Explanation:

This question tests the consultant's understanding of the practical outcome of using the Dataflow to Recipe converter and the implications for downstream assets like dashboards.

Why C is Correct:
When you use the Dataflow to Recipe converter, it creates a brand new recipe that outputs to a brand new dataset. This new dataset will have a completely different API Name and internal ID. The consultant is likely looking at the original dashboards and lenses, which are still connected to the old dataset that was created by the dataflow. Since the dataflow is likely still running (or was only recently turned off), the old dataset is the one being updated and viewed. The new dataset from the recipe exists but is not yet connected to any dashboards, which is why they "are unable to see the aggregated data."

Why A is Incorrect:
Turning off the recipe and only running the dataflow would be a step backwards. The dataflow cannot perform the aggregation that the consultant wants to implement. This action would not solve the problem of not seeing the aggregated data; it would just revert to the old, pre-aggregated state.

Why B is Incorrect:
While it is a necessary step in the overall migration process to eventually turn off the old dataflow to avoid conflicts and duplication, it is not the cause of the issue. The core issue is that the visualizations are still pointing to the old dataset. Even if the dataflow is turned off, the dashboards will not magically update to show the new dataset's data; they must be manually reconnected.

Reference
Dataset lineage: In CRM Analytics, dashboards, lenses, and stories are tied to a specific dataset by its unique ID. Converting a dataflow to a recipe creates a new dataset in a new lineage.
Migration Process: The complete process for replacing a dataflow with a recipe is:
Use the converter to create the new recipe.
Run the new recipe to populate the new dataset.
Re-bind all existing dashboards and lenses from the old dataset to the new one.

Then, and only then, turn off the old dataflow to stop updating the old dataset. The consultant in the scenario has likely only completed steps 1 and 2, missing the critical step 3.

Explanation:

This question tests the consultant's understanding of how to manage the refresh schedule for an entire templated app and its associated data pipeline in CRM Analytics.

Why C is Correct:
Templated apps (like Sales Analytics) are designed as integrated, pre-built solutions. When you install such an app, it creates a complex, interdependent set of assets (dataflows, recipes, datasets, lenses, dashboards, and Einstein Discovery models). The App Install History page provides a centralized "master schedule" for the entire app. Scheduling from this location ensures that all the underlying components run in the correct, managed sequence. Scheduling the app itself guarantees that dataflows run first to bring in raw data, then recipes transform it, and finally, any dependent Einstein Discovery models are retrained—all automatically and in the right order.

Why A is Incorrect:
While technically possible, this is a highly inefficient, error-prone, and non-scalable approach. Manually scheduling each individual asset (every recipe, dataflow, connector, and Einstein model) is tedious. More importantly, it breaks the managed dependencies. You risk a recipe trying to run before its source dataflow has finished, or an Einstein model retraining before its source dataset is updated, leading to data inconsistencies and failures.

Why B is Incorrect:
This is an improvement over option A but is still incomplete. Scheduling only the recipes and dataflows would update the core datasets. However, it would not automatically trigger the refresh of the Einstein Discovery models and prediction definitions. These are separate assets that rely on the updated datasets. Using the App Schedule is the only method that encompasses the entire pipeline, including Einstein assets.

Key Concept
Managed App Schedules: The key concept here is that a templated app is a managed package of analytics content. The platform provides a top-level scheduling mechanism (App Install History) specifically to handle the orchestration of all its components. This is the Salesforce-recommended and most robust method for ensuring a consistent and reliable data refresh for the entire application.
Orchestration: A critical part of a consultant's role is understanding data pipeline dependencies. The App Schedule handles the orchestration automatically, eliminating the need for complex manual workflow management.

Explanation:

When CRM Analytics (data sync/direct data) pulls currency fields from Salesforce, the platform converts them by default to the org’s default/corporate currency unless you explicitly preserve original currency values. If UC’s corporate currency is USD, synced currency fields will appear in USD on datasets/dashboards even if the source records are in GBP.

FYI:
Salesforce added a setting for Salesforce Direct Data in recipes to preserve original currency values (i.e., skip the automatic conversion to the default org currency). If you don’t use that, you get corporate-currency values.

Why the others are wrong
A. “ANS local currency is USD” — There’s no CRM Analytics setting called “ANS local currency.” Not a relevant concept/config.
B. “Integration User currency is USD” — The authoritative behavior is that data sync converts to the org’s default/corporate currency by default, not the integration user’s personal currency. (Some blogs/community posts mention the integration user, but Salesforce’s own docs state conversion is to default org currency.)

Reference
Salesforce Help (Release Notes): Preserve Original Currency Values for Salesforce Direct Data — “By default, all currency fields are converted to the default org currency on all data syncs.”
Salesforce Help: Set Your Corporate Currency — explains corporate currency for multi-currency orgs.

Explanation:

This question tests the understanding of how to assign CRM Analytics licenses and permissions efficiently and in accordance with Salesforce best practices.

Why C is Correct:
Permission sets are the standard and recommended way to grant granular access to features in Salesforce without modifying user profiles. In this scenario, the client has three distinct user personas:
Builders (3 users): Need CRM Analytics Data Manager and CRM Analytics Dataflow Manager permissions to build datasets and dashboards.
Dashboard Creators (15 users): Need CRM Analytics Creator permission to create dashboards but not manage datasets.
Viewers (~982 users): Need CRM Analytics Consumer permission to only view dashboards.

The consultant should recommend creating three separate permission sets, each containing the corresponding CRM Analytics permission license (Data Manager, Creator, or Consumer). These permission sets are then assigned to the respective users. This is scalable, manageable, and follows the principle of using permission sets for functional access rather than creating multiple profiles.

Why A is Incorrect:
App permissions (Viewer, Editor, Manager) control what a user can do within a specific Analytics app (e.g., view, modify, or manage the app's dashboards and datasets). However, these permissions are secondary. A user must first be assigned a CRM Analytics permission license (via a profile or permission set) to even log in to the Analytics Studio. You cannot assign app permissions to users who do not have a base license. Option A addresses the second step without solving the fundamental licensing requirement.

Why B is Incorrect:
While creating three new Salesforce profiles would technically work, it is considered a poor practice and is not scalable. Profiles are complex and control a very wide range of system permissions and settings across the entire Salesforce org. Creating multiple profiles for the sole purpose of managing CRM Analytics access is an administrative burden and can lead to unnecessary complexity in the overall user management system. Permission sets are the modern, flexible, and recommended tool for this specific purpose.

Reference
Salesforce Help: Assign CRM Analytics Permissions Licenses
This documentation outlines the process, stating: "You grant access to CRM Analytics by assigning permission sets that contain CRM Analytics permissions licenses to users." It explicitly recommends using permission sets to assign the core licenses: CRM Analytics Consumer, CRM Analytics Creator, CRM Analytics Data Manager, and CRM Analytics Dataflow Manager.

Key Concept: The process is a two-step assignment:

License Assignment: A user gets a functional license (Consumer, Creator, etc.) via a Permission Set.
App-Level Permissions: After being licensed, the user is then granted specific permissions (Viewer, Editor, Manager) within individual Analytics apps to control what they can see and do inside that app.

Explanation:

In CRM Analytics, when working with datasets that include missing values, especially in fields like Income that may be sparse but highly predictive, the best practice is to use the “Predict Missing Values” transformation in recipes.

This transformation:
Uses machine learning to estimate missing values based on patterns in other fields.
Preserves the column for modeling while improving data quality.
Is ideal when the field is informative but incomplete—like Income in this case.

Since the consultant believes Income is valuable for predicting Customer Lifetime Value, dropping it would reduce model performance. Predicting missing values is a scalable and intelligent way to retain the feature.

❌ Why the other options are incorrect:
Option A: Filling with the average is a simplistic method that can introduce bias and reduce variance, especially when only 1% of values are present.
Option C: Dropping the field discards potentially valuable predictive information, which contradicts the consultant’s belief that it’s informative.

References:
Salesforce Help: Predict Missing Values Transformation
Trailhead: Prepare Data with Recipes