Home / Policy Wind-Tunnel
Policy Wind-Tunnel & Cross-Impact Simulator
Four modules take you from a raw signal to a regulation design. Start anywhere: the example below shows a complete run on digital identity, and every field is yours to change.
Weak-signal triage
Scanning without boundaries produces an archive of interesting articles and no decisions. Anchor first, log each signal with the mechanism by which it could scale, score it, and let only independent clusters become themes.
1Anchor dependency
Not a topic like “technology”, but a function your organisation depends on and the assumption that keeps it working.
2–4Log a signal
The two-minute log. Five to ten strong signals per category beat a hundred weak ones.
5Independent cluster graph
A cluster needs at least three independent sources sharing one causal mechanism before it can become a theme.
6–7Themes, drivers and handover
A candidate becomes an Emerging Risk Theme when it passes four tests. Then interpret the driver and hand it to a named decision-maker.
8Signal log & decay
Signals that fail to join a cluster retire automatically after 730 days, so the radar stays clean.
Archived signals (0)
Scenario generator
Choose 6 to 10 descriptors of your system, each with 3 or 4 plausible states. Score how each state promotes or inhibits the others. The solver tests every combination and keeps only the internally consistent futures.
ADescriptors and states
Map a descriptor to a wind-tunnel driver to carry its states into Module 3 automatically (order states from the first to the last level shown).
BCross-impact matrix
Read each row as: “if this state holds, it promotes (+) or inhibits (−) the state in this column.” Pick a value, then click cells to paint. Hatched cells are a descriptor's own states.
CConsistent scenarios
Stress-test & premortem
Describe the policy, its core assumptions and how it affects seven lenses. The engine runs it through every scenario, flags trade-offs and shows which assumptions break.
APolicy under test
BCore assumptions
What must hold for the policy to work? Link each to the driver it depends on and the minimum level it needs.
CImpact profile
Baseline effect of the policy on each lens (−2 harmful … +2 helpful) and the driver each lens is most sensitive to.
DScenarios in the tunnel
Curated presets and any scenario you send from Module 2. Adjust the driver levels to match each story; hover a preset to read its narrative.
EWind-tunnel results
Scores run from −3 (harmful) to +3 (helpful). Score = baseline + (driver level − Medium), minus 1 more when the scenario is turbulent and the score is already negative.
Automated warnings
FAssumption premortem
Imagine the policy has failed in each future. Which assumptions broke? Draft a sturdier replacement and the early-warning signposts that would tell you it is happening.
Regulation configurator
Classify the kind of problem you face, then get the governance instruments that fit it: from codified good practice to a six-month sandbox, a citizen deliberation protocol or a verification blueprint.
ACynefin diagnostic
BContext
CDomain
DRecommended instruments
EGenerated draft
A starting text built from your answers and from Modules 1 and 3. Edit it before use.
FSix principles self-check
After Nesta's work on anticipatory regulation. How far is each principle built into your approach?
Want to run this with your team?
We facilitate signal sprints, half-day wind-tunnels and two-day regulation studios with your own policy and experts.