In 2019, the French Ministry of Defence set up a unit called the Red Team.
It is a group of science-fiction authors and illustrators, working alongside researchers from Paris Sciences & Lettres and officers from the armed forces. Their job is to imagine the conflicts France might face between 2030 and 2060: technological shifts, resource disputes, entirely new kinds of adversary.
Part of the output is classified. The rest is published. Four seasons are online, in text and in video, and anyone can read them.
An army has every reason to keep its threat analysis in a safe. This one decided that some of it was more useful as a story people could read. That decision is worth understanding, because the same logic applies to any organization making choices that commit beyond the current budget year.
A scenario written as a narrative, with characters and objects and a date on it, does something a threat assessment cannot. People remember it. They argue with it. They bring it up six months later in a meeting about something else.
Two methods produce that effect deliberately. One builds the structure, the other builds the object.
What scenario planning gives you
Shell began experimenting with scenarios in the 1960s, and Pierre Wack built the practice through the early 1970s. When the oil market turned over in 1973, Shell’s management had already spent a year discussing what they would do in exactly that situation. They were not right about the future. They were ready for a version of it.
The method most teams use today runs in four steps.
- Map the forces at play over the horizon you picked: regulatory, technological, economic, societal, geopolitical.
- Isolate the two critical uncertainties — strongest impact, least predictable outcome. A variable whose trajectory you already know belongs in the background as a trend.
- Cross the two axes. You get four quadrants, so four distinct and internally coherent worlds.
- Work each world through. What becomes scarce, what becomes ordinary, who wins, who disappears, what your company is still worth in it.
Three benefits come out of this, and any one of them justifies the effort.
You act faster when something moves. The company can say what it would do if a raw material became unavailable, if a use was banned, if a market closed — because the case was worked through before it arrived. What you gain is reaction time.
You get a resilience read of your portfolio. The exercise lays out what stays relevant across all four worlds and what only holds in one. It is the fastest strategic diagnosis available, and it produces a list of low-cost options you can start now to stay present in the futures you did not bet on.
Long-horizon investments become arguable. A technical capability is hard to defend in the abstract. Attached to a world that makes it necessary, it turns from a lab bet into a reasoned choice.
What design fiction adds
Design fiction represents a future rather than summarizing it. You build an object that only makes sense inside a given world: a product sheet, a newspaper page, a user manual, an interface, a short film. The reader reconstructs the world from the object. Bruce Sterling coined the term in 2005 and Julian Bleecker turned it into a practice in 2009. The Red Team is this method applied at national scale.
Five rules separate an artefact from a gadget.
Push the hypothesis all the way. The raw material is gone. The use is banned. Demand collapsed. A future that commits makes people commit with it.
Ship an object, not a conclusion. People reconstruct the reasoning themselves, and that reconstruction is what produces agreement.
Stay plausible in the details. Price, format, regulatory mentions, vocabulary, layout. Credibility is what separates this from science fiction.
Accept some discomfort. The useful artefact shows a future where the company, as it exists today, has to change to have a place.
Open the discussion. The artefact settles nothing. It engages a register no analytical document reaches, and the decision comes afterwards with the usual tools.
Two more benefits come from this half, and they are the ones most people underestimate.
Your long-term work becomes visible. An analysis is read by a handful of people. An object is shown in three minutes to a committee and quoted in meetings afterwards. For any function whose production is intangible, that visibility is the difference between existing at board level and existing in a shared drive.
The same work feeds your external voice. The artefacts built for internal decision-making are exactly what you publish. Concept cars work this way: the object will never be sold, and it installs the idea that the company showing it sees further than the others.
It also pays commercially
In 2011 Corning, the specialty glass manufacturer, made a six-minute film called A Day Made of Glass. It follows an ordinary day in a near future where the company’s glass is everywhere: kitchen windows, car dashboards, classroom walls, hospital displays. No product demo, no commercial promise. A family gets up, goes to work and school, comes home.
It was made for an annual investor meeting.
It became the most watched corporate video in the world, somewhere around 26 million views, and Corning’s executives have described what followed: inbound enquiries about products that did not exist yet. They shot a sequel, produced explainer videos for each technology in the film, and ran a symposium off the back of it.
A materials manufacturer bought itself a decade of narrative with one short film.
Superflux, a London studio, works at the other end of the scale. They built a complete apartment from a 2050 in which food supply has broken down. Visitors walk in, open the pantry, sit at the table. Nothing is explained; the world is deduced from the room.
Three very different budgets, three very different subjects, the same mechanism.
What a cycle actually costs
Smaller than people expect.
The analytical half runs in workshops with a documentation effort behind it. The artefacts, which used to require studio resources, now come out of generative AI tools for text, image and video — the quality jump of the last two years is what makes this accessible to teams that could never have afforded it before.
A full cycle: two validated axes, four documented worlds, one or two artefacts each, a portfolio read, and a list of no-regret moves. Three to six months, no heavy investment.
Where to start
Pick your two critical uncertainties and get them validated at the level that matters — unshared axes disqualify everything downstream. Work the four worlds through properly. Then build one object from the world you find least comfortable, and put it on the table at the next committee.
The conversation that object starts will teach you more than the report you would have written instead.
We run this as a three to six month cycle, and the output is something you can show. If that is useful where you are, get in touch.