During the GTD craze of 2005, I came up with a system that has served me well over the years. My primary goal is to minimize the amount of time I spend looking

During the GTD craze of 2005, I came up with a system that has served me well over the years. My primary goal is to minimize the amount of time I spend looking for files. Rule #1: All data must be stored in a single root folder. As others on this thread have mentioned, this makes it easy to backup data and move it around. On Windows, this means I put everything in: * `c:\data` File systems are typically treated as hierarchies, so we now have to decide the most important aspects of data files that can be used to group and organize them. I believe that ownership is of first-order importance. Is this a file that I personally own, or is it a product for my employer? Is this a file I want to grant my family access to, or my wife? By separating things by owner first, it makes it easy to apply sensible permissions to data, apply organization-specific data retention policies, and copy it around when relationships change. * `c:\data\acme-corp` * `c:\data\benjamin` The next most important aspect is source of truth. Is my machine the primary storage location for this file, or am I keeping it sync with other machines? This info is critical for setting up automated jobs that safeguard your data. * `c:\data\acme-corp\columbia` (the name of my machine) * `c:\data\acme-corp\git` The final decision to be made is whether the data is "reference" data or "project" data (this is a GTD concept). An example of reference data would be your resume -- you want it to always be handy and current. But really, 99% of my data ends up being project data -- data that's very important to me right now, but will probably of little importance in six months. So, my reference data is typically a handful of files single folder, and my project data is stored in a series of monthly folders. * `c:\data\acme-corp\columbia\projects\2018\12` * `c:\data\acme-corp\columbia\projects\2019\01` * `c:\data\acme-corp\columbia\reference` I absolutely love the monthly folders. I create the new monthly folders as a I need them, and older files silently slip out of my view as time goes on. Months later, when I'm searching for a file for an older project, I typically find several related files bunched together, which is hugely helpful. And organizing data by date makes it simple to adhere to corporate data retention policies. Beyond that, I create sub-folders as needed, if there's a lot going on in that area that month. * `c:\data\acme-corp\columbia\projects\2019\01\financial` * `c:\data\acme-corp\columbia\projects\2019\01\legal`