Overview
This section is about the front door of Data 360. Its three objectives cover data transformation capabilities, the processes and considerations for ingesting data from varied sources, and Data 360's data collaboration capabilities, including Zero Copy.
Sources include the Salesforce CRM connector, Marketing Cloud Engagement and Personalization, B2C Commerce, cloud storage such as Amazon S3, Google Cloud Storage and Azure, the Ingestion API in streaming and bulk modes, the Web and Mobile SDKs, MuleSoft and other connectors. Each data stream needs a category (Profile, Engagement or Other), a primary key, an event time field for engagement data, and a refresh mode. Formula fields can clean values at ingestion, while batch and streaming data transforms reshape data after it lands. Zero Copy lets Data 360 query Snowflake, Databricks, Google BigQuery or Amazon Redshift data in place and share Data 360 data back to them without building copy pipelines.
How it's tested
Questions describe a source system, data volume, freshness requirement and data quality problem. You choose the connector, data stream settings, transform type or Zero Copy option that meets the requirement with the least effort and cost, keeping in mind that early choices such as category and primary key cannot easily be undone.
-
Pick the path
A source and freshness need are described, and you choose between a native connector, cloud storage, the Ingestion API, the SDKs or Zero Copy.
-
Set the stream right
A dataset is described, and you choose its category, primary key, event time field or refresh mode.
-
Shape the data
Data needs cleaning, joining or aggregating, and you choose a formula field, a batch transform or a streaming transform.
A retailer keeps five years of order history in Snowflake, refreshed by its own pipelines, and wants to segment on lifetime order value. Copying the tables into Data 360 would duplicate storage and add a pipeline. Zero Copy federation lets Data 360 map the Snowflake tables as external data lake objects and use them in insights and segments, so it is the better choice.
Objectives 3
Official objectives from the Salesforce outline, in outline order.
3.1 Explain data transformation capabilities within Data 360.Formula fields, batch transforms and streaming transforms.
- What it means
- Data 360 can reshape data at three points. Formula fields on a data stream compute or clean values during ingestion, such as normalizing phone formats. Batch data transforms run on a schedule over data lake objects to join, aggregate, filter or pivot data into new lake objects. Streaming data transforms process records continuously as they arrive.
- What to understand
- Use formula fields for simple row-level fixes that need no other rows. Use batch transforms, built visually or with SQL-like logic, for joins, aggregations and complex reshaping where scheduled refresh is acceptable. Use streaming transforms when the output must stay current with incoming events and the logic is simpler, such as filtering or reformatting a single object stream. Transform outputs are new data lake objects that you then map to the data model.
- How it shows up
- A data quality or reshaping need is described with a freshness requirement, and you choose formula field, batch transform or streaming transform.
- Common confusion
- Using a formula field for logic that requires joining two datasets; joins need a data transform.
3.2 Describe processes and considerations for data ingestion from varied sources into Data 360.Connectors, data streams, categories, keys and refresh modes.
- What it means
- Ingestion uses a connector to reach a source, then a data stream that selects objects or files, sets fields and types, assigns a category and a primary key, and schedules refreshes. Native connectors cover Salesforce CRM, Marketing Cloud, Commerce Cloud and cloud storage; the Ingestion API and SDKs cover custom apps, websites and mobile apps.
- What to understand
- Choose the category carefully: Profile for people and accounts, Engagement for time-stamped events that need an event time field, Other for reference data. Primary keys must be unique, and fully qualified keys help when several sources reuse the same key values. Refresh modes include full refresh and upsert, and streaming or near real-time options exist for CRM and API sources. Consider volume, file format, schema changes, credentials and credit cost when choosing frequency and method.
- How it shows up
- A source system and data shape are described, and you choose the connector, category, key or refresh setting.
- Common confusion
- Ingesting order events as a Profile category stream, which removes the event time handling engagement data needs.
3.3 Describe Data 360's data collaboration capabilities, including Zero-Copy.Federation in, sharing out and clean room style collaboration.
- What it means
- Zero Copy covers two directions. Data federation lets Data 360 query data that stays in external platforms such as Snowflake, Databricks, Google BigQuery and Amazon Redshift, exposed as external data lake objects. Data sharing lets those platforms query Data 360 objects directly. Data 360 also supports collaboration features for working with partner data in governed ways.
- What to understand
- Federation suits large or frequently updated warehouse data that the customer already manages, avoiding duplicate storage and pipelines. Live query reads data at use time, while acceleration caches results for performance. Sharing out suits data science or BI teams that work in the warehouse and need unified profiles or insights. Consider latency, query cost on both sides and which platform holds the system of record before choosing Zero Copy over ingestion.
- How it shows up
- A customer with an existing warehouse wants to use or share data, and you choose federation, sharing or standard ingestion.
- Common confusion
- Believing Zero Copy means no cost at all; queries still consume resources and credits on one or both sides.
High-yield concepts
The terms, rules and formulas this domain turns into questions most often.
-
Connector
The configured connection to a source such as Salesforce CRM, Marketing Cloud, S3 or an API.
- Remember
- A connector is set up once and reused by many data streams.
- Tested as
- Source selection questions.
- Trap
- Confusing the connector with the stream.
-
Data stream category
Profile, Engagement or Other, set when the stream is created.
- Remember
- Engagement requires an event time field; the category cannot be changed later.
- Tested as
- Stream design questions.
- Trap
- Choosing Profile for event data.
-
Primary key
A field that uniquely identifies each record in a stream.
- Remember
- Duplicates cause records to overwrite or fail.
- Tested as
- Data quality questions.
- Trap
- Using email as a key when people share addresses.
-
Fully qualified key
A key combined with a key qualifier that identifies its source.
- Remember
- Prevents collisions when sources reuse ID values.
- Tested as
- Multi-source joins and insights.
- Trap
- Joining on raw IDs from two systems.
-
Ingestion API
REST API for streaming small batches or bulk loading data from custom systems.
- Remember
- Needs a schema and connector defined first.
- Tested as
- Custom application sources.
- Trap
- Using it where a native connector exists.
-
Web and Mobile SDK
Libraries that capture behavioral events from websites and apps into Data 360.
- Remember
- Events arrive as engagement data.
- Tested as
- Digital behavior tracking scenarios.
- Trap
- Expecting the CRM connector to capture web clicks.
-
Formula field
A computed field added to a data stream during ingestion.
- Remember
- Row-level only, no joins.
- Tested as
- Simple cleanup questions.
- Trap
- Using it for aggregations.
-
Batch data transform
Scheduled transformation that joins, aggregates and reshapes data lake objects.
- Remember
- Outputs a new data lake object.
- Tested as
- Complex reshaping questions.
- Trap
- Choosing it for real-time needs.
-
Streaming data transform
Continuous transformation applied to incoming records.
- Remember
- Keeps output current with simpler logic.
- Tested as
- Near real-time reshaping questions.
- Trap
- Expecting complex multi-object joins.
-
Zero Copy federation
Querying external warehouse data in place as external data lake objects.
- Remember
- Data stays in Snowflake, Databricks, BigQuery or Redshift.
- Tested as
- Existing warehouse scenarios.
- Trap
- Building a copy pipeline instead.
-
Zero Copy sharing
Letting an external warehouse query Data 360 objects directly.
- Remember
- Shares unified profiles and insights outward without exports.
- Tested as
- Data science and BI scenarios.
- Trap
- Exporting files on a schedule instead.
-
Refresh mode
How a stream updates: full refresh, upsert or streaming, depending on the source.
- Remember
- Frequency affects freshness and credit cost.
- Tested as
- Cost and latency trade-offs.
- Trap
- Choosing the fastest refresh by default.
Side by side
Concepts candidates mix up, compared on the point that decides the answer.
Transformation options
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Formula field | Computes a value per row at ingestion | No joins, applied as data lands | Using it for multi-object logic |
| Batch data transform | Scheduled joins and aggregations | Most powerful, not real time | Choosing it for instant updates |
| Streaming data transform | Continuous processing of new records | Current output, simpler logic | Expecting heavy aggregations |
Getting data into Data 360
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Native connector | Prebuilt link to Salesforce clouds or storage | Least setup for supported sources | Building custom API code instead |
| Ingestion API | Custom streaming or bulk loads | For systems without a connector | Ignoring schema setup |
| Web and Mobile SDK | Captures digital behavior events | Engagement data from sites and apps | Expecting profile data only |
| Zero Copy federation | Queries external warehouse data in place | No physical copy | Assuming it is always cheaper |
Exam traps
Each trap is an answer that looks right. Learn the clue that should change your mind.
-
Streaming everything
- What candidates pick
- Set every source to the most frequent refresh available.
- Why it looks right
- Fresher data seems always better.
- Why it's wrong
- Frequent refreshes consume more credits; daily-changing data should use scheduled batch refresh.
Clue to watch for The source only changes once a day.
-
Formula for joins
- What candidates pick
- Use a formula field to combine order and product data.
- Why it looks right
- Formula fields are quick to add.
- Why it's wrong
- Formula fields work on one row at ingestion; combining datasets needs a batch or streaming transform.
Clue to watch for The requirement mentions two or more datasets.
-
Wrong category
- What candidates pick
- Ingest web click events as a Profile stream.
- Why it looks right
- The events relate to people.
- Why it's wrong
- Behavioral events belong in Engagement with an event time field so insights and segments can use time windows.
Clue to watch for The data has timestamps and many rows per person.
-
Copy the warehouse
- What candidates pick
- Build nightly exports from Snowflake into cloud storage for ingestion.
- Why it looks right
- File ingestion is familiar.
- Why it's wrong
- Zero Copy federation reads Snowflake tables in place and avoids maintaining another pipeline.
Clue to watch for The data already lives in a supported warehouse.
-
Raw IDs collide
- What candidates pick
- Join two sources on their native customer ID fields.
- Why it looks right
- Both fields are called customer ID.
- Why it's wrong
- Different systems can reuse the same values; fully qualified keys add the source context to avoid false matches.
Clue to watch for Two sources use overlapping numeric IDs.
-
API for CRM
- What candidates pick
- Push Salesforce CRM records through the Ingestion API.
- Why it looks right
- It gives full control.
- Why it's wrong
- The Salesforce CRM connector ingests CRM objects natively with less setup and supports frequent updates.
Clue to watch for The source is a Salesforce org.
Answer framework
Use these steps when a scenario asks how to bring a source into Data 360.
- Identify the source and check whether a native connector or Zero Copy option exists before considering the Ingestion API, which suits custom systems that push data through streaming or bulk calls.
- Decide whether the data should be copied in or queried in place, based on volume, ownership and freshness.
- Set the stream category, primary key and event time field to match whether the data describes people, events or reference values.
- Choose refresh frequency from the business freshness need, keeping credit consumption in mind, because more frequent refreshes process more rows.
- Pick the lightest transformation that works: formula field for row fixes, streaming transform for current simple logic, batch transform for joins.
Practice question
Want more like this? The Data-Con-101 practice test has questions for every domain.
Checklist
0 of 12
Tick each one you can do without notes. Your ticks stay in this browser.
Study strategy
Spend the first part of your time building a source decision table. List each source type in the guide: Salesforce CRM, Marketing Cloud, Commerce Cloud, cloud storage, custom applications, websites and apps, and external warehouses. For each, write the connector or method, typical freshness and one setup detail that often appears in questions.
Next, practice stream design. Take three sample datasets, such as customers, orders and web visits, and decide the category, primary key and event time field for each. Explain why the wrong category would hurt segmentation later. Then complete the connectors, Zero Copy, batch transform, streaming transform and fully qualified key modules in the official trail.
If you have a playground, create a stream from a CSV file and add a formula field, then build a simple batch transform that joins two lake objects. Finally, review Zero Copy carefully: federation brings data in without copying, sharing sends data out without copying, and both carry query cost.
FAQ
How much of the exam is Data Source Connection and Ingestion?
Data Source Connection and Ingestion is 18% of the exam, about eleven of the 60 scored questions as a planning estimate, tied with Data Enhancements, Sharing, and Analysis.
Which data stream categories exist?
Profile, Engagement and Other. Engagement streams need an event time field, and the category is set when the stream is created.
When should I use the Ingestion API?
When a custom or third-party system has no native connector. Use streaming mode for small, frequent updates and bulk mode for large periodic loads.
What platforms does Zero Copy support?
Salesforce lists partners such as Snowflake, Databricks, Google BigQuery and Amazon Redshift, with federation into Data 360 and sharing out. Check current Help for the latest list and directions supported.
Do data transforms change the source data?
No. Transforms read data lake objects and write their output to new data lake objects, which you then map to the data model.
Does ingestion frequency affect cost?
Yes. Ingestion and processing consume credits, so more frequent refreshes and larger volumes cost more. Match frequency to the business need.
Other CRM & Business Applications certifications
Where the same subject shows up in other exams. Cards marked "Same subject" open the matching domain.
- HubSpot Inbound Marketing Certification Study guide
- Salesforce Certified Platform Developer I Study guide
- Google Ads Search Certification Study guide
- ServiceNow Certified System Administrator Study guide
- Salesforce Certified Platform Administrator Study guide
- ServiceNow Certified Application Developer Study guide
Test your Domain 3 knowledge
You have covered connectors, data streams, transforms and Zero Copy collaboration. See how quickly you choose the right ingestion path in practice questions.
Keep going
- Training Training Built on your Weak areas
- Practice test Questions by domain, with explanations
- Mock exam Full length, timed like the real exam
- Readiness test Find your weakest domain in minutes
- Cheat sheet Key terms and formulas on one page
- Ask the Data-Con-101 AI tutor Get any concept explained your way