Syllabus · Innovation Territories Sprint
Every sprint reads a brand through six cultural panels, works the evidence with the six frictions of the Curiosity Engine, and judges the candidates on a six-criterion scorecard. This page defines all of it: what each part asks, what a real finding looks like, and the way each one fails. Keep it open while you read an atlas.
The order of a run
Three parts, three jobs. A panel is a stratum of evidence, researched on its own, without knowing what the other panels found. A friction is a move that takes away the answer anyone would give, so the next answer is further out; the engine runs six of them on the panels' evidence and ends with at least ten candidates. The scorecard is a separate tool that judges those candidates the way an innovation committee judges any bet.
The test of whether the method ran: delete the six panels. If a territory still stands, the panels were decoration for a decision already made.
Part one
Six strata of the same brief, each researched separately. Each one sits in one of the four places a diagnosis looks: the company, the category, the consumer and the culture. The point where two or three panels land on the same thing without having spoken is the territory's first proof.
Pre-modern myth, ritual and archetype sitting underneath the obvious symbols of the brief.
What the neighbouring premium categories have already made normal for the same buyer.
Where cohort wants split, and which cohort the brand will have to choose.
The conventions the brand has to refuse for the new thing to be born at all.
Forces with no interest in the category that are reshaping who buys, why, and how.
What the brand's own substrate encodes that the current marketing has buried.
Part two
A language model returns the most probable answer: the centre of everything already written on the subject. There is no setting for less probable. So the engine never asks for something creative; it takes away the answer it was about to give, and the next one down is further out. Six ways of doing that. Five are procedures with a track record older than any model, and each names where it comes from. One is ours.
Each friction runs as a written step in the run file. A move that leaves no written receipt does not fire.
Write down the answer anyone would give, set it aside, and look for at least one direction that comes from a force outside the category, followed to the consequence of the consequence.
The brand’s meaning is found inside the brand: the archive, the object and the word read literally, and the myth the product already sits inside.
Flip the problem, the category convention and your own first reading. Refuse the rule the category runs on. And when two opposite truths both hold, keep both instead of choosing.
Borrow structure from another field: a behaviour the category next door already allows, a subculture, a book, a ritual, the way people talk online right now.
Take a tool out of the hand and solve the problem anyway. The cut falls on what the obvious answers lean on, chosen so its absence says something about the diagnosis.
Move the ground instead of the answer: build a world with its own rules and ask for its paperwork, a job listing or a statute, never for ideas.
Part three
The frictions produce at least ten candidates. The scorecard decides which are worth a bet. It is built from the standard innovation screens: desirable, viable, feasible (IDEO; Tim Brown, Change by Design, 2009), with the questions from George Day’s “Is It Real? Can We Win? Is It Worth Doing?” (Harvard Business Review, 2007), plus one column from the engine.
Every candidate is scored, killed ones included. Each criterion goes from 0 to 3, and every score carries its evidence and its tag. A score with no evidence is a 0.
| Criterion | Lens | The question | A 0 | A 3 |
|---|---|---|---|---|
| Felt tension | Desirable | Do people already feel this? | It tells people a fact | People already carry it and cannot yet say it |
| Size of the prize | Viable | Is the market real, and how big? | No market logic | A sourced figure plus a plausible slice |
| Open space | Viable | Who is already there, and why is it free? | Rivals already do it | Open, with the reason it is empty |
| Worth doing | Viable | Does the money work? | No price or margin logic | Price and margin source named |
| Right to win | Feasible | Can this brand say it, and can the company make and sell it? | Nothing ties it to the brand | A brand asset and a capability, both evidenced |
| Distance | Engine | How far is it from the answer anyone would give? | It is the obvious answer | Far from it, measured |
Two marks travel with every score: the horizon (H1, H2, H3, after Baghai, Coley and White, 1999) and the moment of the tension (residual, dominant or emergent, after Raymond Williams, 1977).
Three clean answers to the brief, asked straight, are generated before any friction runs. A candidate that a separate judge model reads as one of them does not ship without a written reason.
Remove the six panel readings from the run. A territory that still stands came from the news or the market data, and was decided before the cultural work began.
The scorecard is a profile, never a sum. Market size and margin are estimates at this stage and never kill a candidate before the business case is built.
Three territories ship side by side, each from a different friction, each strong in a different place. The killed candidates are shown with the reason they died, because the best idea in a run is sometimes in the killed pile.
Part four
The panels and the frictions produce claims. Two rules decide whether a claim is allowed to carry weight in the deliverable.
Provenance
A claim that fits none of the three does not ship.
Region lock
The brief locks one geography, and every cultural and consumer source has to come from it. A sprint locked on the United States cannot prove a US territory with a UK statistic, even when the underlying pattern is real everywhere. Company-level facts are the one exception, because they are facts about the brand rather than about the locked market.