Tagging help... how best to build a list of tags... looking for precedents


Hi there,
I’ve got my scanscnap ix500 sitting here and MacSparky paperless iBook and a copy of Hazel and am all set to jump in and embrace the paperless workflow… except for one little thing… tags…

Anyone have recommendations about how best to come up with a list of tags? Or a list available to borrow from and customize? I don’t want to get into ISO dates, but instead am trying to avoid reinventing the wheel when it comes to tag lists. But maybe this is silly since every person is different… a distinct possibility. In which case, maybe the question is whether anyone has suggestions about how best to build a list of tags… Mindnode brainstorms? Or just start somewhere and then refine over time?

I am an advocate of letting your tag list naturally develop over time. Start with your current needs. Then once your system breaks and a tag doesn’t accurately describe the item you’re tagging, add more tags to meet the needs you have for that information.

It would be ideal to have this on a regular review cycle, too, so you can clean up (merge, rename, delete) tags that may need adjusting.


I don’t have specific tags to recommend. I think the specific words have to be personal, but there will likely be common categories of tags.

If you haven’t already, listen to mpu 390 with Brett Terpstra. Also, read the links in the show notes particularly Brett’s.

I use Brett’s Hazel & Ruby scripts for automatic filing of tagged items from the desktop. As a result all my tags start with a colon (“:”). So all my tags are of the format :tagstring. This has the advantage that you can imbed tags in any data store which has text, even if it doesn’t explicitly support tags, and spotlight, launchbar, and the like can find your tagged items. In all other uses of a colon in general text entry I am always following the colon with a space, so miscellaneous documents don’t tend to show up in my tag searches.

As a result of adopting Brett’s tools and methodology, I have also moved toward having much shallower folder structures than I used to maintain. I am really liking that. It’s much simpler to find things and stay organised.

My tagging system grew from my old hierarchical filesystem. Currently I have three folders in my Documents folder

  • Action - Stuff I need to deal with
  • Working - Stuff I am dealing with
  • Archive - Stuff I’m no longer dealing with, but want to keep anyway.

Inside each folder are all my files, no more nested folders. My Archive folder is fairly good sized. Before I had a filesystem that looked more like this:


  • Personal
    • School
    • Home
    • Kids
    • Vehicles
  • Finance
    • Bills
      • water
      • power
      • water
    • Insurance
      • health
      • dental
      • life
      • auto
      • home
    • Receipts
    • Statements
      • bank 1
      • bank 2
        • 1999
        • 2000
        • 2001
      • Insurance company
  • Work
    • Project
    • Reviews
    • Another Project

You get the picture. The thing is, sometimes things should go in two folders, like insurance statements, should those go under “insurance” or under “Statement/Insurance”? I finally got tired of the endlessly recursive hierarchy and converted all the folder names to tags.

So, now I’ve got tag names like “Finance”, “Bills”, & “Water” applied to my scanned water bill, and “Finance”, “Insurance”, “Statements”, & “Life” applied to my life insurance statement. Everything gets tagged and filed into the Archive folder by Hazel after scanning and OCR.

Now it’s no big deal if a particular file should exist in two places because those places are just tags, and I can apply as many tags as I want. Search works great for finding files, but so does the Finder’s “Tags” section in the sidebar. Combining the two I can always find exactly what I need.

One place to look, one place to backup, one place to file everything that I want to keep.

1 Like

Have a think about what you are going to use the tags for. Also if others will reference the files would their requirements be different?

A Folksonomy (free format tagging as you go) helps with searching where you would tags file with anything you might reasonably search on and want that file returning; often excluding any text or meta data content that would be found instead - I.e. Avoid duplicating as a tag when it exists as file data already.

A Taxonomy (strict, predefined tags) helps with grouping for filtering since you are effectively categorising like you might with a file structure, except tagging is not beholding to hierarchical structure or duplicate association to a single source - I.e. a file exists in one folder but can exist in many groups as defined by many tags.

Beyond that it really is your choice. Is it for work, home, hobby, voluntering/charity, all of the above, something else…? The choice of what you might use to tag with is dependent upon the topic areas being covered and what makes sense to everyone who might need to use them.

Relistened to Podcast again on tagging.

Resurrected a decade-old program “LEAP” that still from what I can see is the best tagging software utility


This is a handy program that also maps all tags in a keyword diagram with size of the tag corresponding to the number of times used in the files.

Hoovering over Tags displays the number of times used