A folder decides where a note lives. A tag describes what it's about. Search ignores both and just finds the words. They feel like rival religions, but they're really three different bets on the same question: how will I get this note back six months from now? Pick the bet that matches how your memory actually works, not the one with the tidiest screenshot.
The three approaches, honestly
| Folders | Tags | Search | |
|---|---|---|---|
| The idea | One home per note | Many labels per note | No structure — find by content |
| Filing effort | High — decide where it goes | Medium — decide what to call it | Zero |
| Handles overlap | Badly — a note lives in one place | Well — a note can be many things | N/A — content is the index |
| Breaks down when | A note fits two folders | Tag names drift ("#idea" vs "#ideas") | You forget the exact words |
| Best for | A few obvious buckets | Cross-cutting themes you reuse | Almost everything else |
Folders: good for a few obvious buckets
Folders are intuitive because the physical world works the same way. The catch is the one home rule: the moment a note could plausibly live in two folders — is this meeting note under "Client X" or "Q3 Planning"? — you've bought yourself a small decision every single time you file, and a coin-flip every time you look. That friction is what quietly kills capture, because forgetting starts within minutes and a thought you didn't save is gone.
Use folders anyway, but only a handful of obvious ones. The disciplined version of "few broad buckets" is the PARA method, which files by how actionable a note is rather than by topic — a smart move, because topics overlap and deadlines don't.
Tags: good for cross-cutting themes
Tags fix the one-home problem: a note can be #recipe and #vegetarian and #quick at once. That's genuinely powerful for themes that cut across your folders. The failure mode is subtler and slower — tag vocabularies rot. You write #idea on Monday and #ideas on Thursday, and now neither view is complete. Tags reward discipline most people don't sustain, so keep the set small and boring. If you want tags to do real thinking rather than just labelling, the connection-first tradition — Zettelkasten and atomic notes — is where they earn their keep, but that's a bigger commitment than most notes need.
Search: the one that keeps working
Here's the uncomfortable truth folders and tags dance around: you rarely remember where you filed something, but you almost always remember something it said. Instant full-text search turns that scrap into a hit in milliseconds, and it costs nothing to maintain because your words are the index. It fails only when you've forgotten the exact wording — and that gap is exactly what AI search now closes, answering a plain-language question ("what did I decide about the lease?") across everything you wrote, and surfacing connections you'd never line up by hand.
So which should you use?
All three, in this order of trust: search first, a few folders for the truly obvious, tags only for themes you'll genuinely reuse. The mistake is inverting that — building an elaborate folder-and-tag cathedral before you've written anything, then abandoning it. Capture should be effortless and organising should be optional; for the full playbook, see how to organise your notes so you actually use them.
Where does an app choose for you? Notion leans on databases and properties (structured, powerful, heavier). Obsidian gives you folders, tags, and backlinks for local-first control. Apple Notes does folders plus solid search as the Apple default. And Clair Mind makes the search-first bet on purpose: one-tap capture, light folders, and AI that connects and answers across your own notes offline — so you only have to remember that you wrote something, not where. That's the right fit if fast, private, phone-first capture is your real problem; if you need a shared team workspace or a hand-built knowledge graph, the others above fit better.
Deciding from your situation instead? Start at best notes apps by use case.