Most note apps make you choose early: a free-form outliner like Roam or Logseq, or a structured database like Notion. Tana's pitch is that you shouldn't have to. You write in a plain bullet outline, and when a bullet needs structure — a task, a person, a meeting, a book — you tag it, and that single tag turns the bullet into a typed object with fields, a template, and a place in queryable views.
We spent two weeks moving a working week's worth of notes, tasks, and reading list into Tana to see whether the model holds up past the demo. It does, mostly. But the gap between "this is clever" and "this is my daily driver" is wider than the marketing suggests.
How the supertag model actually works
Everything in Tana is a node — a bullet. That part is familiar to anyone who has used an outliner. The difference is the supertag. Type #task on a bullet and that bullet inherits whatever you defined #task to be: a status field, a due date, an assignee, maybe a #project reference. The text stays where you wrote it, in the flow of your notes, but the metadata becomes structured.
The payoff comes from live search. You define a query once — "all #task nodes with status not done, due this week" — and Tana renders it as a list, a table, or a kanban board that updates as you tag new bullets anywhere in your workspace. You're not maintaining a separate task database. The database is your notes, read through a filter.
This is the genuinely different idea. In Notion you create a page inside a database. In Tana you write a thought, and tagging it after the fact promotes it into the structured layer. Capture stays frictionless; structure is retroactive.
The daily note (Tana calls it the calendar node) anchors the workflow. You dump everything into today's bullet list, tag what matters, and the tagged items surface in the relevant views automatically. A meeting note tagged #meeting with a #person field shows up later under that person's node, alongside every other time you mentioned them.
Supertags can extend other supertags, the way a class extends a parent. A
#1-on-1tag can inherit every field from#meetingand add its own. This is powerful, but it means your tag schema is effectively a small data model — and like any data model, it rewards planning and punishes sprawl.
Where Tana earns its place
The strongest case for Tana is the knowledge worker who already thinks in outlines and is tired of copy-pasting the same note into a separate tracker. A few things worked noticeably better than our previous Notion-plus-text-file setup:
- Capture-then-structure. Writing first and tagging second matched how notes actually accumulate during a call. We weren't deciding which database a thought belonged to before we'd finished the thought.
- References that compound. Mentioning a project by its node anywhere creates a backlink. Over two weeks the project nodes became real dashboards without any deliberate dashboard-building.
- Tana AI for capture. The voice-to-structured-node feature turned a 40-second spoken brain-dump into tagged tasks with due dates pulled out of the audio. It misread one date out of roughly a dozen we tried — usable, not flawless. AI commands can also fill fields or summarize a node's children on demand.
For comparison, here is how the core model differs from the two tools most people would otherwise reach for:
The table flattens nuance, but the pattern is clear: Tana sits between the free-form outliner and the rigid database, and the supertag is the hinge.
Where it bites back
Tana is not a tool you adopt over a coffee break. The friction is real and worth naming before you migrate anything important.
The learning curve is steep. The first few days felt slower than the setup we left, not faster, because the supertag schema is something you build, not something you're given. Get it wrong early and you accumulate inconsistent tags that your queries silently miss. We rebuilt our tag structure once after realizing #task and #todo had drifted apart.
It is cloud-only. Unlike Logseq's local markdown files, your workspace lives on Tana's servers. There's no offline-first local copy, which matters if you care about data portability or working on a plane. Export exists, but the structured relationships don't round-trip cleanly into a plain format.
Pricing is a consideration, not an afterthought. Tana offers a free tier and paid plans, with the AI features and higher usage gated behind the paid levels. Confirm the current tiers on Tana's pricing page before you commit a team — the AI quotas in particular are the kind of thing that changes, and the free tier may not cover heavy daily use.
Don't migrate your entire system in week one. Run Tana alongside your current setup for one project or one area of your life first. The supertag schema you design after two weeks of real use will be meaningfully better than the one you'd design on day one — and rebuilding a half-migrated workspace is the most common way people bounce off this tool.
If the structured-but-flexible model appeals but the outliner mental shift doesn't, Notion remains the lower-friction default. It gives you databases and relations without asking you to think in bullets, and the onboarding cost is far lower for a team that just needs shared docs and trackers.
Two weeks isn't long enough to call Tana a forever-tool, but it's long enough to say the idea is sound and the execution is real. If you already live in outlines and your notes and tasks are duplicated across two apps, the payoff is genuine. If you don't, budget for the climb — or pick the tool that meets you where you already are.
Originally published at pickuma.com. Subscribe to the RSS or follow @pickuma.bsky.social for new reviews.
Top comments (0)