No video of the event yet, sorry!

PostgreSQL was among the first relational databases to introduce a JSON data type, emerging at roughly the same time MongoDB was gaining traction as a NoSQL document database. While both PostgreSQL's JSONB and MongoDB's BSON formats store JSON, they differ significantly in their implementations and goals: PostgreSQL adheres to SQL, stores data as rows within blocks, and relies on extensions to provide a general purpose database, whereas MongoDB provides native features designed specifically for document-oriented data models. As PostgreSQL's popularity grows, an important question arises: can it effectively replace a native document database, or is it more advantageous to use JSONB in combination with a relational model?

Both databases have evolved considerably in recent years, yet some long-standing myths persist. This talk will compare PostgreSQL with JSONB and MongoDB for document-oriented use cases, focusing on indexing flexibility, update performance, transactional behavior, schema adaptability, and data locality in physical storage.

Our aim is not to declare a "winner" but to offer practical guidance on choosing the right approach for your application — whether that means tables, documents, or documents within tables. We will examine how to leverage each system's strengths while avoiding misconceptions born from their superficially similar JSON interfaces but fundamentally different architectures. We will also discuss what PostgreSQL would need to improve for JSONB to compete fully with MongoDB's native document capabilities.