For a Visit Recommendation implementation that must scale to production business volumes, the recommendation-selection logic should be implemented using an Apex class.
Salesforce supports Flow-based strategies for Einstein Visit Recommendations, but its documentation explicitly states that the flow-only approach is not scalable for a business use case and recommends using Apex for scalable Visit Recommendation processing.
With Apex, a developer can implement the business logic needed to identify the Retail Stores for which Einstein should generate recommendations. The Apex method can be exposed through an invocable action and incorporated into the recommendation strategy or an associated flow. This architecture provides more control over bulk processing, complex filtering, governor-limit management, reusable logic, and enterprise-scale execution.
Salesforce ' s Visit Recommendation setup documentation supports both Flow and Apex approaches. It specifically provides an Apex-based option to select the list of stores for which recommendations should be generated and then use that Apex action within the recommendation framework.
Workflow Rules and Process Builder are not the supported architectural mechanisms for implementing scalable Einstein Visit Recommendation strategies.
Study Guide Topics:Visit Planning - > Einstein Visit Recommendations - > Apex - > Scalability - > Recommendation Strategy
===============
