Heavy user of Postgres and Mongo. I use both in production for years. I don't think that there are good or bad databases. Every database has a different focus a

Heavy user of Postgres and Mongo. I use both in production for years. I don't think that there are good or bad databases. Every database has a different focus and thus respective trade-offs. You know this feeling when you have an idea and want to create a quick prototype to test the idea? By quick I mean you want to have this thing running in a few minutes without any effort. This is possible both with Mongo and Postgres. But with Mongo you even don't need to create databases or collections (tables) before. You just write to the DB and everything will be created if not there yet. This small thing let the code stay super simple and no migrations are required. With Docker and the official Mongo image you can setup a webserver with Mongo with authentication in seconds. Accessing the DB without an ORM is still super comfortable since everything is JSON (in case you work with Node and the Mongo native driver). So, one of the biggest reasons for me is that I start after all and not find an excuse not to start this prototype. Because I am lazy and most time for creating a prototype is always spent on setting up the development environment. This is also why people like JSFiddle and Codepen so much: Just start coding. It's great for prototyping and most of my prototypes stay prototypes. The few which got successful got migrated to a more mature architecture. Sometimes they are employing Postgres, sometimes Cassandra, but often enough they stay with Mongo but including validators, etc. All successful projects I started, began quick and dirty and slowly evolved into a mature architecture. The best projects were those where I had a prototype running in less than two hours. It's really the use case. I think disliking a technology just limits your options.