A Year Into PARA, and the Archive Is the Only Part I Still Use

PARA sorts everything you keep into Projects, Areas, Resources and Archive, and the appeal is that the categories are defined by actionability rather than subject, which solves the classic problem of a note that belongs in three places. I adopted it properly in September last year with about 200 existing notes, and I now have around 900. Here is what a year of real use did to the four buckets.

Projects worked exactly as described and I have no criticism. A project has an end, the folder empties when it ends, and having a container that is supposed to become empty turns out to be psychologically different from every other kind of folder. This is the part of PARA I would keep if I could only keep one.

Areas is where it started to go wrong, and I think the problem is structural rather than a mistake I made. An Area is an ongoing responsibility with no end date — health, finances, a team you manage. The trouble is that notes in an Area accumulate forever and are never reviewed, because review is triggered by completion and there is no completion. By month seven my Areas folders had become exactly the undifferentiated pile the system was supposed to prevent, only now with a respectable name.

Resources — reference material not tied to a responsibility — merged into Areas in practice within about four months. The distinction requires you to decide, at capture time, whether something is connected to an ongoing responsibility, and at capture time you usually do not know. I would save an article about scheduling and it could plausibly be a Resource or part of my team Area, and the decision took a real few seconds, and after a few hundred repetitions I stopped making it carefully and started putting everything in whichever I had open.

The Archive is what survived, and it survived for a reason I did not anticipate: it is the only bucket where the entry criterion is unambiguous. Something goes in the Archive when it is done or dead. There is no judgment call, so there is no drift, and consequently the Archive is the only folder in my system whose contents mean what the folder says. About 600 of my 900 notes are there now.

What I actually do today is closer to two buckets than four: active, and archive. Everything I am currently working on is in one flat space, and when it stops being current it goes to the archive with a date. Retrieval is by search rather than by location, which was the part I resisted longest and which works better than the filing did, because my search terms are the words I actually used and my folder names were the words I thought I would use later.

I want to be fair to the method, though, because “I stopped using three of the four parts” reads as a failure and I do not think it was. Going through the exercise of classifying by actionability changed how I decide what to keep at all, and my capture rate dropped by roughly half in the first two months — not because a rule told me to be selective, but because having to answer “is this a project, an area, or a resource” made it obvious how many things were none of them and were being saved out of a vague sense of maybe.

If I were advising someone starting today, I would say: use Projects and Archive, put everything else in one place, and revisit in six months. The two buckets with clear entry criteria will still be clean. The ones that require a judgment call at capture time will not be, and that is not a discipline problem.