How to Build a Story Franchise: The Canon Architecture Behind a World That Holds More Than One Story

A franchise is a world that can pay for more than one story

The word gets used to mean a successful thing with sequels, which is the outcome rather than the structure. Structurally, a franchise is a world with more story in it than any single work extracts — and, critically, one where extracting more does not require the original author to be present.

That second clause is the whole discipline. A world that only its creator can safely write in is not a franchise; it is a personal project with a queue. It can still be enormously successful, and plenty of great work is exactly that. But it will never scale past one person's working hours, and every proposal to expand it will arrive as a risk rather than an opportunity.

The distinction shows up immediately in how the two are planned. A single story is planned toward an ending. A franchise is planned toward a middle — a stable, well-documented state of the world from which many stories can depart. Ask what is true in your world when no plot is running, and a franchisable world has a long answer.

This is also why franchise development starts earlier than most creators expect. The decisions that make a world extensible are cheap to make before the first work is finished and expensive afterwards, because by then the first work has silently answered them all by implication. The craft of making multiple properties cohere is its own subject, covered in how to build a fictional universe; this pillar is about the decisions underneath it.

Not every good world can carry one

It is worth being honest about this early, because the alternative is spending two years on infrastructure for a world that did not need it. Three tests, and a world that fails the first two is better served by being written well than by being franchised.

**Does the world generate stories you have not told?** Not sequels to the story you told — different stories, about different people, that could only happen there. If every genuinely interesting thing about the world was consumed by the first work, the world was a setting rather than a franchise. This is a common and entirely respectable outcome. The test is whether you can name three stories in the world that do not involve your protagonist, and want to read all three.

**Can a stranger hold the premise?** A franchise is built by people who did not invent it. If the world takes forty minutes to explain before anyone can write in it, that cost is paid again by every new person, forever. The worlds that extend well tend to have a compressible core — a rule, a situation, a question — that survives being stated in two sentences, with the depth underneath it rather than in front of it.

**Is there a line between fixed and open?** Not whether you have written it down yet, but whether it exists at all. Some worlds are load-bearing everywhere — every detail is essential and nothing can be added without disturbing something. Those are often the most beautiful worlds and the hardest to franchise, because there is no room in them for anyone else.

A world that passes all three is worth the structural work. A world that passes the first but fails the second is usually fixable — the problem is presentation, not architecture. A world that fails the third is a much bigger decision, because making room in it means deciding that some things you invented are no longer sacred.

The canon architecture: fixed, open, and the line between

This is the central act of franchise development and the one most often skipped. Every fact in your world belongs in one of two categories, and the franchise's health depends on which is which being explicit rather than felt.

**Fixed canon** is load-bearing: the facts other facts depend on. The rules the world runs on, the events that produced the present situation, the things about the central characters that cannot change without a different story. Contradicting fixed canon does not create a small error; it invalidates work that was built on top of it. Fixed canon should be small — much smaller than instinct suggests — because everything in it is a constraint on everyone who comes later.

**Open canon** is everything a new creator may invent freely: regions nobody has visited, decades nobody has dramatized, institutions that exist but have never been shown from the inside. Open canon is not the unimportant part of the world. It is the part with room in it, and a franchise with too little of it suffocates.

The line between them has to be written down and stated positively in both directions — this is fixed, that is yours. A bible that only lists what is true, without marking what is fixed, fails in one of two ways, and which one you get depends on the temperament of the person reading it. Cautious creators treat everything as sacred and produce timid work that adds nothing. Confident ones treat nothing as sacred and break the spine. Both failures are the bible's fault.

There is a third category worth naming explicitly, because leaving it unnamed causes most of the arguments: **provisional canon**, things that are currently true but that the canon authority is willing to change. Marking something provisional tells a new creator not to build load-bearing work on it, which is exactly the information they need and almost never get.

A useful discipline when drawing the line is to ask, of each fact, what breaks if this changes. If the answer is nothing, it does not belong in fixed canon no matter how much you like it. The related question of keeping the whole record consistent as it grows is canon management, and the underlying principle — one authoritative record rather than several that disagree — is a single source of truth.

The second property is the one that defines the franchise

The first work establishes that the world is good. The second establishes that it is a world at all, rather than scenery behind one plot. Everyone who considers building in it later will read the second property as the statement of what this franchise is — which makes it the highest-leverage decision in the whole enterprise, and one routinely made on availability rather than strategy.

The instinct is to continue: same characters, later events, the next chapter. That is the safest choice commercially and the weakest one structurally, because it proves nothing about the world's capacity. A direct sequel demonstrates that the story continues, not that the world holds. If the second property is a sequel, the third has to do the structural work instead, usually at a point where the franchise's identity is already set.

The stronger move is lateral: a different corner, a different cast, a different scale, in the same world under the same rules. It is riskier and it answers the only question that matters — whether the world is generative or decorative. It also produces the most useful artifact in franchise development, which is a second creative team's honest experience of trying to work from your canon. Everything ambiguous in your bible surfaces in that first handoff.

Choosing the medium is a separate question from choosing the story, and conflating them is a common error. A world does not extend to a new medium simply because it was popular in the last one; each medium does something specific well, and the choice should follow from what the next story needs rather than from what is currently financeable. That decision has its own guide in transmedia storytelling.

Somebody has to be allowed to say no

Every franchise that stays coherent has a canon authority, whether or not it uses the phrase. Every franchise that loses coherence lost it by leaving the question unanswered, at which point authority defaults — usually to whoever produced the most recent work, which is how a universe ends up with three incompatible accounts of the same event.

The arrangement that works is narrow authority and wide proposal rights. A small number of named people can approve a change to canon. Everyone else can propose one, through a route that actually exists and gets answered. The proposal route matters as much as the authority: a franchise where the only way to add canon is to be one of the two people who can approve it will get exactly as much new canon as those two people have time for.

Write down what requires approval. In practice the list is short and predictable: anything touching fixed canon, anything that changes a character's established state, anything that dates events relative to existing ones, and anything that adds a rule to the world. Everything else — new locations in open regions, new minor characters, new detail that contradicts nothing — should not need a meeting, and requiring one is how you teach collaborators to stop asking.

Record decisions, not just outcomes. The most expensive kind of franchise argument is the one that was already settled two years ago by someone who has since left, with no note explaining why. A one-line reason attached to a canon decision is worth more than the decision itself, because it lets a future team apply the same logic to a case nobody anticipated.

For a room working on a single property at a time, this is the same discipline as writers' room canon management; at franchise scale it simply persists between productions rather than resetting with each one.

The record other people can actually work from

A franchise bible is not a longer story bible. It is a differently-shaped one, built for a reader who did not invent the world, is under time pressure, and cannot ask you anything.

It has to be navigable before it is complete. A buyer deciding in an afternoon and a staff writer checking a fact mid-draft have the same requirement: find the answer in seconds. Canon buried in beautiful prose that nobody can search is functionally unavailable no matter how thorough it is — and thoroughness that cannot be retrieved actively hurts, because it creates the belief that the answer exists somewhere while making it impossible to find.

It has to mark status, not just state facts. Fixed, open, provisional — for every entry. This is the part that makes the difference between a document that enables other creators and one that merely informs them.

It has to be demonstrably consistent. Asserting that a world holds together is not the same as showing it, and at franchise scale nobody can verify it by reading. The number of pairs of facts that must agree grows roughly with the square of the number of facts, which is why worlds that were fine at one property start producing contradictions at three — not because the team got worse, but because the checking problem outgrew human memory somewhere between the two.

The full specification of what an IP-grade bible has to contain is in the story bible as an IP asset, and the structure for a team working from one canon is in the shared story bible.

Continuity stops being a memory problem

At one property, continuity is a discipline: read carefully, keep notes, catch your own mistakes. At franchise scale it becomes an infrastructure problem, and the methods that worked for a novel fail quietly rather than loudly.

The specific change is that nobody has read everything. A new writer on the fourth property has read the first, skimmed the second, and been briefed on the third. They will contradict something in good faith, and no amount of care prevents it — the information required to avoid the error was never in their head. This is why franchise continuity is a system rather than a virtue.

What the system needs is unglamorous. One authoritative record rather than several. A way to check a new piece of work against it that does not depend on someone happening to remember. A defined moment in the production process where that check actually happens, because a check with no scheduled home does not occur. And a way to record the fact that a contradiction was noticed and deliberately allowed, which is a normal and healthy thing that becomes a disaster when undocumented.

Tooling helps precisely at the point where the pairwise checking exceeds memory. CanonBoard's Conflict Detection reads the cards on your canvas and reports contradictions it finds — it is available on Core and above and costs one credit per scan. It never writes or invents anything; it reads what you have recorded and tells you where two things disagree. Anything you are not ready to have checked can sit in the Locked Vault, also Core and above, which the scan leaves alone entirely.

The operational detail of running this across a universe with multiple concurrent productions is covered in franchise continuity management, and the long-horizon version for a single continuing story is tracking canon across a series.

Letting other people build in it

At some point a franchise stops being extended only by you. That happens through hiring, through licensing, or through an audience — and the structural requirements are the same in all three cases, which is convenient, because it means the work is done once.

What every outside creator needs is identical: a clear statement of what is fixed, explicit permission over what is open, a route to propose a change, and someone who answers. A franchise that has done its canon architecture properly can onboard a licensee in a week. One that has not will spend that week discovering, in real time, which of its assumptions were never written down.

The commercial and legal shape of this — what a licensee may do, what stays yours, how credit and approval work — is its own subject, covered in licensing your story world.

The audience version is newer and more interesting than it first appears. Inviting readers to pitch into a defined corner of a world is a way to find out what people want more of, before committing a production budget to the answer. It works on exactly the same architecture: a public region that is genuinely open, a private remainder that is genuinely sealed, and canonization staying entirely with the creator. That is what The Board is built to do — the guardrails are set out in opening your world without losing it, and what canonizing a contribution does and does not obligate is in credit, not compensation.

What a buyer is actually evaluating

Franchise development eventually meets a buyer — a studio, a publisher, a partner — and the thing being evaluated is not the story. It is whether the world can carry more than the thing you are showing them, and whether working in it will be expensive.

That second question is the one creators underestimate. A buyer is estimating how much work it will take to make more of this, and a world whose canon is unwritten or contradictory is a world that will need an expensive year of development before anyone can write in it. A coherent, navigable, explicitly-scoped canon is not a nice extra in that conversation; it is a substantial part of what is being valued.

This is why the bible does commercial work rather than administrative work. It is the evidence for the claim that the world holds. Showing that you already know what is fixed, what is open, and who decides, answers most of what a buyer needs to know about the risk of building on you.

The full picture of what makes a world into an asset is in how to develop a story IP, and what actually happens in the room is in pitching original IP to studios.

Where to start, if you are starting now

The work compresses into a short sequence, and doing it in order matters more than doing all of it.

Write the fixed list first, and make it short. Everything the world cannot contradict without invalidating existing work. If it runs past a page or two, you are describing the world rather than constraining it, and a fixed list that long will strangle every creator who comes after.

Then mark what is open — positively, by name. Regions, eras, institutions, kinds of story. This is the part that reads as generous and functions as engineering, and skipping it is why so many franchise bibles produce timid work.

Then name the authority and the route. Who approves a change to fixed canon, and how someone who is not them proposes one. Two sentences is enough. An undefined process defaults to the loudest recent voice.

Then make it retrievable. A record nobody can search is a record nobody uses, and at franchise scale the difference between available and retrievable is most of the value. Structured canon — where anyone can find what is true about a character, a rule, or a moment in seconds — is the entire point of the exercise. That is what CanonBoard is for: one authoritative record of a world, structured so a whole team can work from it and check against it. You can start building free, and the Locked Vault keeps the unreleased parts sealed while the rest of the world stays open to the people you have invited.

None of this makes the stories good. It makes them possible to keep making, by more than one person, for longer than one project — which is the only thing that separates a franchise from a world that had one very good year.

Frequently asked questions

What is franchise development?

Franchise development is the work of turning a story world into one that can support multiple properties over time — more than one story, often in more than one medium, frequently made by people who are not you. It is distinct from writing a good story and distinct from selling one. The central task is structural rather than creative: deciding which facts about your world are load-bearing and permanent, which are open for other creators to fill in, and who has the authority to change either. A world can be brilliant and completely unfranchisable if nobody but its author can safely add to it.

What makes a story world franchisable?

Three properties, and popularity is not one of them. First, the world has to generate more stories than the one you told in it — if every interesting thing that could happen there already happened in book one, there is no franchise, there is a novel. Second, it needs a premise that a stranger can grasp fast enough to build on, because a world only you can explain is a world only you can write. Third, and most often missing, it needs an explicit line between what is fixed and what is open. Without that line, every new creator either freezes, afraid to invent anything, or breaks the spine because nothing told them it was load-bearing.

What is the difference between a story bible and a franchise bible?

Scope and audience. A [story bible](/blog/what-is-a-story-bible) is the authoritative record of what is true in a world, usually written for the person or room already working in it. A franchise bible is that record rewritten to survive being handed to people who did not build the world and cannot ask you questions — which means it must be navigable rather than merely complete, must state what is fixed versus open, and must prove the world is internally consistent rather than assert it. Most private bibles fail that test badly, because they were never meant to pass it. The full requirements are in [the story bible as an IP asset](/blog/story-bible-for-ip).

How many properties does it take to be a franchise?

Two, and the second one is the hard one. The first story establishes the world; the second proves the world was real rather than a backdrop for one plot. That is why the choice of second property matters more than any decision after it — it sets the precedent for what kind of stories the world tells, and every later creator will read it as a template. A weak second property does not just underperform; it narrows what the franchise is understood to be, and that is much more expensive to undo than to get right the first time.

Who should be allowed to add canon to a franchise?

As few people as possible, with everyone else able to propose. The failure mode at both extremes is well documented: one person as the sole bottleneck means the franchise moves at the speed of their attention, while open authority means canon drifts within a year and nobody can say what is true. The workable arrangement is a small named canon authority who can approve changes, a written record of what has been approved, and a defined route for everyone else to submit. What matters is less who holds the authority than that the answer is written down — an undefined process defaults to whoever spoke most recently.