The method · Version 1.0
And for staying right about it as the world changes underneath you. It runs four movements around a fixed centre. Half the work is listening, a quarter is imagining, a quarter is validating. The output is a roadmap of validated customer experiences, each carrying the evidence that put it there.
And it never ends: from the first idea until the day you kill the product, the same loop runs — listen, imagine, pilot, adapt.
At the centre sits the Mission — the one thing that does not move.
Article III
Everything in the method is an application of one of these. When a situation is not covered by a rule, decide from the value.
III.1
“I would totally pay for that” is noise. “I have been doing this by hand every week” is signal. Every input to the system — trends, heroes, pilot results — is made of what people already do, never of what they say they would do. Payment is the only opinion that counts.
III.2
Not your product, not your platform, not your org chart. A roadmap that opens with “our platform” is the lattice talking. Every real trend has humans driving it; the work is to find them and lead with their story.
III.3
Nothing gets pitched without evidence, and evidence beats opinion regardless of whose opinion it is. This is the one hard rule of the operating system — and the one your organization will test first, usually via someone senior with a strongly-held view and no pilot.
III.4
Mercenaries chase whatever pays; missionaries pursue a mission and build the community that shares it. Missions attract co-creators; mercenaries attract nobody. Everything is permitted to change except why you exist.
III.5
Adoption can be engineered unethically; it just does not compound. Extraction burns the trust networks that adoption travels through, and those are the slowest asset to rebuild. Put AI everywhere in the work and humans firmly on the value decisions.
Article IV
IV.1
The winning playbook becomes process, process becomes structure, structure becomes identity. Nobody made a mistake — what went right simply does not stop. In the AI era, that hardening is fatal.
IV.2
A crystal is a structure where nothing can move without everything moving. Can pricing change without the product changing? Can a product die without a career dying with it? Each is a degree of freedom, and the loop exists to add them.
IV.3
A crystallized company that accelerates builds the wrong things faster. This is why the copilots did not move the needle: acceleration was bought for what was actually a shape problem.
IV.4
Revans’ law — L ≥ C: an organization survives only while its learning rate matches the rate of change around it. AI drove C up and made programmed knowledge free, so the whole burden falls on questioning insight. The method is an operating system for that.
IV.5
A fluid trend is real behaviour that is accelerating but not yet obvious, named, or crowded. By the time it is crystal clear, it has crystallized — consensus is crystallization with better PR. Build toward the emerging behaviour and ride it across the chasm.
IV.6
Nobody wants your product; they want what happens to them. Write the scene first — a specific person, in a specific moment, having something happen to them worth having happen. Products are delivery mechanisms you choose later.
IV.7
A pilot without a pre-agreed threshold is a launch with a modest budget. A graveyard of ideas killed by evidence is a vital sign, not an embarrassment.
IV.8
Nothing settles anymore, so every transformation project — anything with an end date — fails by design. Whatever you build must be permanent, running, and self-measuring: infrastructure, not initiative.
IV.9
Invention is creating something new; innovation is creating something new that gets used — repeatedly, voluntarily, creating value. Features shipped, demos applauded, pilots launched: all producible by an organization achieving nothing.
IV.10
Diamond and graphite are the same carbon — the atoms were never the problem, only the arrangement. Your people, data and customer relationships do not need replacing. The attacker cannot buy your assets; you can build their fluidity. Their gap closes in years; yours closes in quarters.
Each movement decouples something that was welded. Trends decouples sensing from planning — you can look without committing to act. Stories decouples understanding the customer from having a product. Experiences decouples designing from building. Pilots decouples trying from committing. That is the whole mechanism, and why the loop works where speed purchases do not.
The first turn takes real effort; every turn after it stores momentum. A first full turn, run seriously by a small team, is six to ten weeks. The second is four. By the fourth, the questions have sharpened and the evidence arrives faster. That is why it is a flywheel.
A real one — co-creators gathered around the mission, with an honest sales expectation built in from day one — and a synthetic one: AI digital twins of your heroes, critiquing pilots at machine speed so real people only ever see pilots that synthetic people have already sharpened. Synthetic signal is directional only; it never substitutes for a real person paying.
An experience graduates only with three sign-offs: synthetic (the twins converged positive), real-customer (real people validated it, including genuine commitment to pay), and leadership (the team commits to build it). What happens after the gate — how you build — is deliberately outside the method.
Agile makes you fast, Lean makes you unwasteful, Fluid makes you right. It replaces nothing. Design Thinking hooks into Experiences, Jobs to Be Done into Stories and Experiences, Lean Startup into Pilots, Agile into the build phase.
It loses to one precise strike on a cleavage plane. And the people who know where your cleavage planes are learned the trade inside companies like yours.
Article V
What to actually do, the artifact to fill in, and the specific way each step goes wrong.
Before you collect a single trend, write the mission — the change you intend to cause in the world, stated at a level that survives the death of any particular product. You do not need a product, a solution, or even a problem yet.
Then start the co-creation community — alongside Movement 1, not after it. State the premise in the first welcome message: we are going to build something and offer it to you. That expectation is not manipulation; it is the honest premise of the relationship, and it is what makes the eventual first sale natural instead of awkward.
Reward co-creators concretely — attribution, first access, discounts, contests; not equity. If you are an executive with existing customers, you already have the community; your work is to invite them into this initiative as co-designers. In parallel, build the synthetic community: digital twins of the heroes you will develop in Movement 2.
Worked example
Bring meaningful analogue experiences back into a digitally saturated world.
No product, no segment, no channel, no technology. Every one of those may change; the sentence does not.
Failure mode
A mission that lives in a deck and is never cited in a kill decision is decoration. An audience that consumes but never changes what you make is not a community. A community recruited on free stuff optimizes for free stuff.
You create nothing here. The work is a collection system, not a collection: 8–15 named sources, each with an owner, scanned weekly.
Sources in rough order of value: deep customer interviews, internal data patterns (usage, tickets, feature requests), market bets (acquisitions, new entrants), behavioural spikes, social mining — what people say to each other, unprompted — podcasts, market press.
Cluster signals into categories. Snail-mail clubs, sticker collecting, paper journaling and vinyl all point at one category: analogue experiences. One signal is an anecdote; five converging signals in a named category is a current. Then position against your mission — not a product, a category, ideally an existing market with a new category where fluid trends are converging and the space is up for grabs.
Why three to five and not fifteen: the constraint forces the ranking, and the ranking is the output. The confidence number is Bayesian discipline made concrete — write in advance what evidence would move it, or you will move it retroactively.
Failure mode
Collecting everything and prioritizing nothing. Technology fads mistaken for value shifts — the question is never “what is now possible” but “what are people now doing.” A quarterly trend-report ritual instead of a maintained instrument. Sources with no named owner: the most boring failure and the most common.
If a trend is real, humans are driving it. Develop them as heroes, with a novelist’s depth, extrapolated from the trend data you already hold.
Who are they, what world do they live in, what do they want, what is in their way, what transformation are they after. Capture their voice verbatim from the threads and interviews — you will write pilots in their words, not yours.
Then map their current journey at the value layer — each step a value handoff, never a screen — and find the friction and the value peaks to protect. The journey map is your allocation instrument: it tells you where to point people and money.
The Storyboard holds two to five heroes, no more. Pilots will confirm some, kill others, and reveal ones you never imagined.
Failure mode
Personas instead of heroes — “aged 35–50” produces nothing. Skipping to solutions: ban “we could” from this movement entirely. Inventing the voice instead of sourcing it. A hero whose obstacle is conveniently solved by a capability you already have.
Now, for the first time, you create — and what you create is scenes, not products. Take a hero and write, in vivid multi-sensory prose, experiences she would love.
A specific moment, what happens, what she feels, what value is handed to her. No feasibility constraints: “can we build this?” and “what would it cost?” are banned here. Premature practicality collapses the possibility space toward what you already know how to build.
Every scene notes three things — what happens, what the hero feels, and what value is delivered. The third is what makes it testable. Every experience also lists its riding assumptions with confidence levels, written while you are still in love with the idea.
The Portfolio holds five to fifteen experiences per positioning. At least two should make someone senior uncomfortable — a portfolio of safe swings produces no new value.
Worked example
Still in her pyjamas, she walks to the mailbox — a thick, hand-addressed envelope, unlike anything in the pile. Heavy paper, faint lavender. This month’s letter, and a small collectible sticker for the fridge. She reads standing in the kitchen and tears up a little. Her day starts differently now.
The big swing: letters written as smuggled correspondence from a priest imprisoned in 1950s communist Romania — paper aged, one corner burned, a fragment of prison-uniform cloth folded inside. She is not reading about history; she is holding it. Feasible? Worth $10–15 a month? No idea — that is what Pilots are for.
Failure mode
Feasibility leaking in — “we could do that on the existing platform,” and ten minutes later the portfolio is a feature list; ban “we” from scene-writing. Products disguised as experiences: if you can build it directly from the sentence, it is not a scene. Writing for the buyer instead of the hero. Scenes with no value handoff.
A pilot is an invitation to real people to try an experience with you — a minute, or a month. Not an experiment run on subjects; something people participate in, often as co-designers.
The threshold comes first. Before any pilot runs, in writing: what assumption is under test, what confirms it, what kills it, what you will do in each case, and the decision date. If the kill condition cannot be written, it is not a pilot — it is a launch with a modest budget.
Synthetic first. Twins go through the pilot, generate feedback on value, delight, confusion and fit; an assistant agent tweaks; repeat until strongly positive. Machine-speed quality control, so real attention is never spent on an obvious miss. Directional only — it never substitutes for payment.
The strongest format is dynamic co-design: preflight questions personalize the pilot to their actual life; an interactive narrative unfolds beat by beat with the participant choosing; a live customization panel redesigns the experience in real time. “I hate lavender.” “What do you love?” “Roses.” Each participant hands back a personalized version they co-designed until they liked it — and people want what they helped shape.
Every pilot result flows back: update the Compass, the Storyboard and the Portfolio. Kill the belief, not just the idea, or you will regenerate the same idea in two quarters.
Graduation is three sign-offs: synthetic, real-customer with genuine commitment to pay, and leadership. Only then do you build — and you sell back to the community first. If they do not pay, they were never customers, and the loop needs to hear that too.
Failure mode
No threshold — most innovation theatre is this one omission. Enthusiasm counted as validation; compliments are the cheapest thing a human produces. Pilots run on friendlies. Ambiguous results resolved upward because stopping is socially expensive. Synthetic sign-off substituted for real payment: self-deception at machine speed.
Article VI
Every transformation project reverts because of four break mechanisms. An installed loop has one component per mechanism.
The test of installation
It runs when no one is excited. Budgeted like infrastructure, governed like security, survives leadership attention cycles.
Article VII
Plus a Scoreboard, reported to the board quarterly like financials.
Split failure publicly in two: an invalidated assumption that produced learning is celebrated; an avoidable mistake is managed.
What happened, publicly, to the last person who killed their own project with evidence? Your organization already knows the answer and is behaving accordingly.
The community starts on day 1 and never stops. Then keep turning — the second turn is half the effort of the first.
Version 1.0, September 2026. Reconciled from the Key Ideas set, the methodology appendix and outline v4. Where those disagreed, the four movements follow the two most recent sources — Trends, Stories, Experiences, Pilots — and the earlier Journeys/Experiments naming is superseded, with the journey kept as a technique inside Stories. Nothing was silently chosen.