Your are on to a good start. All advice here is good. End of the day, how you do it has to suit what you do, can do, and need to do. You can start and achieve paperless without a lot of messing with fancy computer tools. Exploit the tools to your benefit when the need and value is clear.
I’ve been pretty much paperless for years and in a nutshell, I would offer the following:
: For me, tagging is probably a waste of time. Feels good to tag, and I do it; but I never use them.
: Having a deep folder structure is probably also a waste of time. Feels good to have a structure, and I probably over-do it. Deep folder hierarchy has deep history when everything paper involving many clerks and admin staff with file cabinets, drawers, and filing folders in the drawers. Probably started by the project team that built the pyramids! Those days gone. Simple folder structure and then rely on that and computer navigation (Finder) and search (Spotlight et. al.) to find stuff.
: Remember that the containing folders are not “sticky” to a document if the folder name contains important information to identify the file. File names are the best you have to label the document and remain unchanged, normally. When files move around and if they depend on the folder names above to describe the document, then key information is lost. I use Hazel to help with naming files, but sometimes I name/rename manually. As I’ve used Hazel for years, I of course have some complex rules. Make Hazel shine is fun. Notionally I name all files something like: [Scan or Incoming Date] [Identifier] [Description including document date if appropriate].pdf. PDFs are OCR’d. I rely on the computer’s search to find things more than hunting through folder structure.
[Scan or Incoming date] is YYYYMMDDHHMMSS as can’t depend on file system dates “sticking”. This format is sortable.
[Identifier] is normally the same as the folder targets (to help me or Hazel move, and may include initials of person if personal
[Description] as it says. Vendor, what it is, for what, etc. If a bill/invoice, I’ll include the date in readable format, e.g. 4 June 2019 or something. Be descriptive. Hazel can be setup to extract all that, but easy to go over the top.
I’ve learned that a few top-level folders is best. I minimise use of subfolders for reasons as described above.
~Personal/ with a subfolder for each person in the family with stuff pertinent to them. Subfolders by person are as appropriate. Some of these subfolders by person are shared across network with that person.
~Finance/ with a subfolder for each year. I dump everything for that year in the relevant year folder. No other subfolders (relying on file name to distinguish the files)
~Interesting/ with a few subfolders of interest area.
I have a few more top-level folders which have emerged over the years but too detailed to explain here. I hope you get the gist.
The Finance and Interesting documents are actually now imported fully in to DEVONthink databases (one for each). A third DEVONthink database is “Personal” which is mostly indexed back to the ~Personal/ folder–I do that so that I can keep sharing with the family members secure and easy. they don’t use (yet) DEVONthink. I’ll use DEVONthink for searching/finding/etc.
DEVONthink will use the tags that were assigned into the OSX file system, but yet again, i find that while I do it, I never use them. Gotta stop, I think. DEVONthink’s structure (Groups, Smart Groups, etc.) is great.
You of course don’t need to use DEVONthink, especially at start. If you chose to get there, you can drag and drop your files to import (trivially easy), or you can index pointers back to the files still residing in the file system. All this described fully in DEVONthink’s terrific manual.
And finally, as someone else mentioned, files inside DEVONthink are still in native format and not proprietary as other applications. Important, I think.