DE
EN
The business stops buying
A Tuesday morning in 2036: the owner of a twelve-person business checks utilisation on three rented machines from his phone and hands one back, because the job never came in.
Future Round Beta
Strategic foresight with a panel that disagrees.
One question about the future. Five to sixteen synthetic personas answering individually — not as a chorus. Out of that come contrasting future images and a roadmap that runs backwards from the goal to next quarter. Not an oracle, a thinking tool.
The difference sits in one detail: the two axes of uncertainty everything ends up on are not set in advance. They come out of the answers.
Status: The Future Round is built and running — we are opening it gradually and with support for beta participants rather than rolling it out quietly. All voices are synthetic.
§ I How it runs
From one question to steps that fit into a quarter.
- 01Panel round
Every voice answers three times over.
You ask a question about the future and pick a horizon — five, ten, fifteen, or twenty years. Five to sixteen personas then answer individually, each from their own biography: what they expect, what they wish for, what they fear. Plus one concrete moment from their day in the target year, and one possibility that still sounds absurd today.
Expects · Wishes for · Fears
- 02Future images
Many voices become a few images.
From the individual answers, the Future Round distils three to five contrasting future images. Each carries a narrated scene from the target year, the opportunities it opens and the tensions it creates — and the names of the personas who carry it. From there you are one click away from the individual voice.
Three to five images · each with named support
- 03Roadmap
Backwards from the goal to next quarter.
You pick one or two images. For each, the round works backwards: target state, obstacles, opportunities, then milestones in two-year steps. At the end there are three to five first steps today — small enough for a single quarter.
Target state → milestones → first steps today
§ II The map
Two axes that come from the panel — not from a template.
In classic scenario work, an analyst fixes the two axes of uncertainty up front. Here it runs the other way: the axes are distilled from the panel's answers. Whatever genuinely pulls the panel apart is what spans the map.
Every future image gets a position on it. If two images sit too close together, a check discards them before you see them — otherwise you would be holding four cards that say the same thing.
§ III Example
What a result looks like.
Illustration · invented example, not a real round
- Question
- How will trade businesses in our market buy their machines ten years from now?
- Horizon
- 10 years · target year 2036
- Panel
- 12 voices from the trades, procurement, succession, and financing
Two of four future images
Succession decides
A Tuesday morning in 2036: the daughter took over two years ago, and for the first time the investment decision is not made over a beer but on a comparison sheet she built herself.
First steps today
- Ask ten existing customers who actually picks the machine — and who only signs off.
- Review the quote template for rental options: is the purchase price still the first line?
- Sit in on one business handover in your customer base and write down what the successor asks.
In the report, every future image hangs on the voices that carry it — from the map you are one click away from the individual persona and can read what her position is made of.
§ IV Methodology
Where the structure comes from — and why it isn't just a longer prompt.
Futures research has spent decades building defences against the obvious failures: extrapolating the present, settling on one picture, stopping at the pretty picture. Those defences sit in the Future Round as fixed rules — here is which ones.
Ethnographic Futures Research
Expects · Wishes for · Fears
The triad comes from ethnographic futures research: people answer differently depending on whether they expect a future, wish for it, or fear it. We ask all three separately — otherwise the worry disappears behind the expectation.
After: Robert Textor, Ethnographic Futures Research
Dator's Second Law
The wildcard requirement
Every persona has to name a possibility that still sounds ridiculous today. That is not a party trick but a countermeasure: models, like people, extrapolate the present. Without the requirement, every round stays well-behaved.
After: Dator, “Any useful idea about the futures should appear to be ridiculous”
Future persona / design fiction
The narrated scene
Every future image carries an everyday moment from the target year — “A Tuesday morning in 2036…”, three to five sentences. A scene sticks and can be argued with in a workshop; a list of drivers cannot.
After: Fergnani 2019, Foresight (Emerald)
Participatory backcasting
Backwards, not forwards
Instead of extrapolating forward from today, the round sets the target state and works its way back. That is exactly why it lands on steps that fit into a quarter rather than a wish list.
After: Robinson 1990 · Quist & Vergragt 2006
Amara's observation, structurally enforced
The Amara correction
We overestimate technology in the short run and underestimate it in the long run. That correction sits as a fixed rule in how milestones are derived — not in the intuition of an individual persona.
After: Attributed to Roy Amara, Institute for the Future
A design decision
Enforced divergence
Language models pull toward the middle. The Future Round is built against that tendency: a deterministic check discards future images whose axis signatures are too similar — before you ever see them. Otherwise you end up with four cards saying the same thing.
All of the traditions named here were developed for human respondents. Transferring them to AI personas is an adaptation, not a validated use case of the research cited.
Not built yet: Delphi, the futures wheel, wind-tunnelling, three horizons.
§ V Limits
What the Future Round is not.
Not a forecast.
The Future Round assigns no probabilities and does not say what will happen. Every technical assumption in the report carries the label “modeled assumption, not a forecast”.
Synthetic voices, not a survey.
All voices are AI-modeled personas. A Future Round replaces neither an expert survey nor market research — it structures the thinking that comes before them.
These methods were made for people.
The traditions cited here were developed for and applied to human respondents. Transferring them to AI personas is an adaptation, not a validated use case of the research cited.
Projection, not transformation.
The personas are not aged and not turned into future humans. They stay who they are and look forward from there.
§ VI Who it's for
Built for the moment before the strategy decision.
For innovation and strategy leads, for consultancies, for teams who have to work through a question about the future before budget is committed. Typical occasions: a market entry that has been discussed in circles for months. A portfolio decision where two camps block each other. A regulation whose effect only shows up years out. Or simply a strategy offsite that should not end at the gut feeling of the loudest person in the room.
The report is built as a workshop handout: overview, map, roadmap, and voices — four tabs, plus a PDF report you can print and put on the table. A round is repeatable: with a new question, a different panel, a different horizon.
Why foresight workshops so often end in innovation theatre — and what changes that structurally: the column “We had the nicest poster in the building” from the perspective of an innovation manager.
§ VII Questions
Frequently asked questions about the Future Round.
What is a Future Round?
A Future Round is a mode in Radical Personas built for strategic foresight. You ask a question about the future and choose a horizon; a panel of five to sixteen synthetic personas answers individually, three to five contrasting future images are distilled from those answers, and for the images you select the round works backwards to first steps for the coming quarter. The result is a report built as working material for strategy workshops.
Is a Future Round a forecast?
No. No probabilities are assigned and no claims are made about what will happen. Every technical assumption in the report is labelled “modeled assumption, not a forecast”. The value lies in surfacing assumptions and tensions solid enough to carry a strategy conversation.
Why not just ask ChatGPT for future scenarios?
Because a single prompt returns a single, smoothed answer. The Future Round questions five to sixteen fully developed personas separately, keeps expectation apart from wish and fear, requires one uncomfortable possibility per voice, and discards future images that are too similar to one another. The disagreement is the product, not the consensus.
What happens if the panel agrees?
Then that is what the report says. Agreement is a result and is not broken up artificially — but it is not sold as confirmation either. You see which voices carry it and where it starts to fray.
Who is this for?
For innovation and strategy leads inside companies, for consultancies, and for anyone who has to work through a question about the future before budget is committed. The report is built as a workshop handout, not as a dashboard.
How do I get into the beta?
Through the contact form. Tell us briefly which question about the future is on your mind and what field you work in — we will get back to you and set up access. We are keeping the beta small and hands-on.
Where the method hits its limits
Synthetic personas represent average perspectives — not edge cases, and not real conversion data. For final validation, there is no way around real users. Synthetic personas complement real user research — they don't replace it.
§ VIII Getting started
We are keeping the beta small and hands-on.
Tell us which question about the future is on your mind and what field you work in. We will get back to you, set up access, and walk you through the first round.
We reply within 24 hours on business days. You will find out in the conversation when and at what scale we can take you on — nothing is committed before that.