All websites require a source for their data (i.e. a database), but many websites don't really need to be directly connected to a SQL database for nearly every page request or similar and can just use really simple files in S3 or something.

Many, many informational web sites do not require a database, including many non-trivial sites I've done for clients. Hugo is particularly adept for non-trivial static sites (via it's ability to use built-time API json data, etc).

Most corporate blogs with comments disabled would be fine as a filesystem, rather than a database.

JSON files in a git repo will get you really far if you want it to.