The administrator must enable Collaborate with Slack and Tableau Next, making A correct. Salesforce's documented setup requires connecting the Salesforce org to Slack and then navigating to Tableau Next Administration settings to confirm that Collaborate with Slack and Tableau Next is enabled.

For Inspector Proactive Data Alerts specifically, Salesforce states that Slack must first be enabled for Tableau Next. Users can then receive Inspector alerts through Slack after the Agentforce app and associated connection are configured.

Option B is not the documented prerequisite. Tableau Next relies on appropriate Tableau Next permissions plus Salesforce/Slack identity and connection configuration; assigning a generic “Slack User” permission set is not the required answer.

Option C is also inaccurate as written. A properly configured Analytics and Visualization agent must contain the relevant Data Analysis and Data Alert Management capabilities, but administrators do not necessarily need to create an entirely new agent solely because Slack is being enabled; an existing qualifying agent can be configured and activated.

References/Topics: Embedding, Cross-Cloud, and Interoperability - > Tableau Next with Slack - > Inspector Alerts - > Collaboration Settings - > Agentforce App.

===========================================================================

Enable Query Cache is the appropriate short-term performance optimization for a frequently accessed Tableau Next dashboard. Query caching reduces repeated execution of expensive live analytical queries by temporarily reusing previously calculated query results.

Salesforce specifically positions dashboard query caching as a mechanism to accelerate page load times for frequently visited dashboards. Current Tableau Next functionality can store and reuse query results for up to approximately 30 minutes, thereby reducing system overhead and improving the end-user experience.

This is precisely appropriate when the consultant wants a rapid tactical improvement before investigating deeper issues such as semantic-model complexity, source performance, query design, data architecture, or excessive dashboard density.

Disabling Reflow concerns dashboard layout behavior and responsive presentation rather than query-processing latency. Turning off Tableau Agent likewise does not address the underlying query execution performed to render dashboard widgets.

Caching does involve a freshness tradeoff because viewers may temporarily receive cached rather than newly executed query results. For dashboards requiring immediately current information, live mode may still be necessary. The consultant should therefore treat caching as a deliberate performance-versus-freshness decision.

References/Topics: Visualizations and Dashboards - > Dashboard Performance - > Query Cache - > Cached Data Mode.

===========================================================================

The semantic model should be enriched so the agent can associate the organization's business language “Earnings” with the existing Profit Margin concept. Therefore, B is the correct answer. Semantic models provide the business-context layer that allows Tableau Agent to translate natural-language terminology into governed analytical definitions. Salesforce recommends clear, contextual names and descriptions and explicitly highlights synonyms as an important semantic consideration for agent readiness.

Salesforce's Tableau Semantics training also explains that synonyms broaden the agent's understanding of terminology used by business users, allowing alternate terms to map back to authoritative metrics and definitions.

Creating a duplicate metric merely to support another phrase introduces redundant semantic definitions and increases ambiguity. Salesforce specifically recommends eliminating redundant or overlapping calculated definitions rather than multiplying them. Changing aggregation from Average to Sum would alter mathematical behavior and does nothing to resolve a terminology-discovery problem.

References/Topics: Data Setup - > Create Semantic Models - > Design Semantic Models for AI Readiness - > field descriptions, synonyms, ambiguity resolution.

===========================================================================