Data 360 data ingestion and modelling
Data streams, connectors, DLO to DMO mapping, transforms and zero-copy federation for Data 360, designed so identity resolution and segments work later.
Everything in Data 360 depends on how data comes in. Identity resolution, calculated insights, segments, personalisation and agent grounding all read from the objects you create at ingestion, and some ingestion settings can't be changed later. Our Data 360 data ingestion service designs and builds that foundation: data streams, connectors, mapping to the data model, transforms and zero-copy access.
This service covers the foundation layer. For the full programme, see Data 360 consulting. For unifying profiles afterwards, see identity resolution.
What's included
| Area | What we do |
|---|---|
| Source inventory | List sources, owners, volumes and freshness needs, and choose ingestion or zero copy for each |
| Data streams | Configure each stream's category (Profile, Engagement or Other), primary key and, for engagement data, event time |
| Connectors | Use the zero-cost connectors for Sales, Service, Industries, B2C Commerce and Marketing Cloud Engagement, and the Ingestion API, CSV or unstructured sources for others |
| Mapping | Map DLOs to DMOs in the Customer 360 Data Model, using starter data bundles where they fit |
| Transforms | Batch or streaming data transforms to clean, join and reshape data |
| Zero copy | Query federation, data sharing and Live Query, with optional cached acceleration |
| Documentation | A data dictionary of streams, objects, keys and owners |
How we work
- 01
Discovery
Inventory sources and the use cases they serve, and agree which data must be real-time and which can be batch.
- 02
Design
Design streams, keys, DLO to DMO mappings, transforms and zero-copy choices, and record them in a data dictionary.
- 03
Build
Create connectors and streams, build mappings and formula fields, and configure transforms.
- 04
Test
Reconcile record counts with source systems, check keys and event times, and test refresh behaviour.
- 05
Launch
Switch on schedules, watch the first runs, and hand over monitoring.
- 06
Support
Add sources, adjust mappings and watch job health as use cases grow.
Decisions that are hard to reverse
- Refresh Mode cannot be changed after a data stream is created.
- Stream category and primary key affect everything built on the stream later, including identity resolution and segments.
- Ingestion API data can take up to an hour to show.
- Standard web and mobile streams ingest engagement data every 15 minutes and profile data hourly. Real-time use cases need the real-time layer and real-time data graphs.
Zero copy, with status
Salesforce lists AWS, Databricks, Google Cloud and Snowflake as zero-copy partners, along with Microsoft and IBM. It names out-of-the-box support for Snowflake, Databricks, Redshift and BigQuery. Its zero-copy page labels file federation "In Beta Now", which may be out of date, so we check Help before relying on it. Salesforce described its BigQuery zero-copy expansion as "GA Fall 2026". Python code extensions on zero-copy data come through a CLI plugin in Beta.
Prerequisites and licences
Data 360 must be provisioned in your org. Processing has cost implications, so we review your contract's usage terms with you before scheduling jobs. We don't quote prices.
Common mistakes
- Picking a refresh mode in a hurry, then rebuilding the stream.
- Mapping everything to custom objects instead of the standard data model.
- Copying data that zero copy could query in place, or the reverse.
- Skipping data quality checks before identity resolution. See data quality checks before identity resolution.
Related reading
DLO vs DMO
How Data 360 models data
→Choosing ingestion patterns
Batch, streaming and API options
→Virtualisation vs replication
When to use zero copy
→Data 360 architecture guide
A reference for the full stack
→Questions we hear
Should we ingest data or use zero copy?
Can we change a stream after creating it?
Do starter data bundles do the mapping for us?
When should we use streaming transforms?
Sources
Design your Data 360 foundation
List your main sources and what you want Data 360 to do. We'll suggest an ingestion plan and the decisions to make first.
