Launching apps, from idea to the play-store is feels amazing.
Now that non technical founders can also leverage AI to build their own apps, thanks to the vibe-coding, reality is a little different.
If you have ever built or shipped an app, you will agree that there is more to this.
You want your app to be profitable, bug free and honestly your architecture, to be scalable.
For most b2c apps, a serverless architecture is more than enough provided you deal the right deck.
I will be blunt and open, if your app needs data and your data has a structure, links with each other and you need aggregates, you need a proper database, since there is no cost optimized solution for a nonSQL db in my opinion (not talking about AWS, or other cloud solutions), you can close your eyes and place your best on PGSQL with supabase.
Supbase data-api is perfect and you can manage your data structure with migrations.
If your app doesn't rely on data that much, you can use firestore, keep in mind firestore charges for read/write, meaning if you just read 10000 records in a go, everytime, it will hit your wallet, even though the actions itself are very cheap.
I propose the using a combination of both supbase & firestore. With this combo you can cater any production grade b2c app needs and scalabilty won't be an issue ever.
Tip: If you are using firebase, think in reads/write and if you are using pg SQL think how largw the dataset is. Also queries with PGSQL can be very optimized, enhacing user experience with your app. Firebase is a heaven for apps that needs occasional reads or writes and need to store data in key/value pairs.