In relational databases, data is stored across multiple normalized tables with rigid schemas. In contrast, MongoDB uses flexible, JSON-like documents to store related data together.
Data modeling in MongoDB prioritizes understanding application workloads and modeling relationships between entities, following MongoDB’s “golden rule”—data that is accessed together should be stored together. This minimizes retrievals and leverages MongoDB's dynamic document model to improve performance and better enable scalability.
If you're transitioning from a relational database, MongoDB’s Relational to Document Model Skill Badge can help you learn the document model, workload analysis, and relationship modeling in MongoDB.