MongoDB and other noSQL databases became popular around 2010. They were web-scale. It was the new, the shiny and sense everyone in the startup world would build an app for millions of users they might as well start using noSQL right away.

Regular SQL databases felt old and you did not want to admit you where running SQL in your stack. That was not web-scale!

noSQL required no planning. Just start with an empty {} and start adding your data. Quick and easy to get started!

Short term losses offset long term gains

Everyone wanted "web-scale", but at the same time their app had at most 2000 users. This does not make sense.

Libary of Stuttgart
Photo by Gabriel Sollmann / Unsplash

Google Firebase

Google Firebase is for developers to build apps. Google scale apps. Currently there are 2M active Google Firebase apps with a Google scale backends.

This is of course amazing. Google has made it easy for anyone to have a Google Scale backend. But how many of those 2M apps really need Google Scale?

Yes, your data in Firebase can be queried super fast. But is it worth limiting your query capability because you want to save a few milliseconds?

No joins, no complex WHERE, not data consistency, no transactions.

There is significant downsides to Google Firebase if your app does not need Google scale. You are stuck with the downsides of noSQL and do not benefit from the upsides of Google scale.