data-modeling

Use Master-Detail for tightly coupled objects requiring cascading deletes and roll-up summaries, and Lookup for loosely linked records. Salesforce data modeling decisions depend on whether you need strict data integrity or independent record sharing and deletion behaviors.

To model many-to-many relationships, create a junction object with two Master-Detail fields referencing the parent objects. This Salesforce junction pattern ensures data integrity while allowing flexible linking and reporting across both parent record sets.

External IDs mark custom fields as unique identifiers to prevent duplicate records during data migration and integration. Using External IDs in Salesforce data modeling allows upsert operations to match incoming source data against existing records efficiently.

Use Record Types when objects share the same core fields but need different page layouts and picklist values, and separate objects when data structures fundamentally diverge. Choosing correctly prevents data skew and maintains Salesforce performance.

Big Objects manage massive data volumes for archival and audit purposes in Salesforce, but carry strict indexing and querying constraints. Salesforce data modeling with Big Objects requires careful planning of indexed fields to avoid performance limitations.

Data skew occurs when too many child records are associated with a single parent record, causing locking and performance issues. Prevent Salesforce data skew by distributing ownership and relationship patterns across multiple parent records.