The worldbuilding-bible-versus-writing-as-you-go debate has a different answer for web fiction than it does for a trilogy you draft in private over five years. A worldbuilding bible is a reference document that pins down your world's rules, names, and history before the story needs them, and writing as you go means inventing each of those details in the chapter that first calls for it. For a serial you publish as you write, the honest answer is neither pure extreme but a living bible: discover the world on the page, and record every rule and name the instant you commit to it, so that three hundred chapters later the magic still costs what it cost in chapter nine. A full bible built up front is mostly wasted on a story you are finding out in public, and pure write-as-you-go collapses under its own continuity debt the moment the archive outgrows your memory.
Do you actually need a worldbuilding bible before you start a web serial?
You do not need a finished worldbuilding bible before chapter one, but you will need a running record long before you think you do, and the authors who skip it pay for it around the time their archive passes fifty chapters. This is the part that gets lost in the usual planner-versus-pantser argument, which was mostly built for novelists who draft a whole book in private and can fix contradictions in revision before any reader ever sees them. Web fiction does not work that way. You publish chapter nine while chapter forty is still a vague shape in your head, and chapter nine is now permanent, read and remembered and quoted back at you in the comments. That one fact changes the whole calculation, because the cost of a worldbuilding mistake is no longer a tidy edit; it is a visible contradiction in a text thousands of people have already read. So the real question is not whether to plan the whole world first, which almost nobody doing a long serial can honestly manage, but how to keep from contradicting yourself across a document that will eventually be longer than most epic trilogies combined.
What a worldbuilding bible actually buys you
A worldbuilding bible buys you consistency and depth at the cost of time spent writing things readers may never see, which is a trade that pays off beautifully for a planned, finite work and much less cleanly for an open-ended serial. The patron saint of this approach is J.R.R. Tolkien, who spent decades building Middle-earth's languages, genealogies, and histories before and around The Lord of the Rings, to the point that The Silmarillion was assembled from his notes and published posthumously in 1977. Brandon Sanderson works in the same architect tradition, drafting magic systems with explicit rules before he drafts the plot that will test them, which is why Mistborn (2006) feels like a system you could almost play rather than a set of convenient powers. Frank Herbert's Dune (1965) carries the same weight of prior invention in its ecology and politics. My feeling about these writers is that the bible is not overhead for them; it is the actual creative work, and the novel is the part where they show a sliver of it. If you are the kind of writer who cannot draft a scene until you know how the money and the magic and the succession all fit together, a bible is not procrastination, it is your real first draft, and you should build as much of it as the story genuinely needs. The trap, which I will come back to, is building the parts the story will never ask about.
Why writing as you go fits web fiction's rhythm
Writing as you go fits the serial form because the serial rewards momentum over architecture, and a published chapter a week beats a perfect world bible that never ships. George R.R. Martin gave us the cleanest language for this split when he described writers as either architects, who design the entire structure before laying a brick, or gardeners, who plant a seed and tend what grows. Martin calls himself a gardener, and plenty of the most successful web fiction was grown rather than drafted. Wildbow's Worm, serialized from 2011 to 2013 at a punishing multiple-updates-a-week pace, has the unmistakable texture of a world being discovered in real time, its power system deepening as the author found new corners of it. pirateaba's The Wandering Inn, one of the longest web serials in the genre, reads like a world that grew outward from a single inn rather than one planned from a map down. I think the honest reason discovery writing dominates web fiction is not aesthetic but structural, and it is worth being clear-eyed about it: the schedule does not leave room to build Tolkien's binders first, and a reader who wanted a finished, fully-planned saga would simply wait for the finished book. The people reading you chapter by chapter are there precisely because the world is still open. If you want the fuller case for discovery over planning, I have made it separately in why discovery writing usually wins for web fiction; this piece is about what that choice costs you once the archive gets long.
The hidden cost nobody warns you about
The cost of pure write-as-you-go is continuity debt, and it compounds exactly like financial debt, invisibly at first and then all at once somewhere past the hundredth chapter. Here is the thing about gardening a world in public: a gardener still has to remember what they planted, and human memory is a terrible database for a work that will run to a million words. Even the architects are not immune. A Song of Ice and Fire is one of the most carefully tracked worlds in modern fantasy, and it still carries small, well-documented continuity slips across its enormous length, the kind of character detail or timeline wrinkle that readers catalogue precisely because the work is otherwise so rigorous. That is the point worth internalizing, and it is the reason I lean the way I do: continuity breaks in long serials are almost never failures of imagination. They are failures of retrieval. You did not run out of ideas; you forgot that in chapter twelve you said the artifact could only be used once, and now in chapter ninety the plot wants it twice. Running a platform where authors publish very long serials, I see this pattern constantly, and it is never the writer who planned too little who gets into trouble first. It is the writer who planned a reasonable amount and then never wrote any of it down, so that every established fact lives only in a head that is also trying to hold next week's cliffhanger. The world got too big to remember, and nothing was catching the overflow.
The approach I actually recommend: a living bible
The approach that fits web fiction is a living bible, which means you write the story as you go and record every concrete decision the instant you commit to it on the page, so the reference grows at exactly the pace of the fiction and never gets ahead of it. This is the synthesis that dissolves the whole debate. You are a gardener by method, discovering the world while you draft, because the schedule demands it and the form rewards it. But you are an archivist by discipline, because the moment you name a city, fix what a spell costs, establish who knows the secret, or kill someone, you write that fact into a document you can search later. You are not planning the world in advance; you are refusing to lose the world you have already built. The distinction between what you have shown readers and what you have merely imagined is the whole game here, because you are free to invent and revise anything the story has not yet committed to, and bound only by the things a chapter has already stated as fact. A running record is what tells you which is which at chapter two hundred, when you genuinely cannot remember.
This is also, I will admit plainly, the problem IlorisNovel was built to solve, because keeping a living bible by hand in a separate document is itself a discipline that fails the week you are behind on your update. The editor's connected wiki is meant to be that living bible without the separate-document tax: you note a world fact once, you can @mention it inside a chapter as you write, and the timeline and entry pages hold the names, rules, and relationships so that the record and the draft move together instead of drifting apart. The specific discipline of not losing your own continuity is close enough to its own craft that I have written about it directly in keeping continuity in a long web serial, and it sits alongside the broader question of how systems, maps, and lore scale, which is the subject of the worldbuilding for web fiction guide. Worth separating, too, from a question people often fold into this one: whether to worldbuild before or after your characters. That is about sequence, and I have argued elsewhere that worldbuilding should usually come last. What we are talking about here is not when you build the world but whether you keep a record of it, and my answer to that is the same for everyone regardless of how they plan: keep the record, update it as you write, make it searchable, and let it grow one committed fact at a time until, somewhere around chapter two hundred, you look up and realize you have written a worldbuilding bible after all, the slow way, the only way a serial ever really can.
Common questions about worldbuilding bibles and writing as you go
Do I need a worldbuilding bible to write web fiction?
Not up front, but you will need one by chapter fifty whether you planned it or not. For a serial you are publishing as you write, a full bible built before chapter one is mostly wasted, because the story will contradict your plans within a month. What you cannot skip is a running record: the moment you name a city, fix a magic cost, or kill a character, write it down somewhere searchable. That growing document is your bible, and web fiction's length makes it non-negotiable.
What is the difference between a worldbuilding bible and writing as you go?
A worldbuilding bible is a reference document that fixes your world's rules, names, maps, and history before the story needs them. Writing as you go means inventing each of those details in the chapter that first requires it, with no prior plan. Most working authors land between the two: they discover the world while drafting, which is writing as you go, but they record each decision as they make it, which slowly builds a bible. The split is less planner versus pantser than whether you keep notes.
How do successful authors handle worldbuilding, as architects or gardeners?
George R.R. Martin's own metaphor splits writers into architects, who design the whole structure before laying a brick, and gardeners, who plant a seed and see what grows. Tolkien was an architect who built Middle-earth's languages and history for decades before publishing. Martin calls himself a gardener. Both produced enormous, consistent worlds, so neither method is wrong. What matters for a serial is that a gardener still has to remember what they planted, which is where a running record earns its keep.
How do I keep worldbuilding consistent across a long web serial?
Record every concrete decision the moment you publish it and keep that record searchable, rather than trusting memory across hundreds of chapters. Continuity breaks in serials are rarely failures of imagination; they are failures of retrieval, where you simply forgot what you already established forty chapters back. A living reference that captures names, magic costs, timelines, and who knows what, updated as you write and searchable when you draft, prevents far more contradictions than any amount of upfront planning.
Is it bad to change my worldbuilding after I have published chapters?
Changing unestablished details is fine; contradicting established ones is the problem. Readers never see the parts of your world you have not shown yet, so you are free to invent and revise anything the story has not committed to on the page. What breaks trust is reversing something a chapter already stated as fact, like a magic rule or a death. The fix is knowing which is which, and that is exactly what a running record gives you that memory alone cannot.
by Jacob Tam · October 28, 2026
I run IlorisNovel. If you want to see what writing here feels like, you can try the editor with no account.