Whoo boy is that an understatement! 
@MardeeB First, welcome!
I made a pretty detailed couple of comments about how I structure US & State censuses recently on this post, so I won’t re-hash that.
If I could offer a single Cliff Notes nugget of wisdom … decide on what you consider sources versus repositories. That decision will affect a lot.
I for example, consider Ancestry, Family Search, MyHeritage, etc to be repositories for 95% of their “records”. They didn’t originally create the document / record … they only store it like a library or warehouse. The exceptions have to do with specific indices they’ve actually created themselves. That viewpoint, in turn, means that any records / citations / “sources” I get from those gets re-worked.