Why Are We Still Using Folder Trees in 2026?

Search has changed how we find information, but folders still remain.

It is a question I have been quietly circling for a while. In 2026, with powerful search, indexing, and metadata available almost everywhere, the idea that we still navigate files by drilling down through a hierarchy of folders feels oddly old fashioned. The thought started with email.Last year I read a book called Uptime by Laura Mae Martin, who works at Google as an Executive Productivity Advisor. One idea from that book completely changed the way I handle email.

What email taught me about search versus structure

For most of my career I organised email the way many of us were taught to organise files: folders. Client folders, project folders, internal folders. The logic seemed obvious. Structure helps you find things later.

But Martin challenges that assumption. Instead of carefully sorting email into the one correct place, she recommends pushing everything into a single “Done” bucket and relying on search when you need something again. Once you stop trying to categorise every email, you realise how powerful modern search actually is. You can search by sender, by keywords, by attachments, by date, by conversation participants. The number of possible dimensions is far greater than any single folder structure could capture.

I have been doing this for more than a year now, and it has quietly transformed how I deal with email. I spend less time filing things and more time just getting through the work. When I need something later, I search for it. Which leads to the obvious follow-up question. If search works so well, why are we still organising files in folders?

Why folders felt natural in the first place

Part of the answer is historical. The folder metaphor is older than most people realise. Early graphical operating systems deliberately mirrored the objects people already understood from office environments: documents, folders, filing cabinets. The metaphor worked because knowledge workers already knew how filing cabinets behaved.

But the deeper reality is that hierarchical file systems existed even before graphical interfaces. Operating systems like Multics and later Unix structured files in directories long before we were dragging icons around on a desktop. That hierarchy also became the naming system for files themselves. A file is not just an object, it is an address defined by a path through a tree.

That path concept is embedded everywhere. Applications reference files by paths. Scripts automate processes using paths. Entire software ecosystems assume that a file can be uniquely addressed by walking a directory tree. So the folder tree is not just a user interface decision. It is part of the plumbing of modern computing. Even so, when you step back and think about the human side of the problem, the tree still feels increasingly unnecessary.

Search has already solved most personal discovery problems

For personal data on a computer, I’m not convinced the structure really matters anymore. Search tools like Windows Search or macOS Spotlight build indexes of files and their contents specifically so retrieval becomes fast and flexible.

Modern systems already allow files to be found by type, date, keywords, content, or metadata. Apple even exposes “Smart Folders”, which are essentially saved search queries that dynamically assemble files matching certain criteria.

From a purely human productivity perspective, this starts to look a lot like the shift that happened with email. Instead of manually curating where everything lives, the system can retrieve what you need when you describe it. The problem appears the moment you leave the world of a single user machine. In enterprise environments, folders are not really about finding things. They are about access.

Enterprise folders are really about access control

Most enterprise systems still secure files through containers. Permissions are applied to folders, and those permissions typically inherit down to the files beneath them. That inheritance model is not accidental. It makes access control manageable at scale.

The same pattern appears everywhere. Windows file systems attach security descriptors to files and folders, including ownership and permission lists. SharePoint structures permissions around scopes that frequently align with folder boundaries. Unix-like systems apply default ACLs on directories that cascade to new files created within them. The hierarchy is doing governance work. Once multiple people are involved, the question stops being “where do I find this document?” and becomes “who should be allowed to see it?”

The hierarchy is doing governance work

Search engines in enterprise platforms solve part of this by security trimming results, meaning users only see documents they have permission to access. But those permissions still have to be defined somewhere, and folders remain the most practical container for that boundary. This is the point where the folder model starts to make more sense again. Not as a navigation system for humans, but as a security and governance boundary for systems.

If that framing is right, then the future state of enterprise file structures may look quite different from what many organisations still build today.

The future may be shallow structures and rich discovery

Instead of deep trees designed to mimic how humans think about projects or departments, the hierarchy could become relatively shallow and primarily reflect access boundaries. Folders would exist to define ownership and permissions, while the responsibility for actually finding things shifts to indexing, metadata, and search.

In other words, folders for governance, search for discovery. Interestingly, you can already see this direction emerging in modern platforms. Tools like SharePoint increasingly support metadata navigation alongside traditional folders, allowing users to browse documents by attributes rather than location. The storage layer might remain hierarchical, but the user experience becomes multi-dimensional. From where I sit, this also aligns with how we think about access automation inside enterprises.

Metadata is becoming more important than location

At Activate we work with organisations that are trying to manage access to large file shares and collaboration platforms without burying everything in ticket queues. One of the modules we’ve built focuses specifically on folder access, allowing permissions to be requested, approved, and automated based on who someone is and what role they perform.

Access automation reveals where the real complexity lives

What becomes obvious when you spend time in that space is that most of the complexity lives in the permissions model, not the documents themselves. The folder is simply the boundary that holds those permissions together.

So perhaps the reason we still have folder trees in 2026 is not because they are the best way for humans to organise information. It is because they remain the simplest structure for computers to enforce ownership, permissions, and governance at scale. Search has largely solved the problem of finding things.

Access, it turns out, is the harder problem.