Explanation:

In Salesforce Data Cloud, every Data Model Object (DMO) requires a primary key to uniquely identify each record. If the source data doesn’t have a single field that can reliably serve as a primary key (i.e., there are no unique identifiers), the best practice is to:

→ Create a composite key
This involves combining two or more fields that together can uniquely identify a record — for example, combining email + account_id, or first_name + last_name + birthdate.

You can achieve this in Data Cloud by:

1. Creating a calculated field (formula) on ingestion or during data transformation.
2. Marking that field as the primary key.

This ensures that the identity resolution and deduplication processes in Data Cloud function properly.

🚫 Why not the other options?

A. Use the default primary key recommended by Data Cloud
❌ No "default primary key" exists unless one is mapped from the source. Data Cloud does not auto-generate meaningful unique keys.

C. Select a field as a primary key and then add a key qualifier
❌ A key qualifier (like Email, Phone, etc.) helps with identity resolution, but it doesn’t solve the problem if no field is unique. Choosing a non-unique field would cause data quality issues.

D. Remove duplicates from the data source and then select a primary key
❌ Data Cloud is designed to handle deduplication and resolution internally. Manually removing duplicates is not scalable and doesn’t fix the issue of lacking a unique identifier.

📘 Reference:
Salesforce Help: Define Primary Keys in Data Cloud

Best Practices for Identity Resolution:
“If no field is unique, create a calculated composite key from multiple fields.”

Explanation:
A low consolidation rate in Identity Resolution typically means that many individual profiles are not being unified into fewer unified profiles because the current matching rules are too strict or too few. To increase the consolidation rate (i.e., unify more records), the consultant must broaden the opportunities for matches to occur. Adding more matching rules with different attribute combinations gives Data Cloud additional ways to find matches, thereby increasing the likelihood that records are consolidated without sacrificing data quality.

Correct Option:

B. Increase the number of matching rules.
Creating additional matching rules (e.g., one rule on Email only, another on Name + Phone, another on Name + Address, etc.) provides multiple independent paths for unification. Each new rule acts as an “OR” condition; if any single rule finds a match, the records are unified. This is the most effective and recommended way to raise consolidation rates when the current rate is lower than expected.

Incorrect Options:

A. Change reconciliation rules to Most Occurring.
Reconciliation rules control which attribute value wins when multiple sources conflict (e.g., Most Recent, Source Priority). They have no impact on whether records match and consolidate in the first place; they only affect the surviving value after a match occurs.

C. Include additional attributes in the existing matching rules.
Adding more attributes to an existing rule (e.g., requiring Email + Phone + Name instead of just Email) makes that rule stricter, which usually decreases matches and lowers the consolidation rate.

D. Reduce the number of matching rules.
Fewer rules remove possible match pathways, making unification harder and almost always reducing the consolidation rate.

Reference:
Salesforce Help: “Identity Resolution Ruleset Overview” and “Best Practices for Improving Match Rates” – explicitly states that “adding more matching rules with different field combinations is the primary method to increase unification rates.”

Explanation:

In Salesforce Data Cloud, if a user wants to pause a segment activation but keep the segment available for future use, they should:

→ Stop the publish schedule
This action halts the scheduled activation of the segment to external destinations (e.g., Marketing Cloud, Advertising platforms), but it does not delete or deactivate the segment itself. The segment remains in the system and can be re-activated or scheduled again later.

🚫 Why not the other options?

A. Deactivate the segment
This removes the segment from being evaluated — it’s no longer processed. You’d need to reconfigure it to reuse. Not ideal if you want to “pause.”

B. Delete the segment
Deletes the segment permanently — this is irreversible and definitely not suitable if you want to use it again.

C. Skip the activation
This option doesn’t exist in Data Cloud as a formal action. You can’t just “skip” one activation; you must either unschedule or pause it by stopping the schedule.

📘 Reference:

Salesforce Help: Manage Segment Activations in Data Cloud

Key tip from Salesforce Docs:
“You can stop a segment’s scheduled activation at any time. This doesn’t delete the segment or its criteria, only the scheduled delivery.”

Explanation:
When related attributes are added to a segment in Data Cloud, the system includes all related records by default, not just those filtered by the segment criteria. This can result in activations containing orders older than intended. To ensure only recent orders (e.g., last 30 days) are included, the consultant must apply a filter directly on the related attribute, restricting the records used in the activation to the desired date range.

Correct Option:

C — Apply a filter to Purchase Order Date to exclude orders older than 30 days
By adding a filter on the Purchase Order Date field within the segment or activation configuration, Data Cloud ensures that only orders within the last 30 days are included. This prevents older transactions from being activated, aligns the activation with business rules, and maintains data accuracy for marketing campaigns.

Incorrect Options:

A — Use data graphs that contain only 30 days of data
Creating a separate data graph for 30 days of data is unnecessary and inflexible. It adds complexity without solving the root issue, which can be resolved with proper filtering at the segment or activation level.

B — Apply a data space filter to exclude orders older than 30 days
Data space filters control access or segregation between data spaces, not attribute-level filtering. They cannot restrict the records in a segment activation based on order dates.

D — Use SQL in Marketing Cloud Engagement to remove orders older than 30 days
While technically possible, applying SQL in Marketing Cloud is reactive and inefficient, as it filters after activation. Best practice is to filter at the Data Cloud level to ensure correct data is activated from the start.

Reference:
Salesforce Data Cloud: Activations and Related Attribute Filtering

Explanation:
The core requirement is to structurally transform the source data during its journey into Data Cloud. The source object has two distinct concepts (hotel points, airline points) in a single record that need to be separated. This is a classic data processing task that occurs after ingestion but before the data is modeled for use in segments and insights. The solution must actively split and create new records.

Correct Option:

B. Use batch transforms to create a second data lake object:
This is correct. Batch Transforms in Data Cloud are designed for this exact purpose. A consultant would recommend creating a transform that reads the original ingested data lake object and uses logic to split each source record into two new records—one for hotel points and one for airline points—outputting them to a new, separate data lake object.

Incorrect Option:

A. Clone the data source object:
Cloning the object, whether in Salesforce CRM or during ingestion, would merely duplicate the problem. It would create an identical copy of the data without solving the fundamental issue of splitting the two point systems into separate records.

C. Create a junction object in Salesforce CRM and modify the ingestion strategy:
This overcomplicates the solution by requiring schema changes and data migration in the source system (Salesforce CRM). Data Cloud's transformation layer is built to handle such structural changes without imposing development work on the source system.

D. Create a data kit from the data lake object and deploy it to the same Data Cloud org:
A Data Kit is used to package and transport data model components between orgs (e.g., from sandbox to production). It does not perform the active data processing required to split records within the same org.

Reference:
Salesforce Help - "Transform Data in Data Cloud"