operational monitoring
Track publish times, activation status, destination acceptance counts and middleware error rates. Our segmentation and activation troubleshooting guide lists the most common causes.
Prerequisites and limitations
Data 360 must be provisioned, and each destination needs its own licence and API access. Braze is a separate vendor; Salesforce has announced an agreement to acquire Braze, but until the deal closes they are separate companies and no native integration has been announced.
Related
Data 360 solution centre
Every Data 360 capability in one place
→Segmentation and insights
Building segments and calculated insights
→Real-time segmentation
Streaming vs batch audiences
→Integration solutions
Middleware and API patterns
→Questions we hear
Can Data 360 activate directly to Braze?
Why do destination counts differ from segment counts?
How fresh is an activated segment?
Is CRM a native activation target?
Sources
- https://trailhead.salesforce.com/content/learn/projects/explore-data-cloud-core-functionality/build-a-segment-and-report
- https://developer.salesforce.com/docs/data/data-cloud-dev/guide/dc-dev-overview.html
- https://www.braze.com/docs/user_guide/data/cloud_ingestion
- https://www.braze.com/docs/api/endpoints/user_data/post_user_track
- https://developer.salesforce.com/docs/atlas.en-us.api_streaming.meta/api_streaming/using_streaming_api_durability.htm
Make activation dependable
We can review your targets, freshness needs and reconciliation, and propose fixes.
