Plotiar vs Notion for Novel Writing: When You Should Use Which
Three months I spent building "the perfect Notion novel template." It still exists, somewhere in my workspace, behind a wiki page I titled THE NOVEL OS in block capitals because I thought the shouting would make it feel more official. It has a database for characters with fourteen properties each. It has a relation field linking every scene to every character who appears in it, so I could filter by "which scenes does Marcus appear in" and get an instant answer. It has a rollup that counts total word count per chapter, aggregated automatically, updating live. It is, by any reasonable measure, an impressive piece of information architecture.
I never finished a book in it.
Not "never finished the book I was building it for." Never finished a book, full stop, in the eighteen months I spent as a committed Notion user before I admitted the problem was not my discipline. The problem was that I had built the world's most elegant filing cabinet and mistaken it for a desk.
This is not a hit piece. Notion is one of the best pieces of software made in the last decade, and I want to say that plainly before I say anything else, because most "Notion alternative" posts pretend the tool is bad and it is not. The seduction is real, and it is earned. But there is a specific, structural reason Notion is brilliant at organizing a novel and bad at helping you write one, and once you see the shape of that reason, you stop blaming yourself for the eighteen months, and you start building a setup that actually gets a draft finished.
Why Every Writer Eventually Opens Notion
The pitch is genuinely irresistible if you have ever tried to plan a novel in a stack of loose documents. Notion gives you databases that talk to each other. A character can be linked to every scene, every location, every plot thread that touches them, and the links update automatically in both directions. You can build a kanban board for your chapters, a calendar for your writing streak, a gallery view of your worldbuilding images, and a table of every named character sorted by "first appearance" -- all inside one workspace, all free-form enough that you invent the structure yourself instead of being handed someone else's template.
Tiago Forte's PARA method -- Projects, Areas, Resources, Archives, laid out in his book Building a Second Brain -- was written for knowledge workers organizing client files and meeting notes, not novelists, and yet it maps onto a novel with almost eerie precision. The manuscript becomes a Project. The characters, world, and research become Resources. Old drafts become the Archive. Search "Notion novel template" and you will find dozens of "novel OS" templates built explicitly on this logic, most of them free, most of them beautiful, most of them built by people who, I would bet, have not actually finished the novel the template was designed for. I do not say that as an insult. I say it because I was one of them, and because the template-building itself is where the appeal quietly turns into the trap.
The trap is this: building the database feels like writing. It has the same shape. You open the app. You make decisions. You produce something you can look at and feel proud of at the end of a session. But a character database with fourteen properties is not a scene. It is scaffolding for a scene, and scaffolding is endlessly, pleasurably expandable in a way that an actual chapter of prose is not. There is always one more property to add. There is never a natural stopping point that says "the database is finished, go write." Prose has that stopping point. A chapter ends. A database does not, unless you make it stop.
What Notion Is Genuinely Brilliant At
I want to be specific about where Notion earns its reputation, because the honest version of this comparison depends on it.
As a research database, Notion has no real peer among writing-specific tools. If you are writing historical fiction and accumulating three hundred source clippings, quotes, and half-formed facts about seventeenth-century Amsterdam shipping routes, Notion's tagging, filtering, and cross-referencing will organize that mess better than almost anything built for novelists specifically. As a character database for a series with forty recurring names across six books, the relational structure is genuinely useful -- you can ask "every book where this character appears" and get a real answer instead of scrolling six manuscripts. As a worldbuilding wiki, where every location links to every faction which links to every character who belongs to it, Notion does something close to what a real encyclopedia does, and it does it without you writing a line of code.
There is also a version of this that has nothing to do with fiction at all, which is worth naming because it is the deeper principle underneath the appeal. Niklas Luhmann, the German sociologist, built a physical index-card system he called the Zettelkasten -- roughly ninety thousand cards, each holding one idea, cross-referenced by hand-numbered links to related cards -- and used it to write more than seventy books and hundreds of papers over his career. Notion is, in a real sense, the Zettelkasten's digital descendant: atomized notes, densely cross-linked, designed to let structure emerge from the connections rather than being imposed from the top down. That is a genuinely powerful way to think. It is a fantastic way to build a body of interconnected knowledge. It is just not, on its own, a way to write nine hundred consecutive words of a scene where two people who love each other say the wrong thing to each other in a kitchen.
Where the Disguise Wears Off Around Chapter Eight
Here is where Notion stops being a database and starts pretending to be a word processor, and where the seams show.
The block-based editor that makes Notion so flexible for notes is the same architecture that makes it a genuinely uncomfortable place to write sixty thousand consecutive words of prose. Every paragraph, every line, is a discrete block with its own metadata, its own render logic, its own place in a tree the app has to keep track of. That is fine at the scale of a meeting note or a project brief. It becomes a real performance problem at the scale of a chapter, and worse at the scale of a manuscript, because a long page in Notion is not one document to the renderer. It is thousands of individually tracked objects, and the app has to recalculate all of it every time you type a character. Anyone who has tried to draft a full chapter inside a single Notion page has felt the specific, sinking lag that shows up somewhere past the two-thousand-word mark -- a half-second delay between keystroke and character appearing on screen that does not sound like much until you are trying to hold a sentence's rhythm in your head and the sentence cannot keep up with your hands.
The formatting options that make Notion great for structured notes -- toggle lists, callout blocks, embedded databases, colored text -- actively work against prose. There is no page. There is no sense of where you are inside a manuscript, no pagination, no visual rhythm that tells your eye "this is a chapter, and it has a shape." You are typing into an infinitely scrolling column that looks identical whether you have written one paragraph or eleven thousand words, and that visual sameness is a subtle but real cost. Anne Lamott's "one-inch picture frame" -- the idea that you write better when the unit of work in front of you feels holdable -- depends on the unit actually looking different from the whole. Notion gives you the same undifferentiated scroll no matter how much you have written, which means the discipline of "just write this page" has nothing visual to hold onto.
And there is no offline-first confidence. Notion syncs to the cloud constantly and mostly well, but the editor's responsiveness is tied to that sync layer in a way a dedicated writing tool's is not, and writers who have watched a page freeze mid-sentence while a sync spinner turns in the corner do not forget it. None of this is a knock on Notion as a company. It is a description of what happens when you ask a note-taking tool, built to be infinitely reconfigurable, to also be a manuscript editor, built to disappear underneath the sentence you are writing. Those are different jobs. Notion is honest about which one it was built for. The honest version of the novel-planning story admits that most of us pretended otherwise.
What a Writing-Specific Editor Actually Buys You
Plotiar exists because of exactly this gap, and I want to describe what it does differently rather than just asserting that it is better, because "better" is not a useful word without the specifics underneath it.
The canvas editor renders real pages -- an actual A4 page with pagination, the same visual object a printed manuscript would be, so the document has a shape that changes as you write instead of an infinite undifferentiated scroll. That sounds cosmetic until you have felt the difference. A page that fills up and turns is a unit you can hold in your head the way Lamott's picture frame describes; an infinite column is not. Comments live in the margin next to the line they are about, the way a human editor's notes would, instead of as a separate database entry you have to cross-reference back to a specific paragraph by hand. And because the editor is built for long-form prose specifically -- not repurposed from a note-taking block architecture -- a ninety-thousand-word manuscript stays responsive the way a two-thousand-word one does, because the underlying document model was designed for that scale from the beginning rather than adapted to tolerate it.
The structural planning tools sit in the same project as the manuscript, which is the piece Notion cannot offer no matter how many databases you build, because Notion has no purpose-built flowchart for tracing a plot's cause-and-effect chain, no plotgrid for watching how four subplots intersect chapter by chapter, no family tree renderer, no map layer. You can simulate crude versions of some of these with tables and linked databases -- I built a passable one for character relationships once, using a self-referencing relation property, and it took me four hours and looked like a spreadsheet wearing a costume. Purpose-built visual tools do the same job in minutes and look like what they are. The three-folder architecture -- manuscript, workshop, scraps -- described in how working novelists actually organize a project is a close cousin of what a well-built Notion workspace tries to achieve. The difference is that in Plotiar, the manuscript half of that architecture is a real writing surface, not a database entry pretending to be one.
The Hybrid Setup, When It Genuinely Makes Sense
I do not think the honest answer here is "abandon Notion entirely," and I would be lying if I pretended it were, because there is a real category of writer for whom Notion keeps earning its place even after they move the actual manuscript somewhere else.
If you are running a research-heavy project -- a historical novel with hundreds of primary-source clippings, a series bible with genealogies stretching across six books, a worldbuilding wiki with the kind of cross-referenced density that made the Zettelkasten useful to Luhmann in the first place -- Notion remains a genuinely strong home for that material. The move that actually works, rather than the one I tried for eighteen months, is a division of labor: the manuscript lives in a writing-specific tool built to disappear underneath the prose, and the database of everything the manuscript needs to stay consistent with lives somewhere built to hold a database well. You research and reference in one place. You draft in another. The two do not need to be the same app, and pretending they should be is, I think, the single most common and most expensive mistake in this whole conversation.
| Job | Where it belongs | Why |
|---|---|---|
| Drafting the manuscript | A writing-specific canvas editor | Built for long-form prose, pagination, responsiveness at ninety thousand words |
| Research clippings and sources | A flexible database tool | Tagging, filtering, and cross-referencing hundreds of loose facts |
| Character and plot planning | Purpose-built visual tools | Flowcharts, plotgrids, and family trees show structure that tables cannot |
| Series-spanning continuity | Whichever holds relational data best | Depends on the series; either tool can serve if the manuscript is elsewhere |
Patrick Rothfuss is the case I keep coming back to when I think about the cost of confusing infrastructure with progress. The Name of the Wind came out in 2007. The sequel to the sequel, The Doors of Stone, has not appeared as of this writing, and Rothfuss has spoken publicly, over the years, about the pressure of a fanbase watching him not finish it. I am not claiming to know what happens in another writer's process, and I want to be careful not to diagnose a stranger's book from the outside. But the general shape of the trap he has described in interviews -- perfectionism, an expanding sense of everything that has to be reconciled before the next word can be trusted -- is the same shape I felt building fourteen properties per character in a database that would never be read by anyone but me. Infrastructure has no natural ceiling. A chapter does. That asymmetry is the whole danger.
It is worth noticing the writers who went the other direction entirely. George R.R. Martin has said in interviews, including a much-quoted 2014 appearance with Conan O'Brien, that he still drafts A Song of Ice and Fire on a DOS machine running WordStar 4.0, precisely because it cannot autocorrect, cannot suggest, cannot do anything except put the letters he types onto the screen. I am not arguing every writer should regress to 1985. I am arguing that the correlation between "how much the tool can do" and "how much writing gets done" is not positive, and past a certain point it inverts. A workspace with fourteen character properties per entry and a workspace with no properties at all are both, in their own way, trying to solve the same problem: getting out of the writer's way long enough for a sentence to happen. Notion solves it by offering everything and trusting you to use only what you need. Most of us, myself included, are not disciplined enough to make that trust pay off.
When to Commit to One Tool Over the Other
If you are early in a project and the volume of research or cross-referencing genuinely exceeds what a single writer can hold in their head -- a series bible with real scale, a historical novel with archival-level sourcing -- keep Notion, or something like it, for that layer, and be disciplined about not letting the manuscript live there too. If your project is a single novel of ordinary complexity, and you notice yourself spending more sessions refining the database than producing pages, that is the signal to stop. Not to feel bad about the eighteen months, the way I have made a kind of peace with mine. Just to stop, and move the actual writing into a tool that was built to get out of your way once the sentence starts.
The test I use now, months after abandoning THE NOVEL OS, is simple. At the end of a working session, did I produce prose, or did I produce a place where prose could theoretically go? If it is the second thing three sessions in a row, the tool is not the problem yet. The redirection of effort is. But if the tool itself is making the second thing easier to produce than the first -- if the block editor rewards you with visible progress for building a database and punishes you with lag for writing a chapter -- then eventually the tool becomes the problem too, the way mine did, for a year and a half, before I noticed.
I still open Notion sometimes. I have a research database for my current project that I am, unlike THE NOVEL OS, actually proud of, because it holds exactly what it should hold and nothing else: sources, a timeline of real historical events I need to stay consistent with, a list of names I have already used so I do not repeat one by accident in chapter thirty. The manuscript itself lives somewhere else now, in a tool I have also compared honestly against Scrivener, with a canvas that turns pages instead of scrolling forever. THE NOVEL OS is still out there, unfinished, fourteen properties deep, a monument to a very specific and very common way of not writing a book. I do not delete it. I just do not open it anymore. Some structures are worth keeping around exactly because you built them, and precisely because you learned, eventually, not to live inside them.
Plotiar plays nicely with Notion. Keep your research database wherever it already works; write the manuscript in a canvas built to disappear underneath the prose. Start a project for free and see what changes when the two jobs finally live in different rooms.