Your Folder Structure Is a Mental Model

I reorganized 3,000 files last weekend. Not because the system was broken, but because I realized it had been designed for the wrong audience.

The original structure of my vault reflected how Notion syncs data. There was a bills/ folder at the root because Notion's Bills database synced there. A transactions/ folder because Notion's Transactions database synced there. An agencies/ folder for Florida state agency profiles, a people/ folder for contacts. Each folder existed because a tool needed a place to land its output, not because it reflected how I think about information.

The cost was subtle but real. Every time I needed to find something, there was a small moment of friction. Where does the DEP agency profile live — agencies/ or reference/? Where's a specific person's notes — people/ or crm/? Where do bill records go — bills/ or finance/? Each question was answerable, but answering it required translation. The structure was speaking a different language than my brain.

The Test That Reveals the Problem

There's a simple test for whether a folder structure is designed for you or for the tool. Close your eyes. Think of a piece of information you need. Now open your eyes and ask: where would you go first?

If you'd go to the right place instinctively, the structure is working. If you'd hesitate — if you'd think "it might be here or it might be there" — the structure is working against you. Hesitation is the friction signature of a misaligned mental model.

My vault failed this test repeatedly. I knew intellectually where everything was, but I had to translate. That translation is invisible overhead, the kind you don't notice until you remove it.

Designing for the Mind, Not the Tool

The reorganization principle was simple: group things by how you think about them, not by how they sync. This sounds obvious, but it's harder than it appears because tool constraints constantly push you toward their model.

The most clarifying decision in the reorganization was merging people/ and agencies/ into crm/people/ and crm/agencies/. These are both the same thing — relationship profiles. A person profile and an agency profile contain the same types of information: who they are, what you know about them, when you last engaged, what matters to them. The fact that one is a person and one is an organization is a secondary detail. The primary category is "relationship I manage."

Once I saw it that way, the folder wrote itself. CRM is how I think about these records. Whether the subject is a person or a 30,000-employee state agency is irrelevant to the filing logic.

The same logic applied to finance. bills/, transactions/, and a scattered finance/ folder became finance/bills/, finance/transactions/, and finance/ledger/. When I think about my financial situation, I think about finance as a domain. Bills and transactions are sub-dimensions of that domain. The folder structure now mirrors that hierarchy.

The Sync Problem

The deeper issue is that tool-driven organization creates a system that works well for the tool but poorly for the human. Notion syncs to bills/ and it works perfectly. But when I'm looking for financial data, I'm not thinking "Notion Bills database" — I'm thinking "finance."

This isn't specific to Notion. Every system with sync behavior, import pipelines, or automated filing has this tendency. The tool puts files where it needs them. Over time, the human adapts to the tool's logic rather than the other way around. The friction accumulates silently.

The fix isn't to stop using tools — it's to periodically audit whether the structure still reflects your mental model. The question isn't "is everything in a folder?" It's "if I woke up tomorrow with amnesia and needed to find my rent payment records, would I go to the right place?"

What the Reorganization Actually Changed

After 3,000 file moves and several hours of work, the structural change was this: from 24 root-level folders to 12 domain-grouped folders. Each folder now has a single clear purpose. The test I described above — thinking of information and going to the right place — now passes consistently.

The practical impact shows up in the small moments. I don't hesitate before navigating. I don't maintain a mental index of "bills is here but financial analysis is there." When I want to understand my financial situation, I go to finance/. When I want context on a Florida agency, I go to crm/agencies/. The structure matches how I think.

That's the only design criterion that matters for a personal knowledge system. Not elegance, not symmetry, not tool compatibility. Just this: does the structure match how your mind retrieves information?

If it doesn't, no amount of plugins, templates, or automation will fix the underlying friction. The folder is a mental model. Make sure it's yours.