Skip to content
Agentic·Foresight

Lever 03 · Tunnel

What do the scenarios mean for your strategy?

A Tunnel agent tests your strategy against every scenario, like a model in a wind tunnel, and shows where it holds and where it breaks. From that it derives actions and early-warning indicators; judging, prioritising and deciding remain human tasks.

Sketch: a wind tunnel with a glowing orange fan. The flow passes a stepped strategy model and turns turbulent behind it; a robot takes notes.

01 / Process

From scenario to decision in five steps

  1. Model the strategy

    Business model, resources, dependencies and the assumptions the strategy rests on are written down explicitly and in a testable form.

  2. Run each scenario

    The agent plays the strategy through in every scenario: which assumption holds, which does not, where do bottlenecks or new opportunities appear?

  3. Show robustness and gaps

    Strategic fit is set out scenario by scenario. Robust is what holds in every world; critical is what only works in one.

  4. Derive actions

    Three kinds of response: prepare for what every scenario requires; hedge where an assumption breaks; create options that can be exercised once a scenario starts to unfold.

  5. Set early-warning indicators

    For each scenario, indicators are defined that show early on which direction the world is moving in. They go back to Scout and are monitored continuously.

02 / Division of labour

What the agent does, what humans decide

StepThe agentHumans decide
Strategy modelstructures documents, phrases assumptions so they can be testedconfirm that the model reflects the strategy correctly
Runschecks the strategy against each scenarioreview the reasoning at the breaking points
Robustnesssets fit, gaps and critical assumptions side by sidejudge which gaps matter
Actionsproposes prepare, hedge and option movesprioritise and decide
Early warningderives indicators and monitors themset the threshold at which an indicator triggers action

03 / Limits

What Tunnel cannot do

  • The model is not the company. Whatever is missing from the strategy model, such as culture, relationships or tacit knowledge, is not tested either.
  • The test is only as good as the scenarios. If all scenarios sit too close together, every strategy looks robust.
  • The agent provides reasoning, not certainty. A measured breaking point is a reason to look closer, not a verdict.
  • It cannot decide or take responsibility. Weighing risk, opportunity and resources stays with humans.

04 / In practice

What this looks like in practice

Strategy Review exposes a company to artificially generated future worlds and measures where it holds and where it breaks. The result is a decision paper with implications for each future world, assessed options, breaking points and early indicators.

Have your strategy tested in the wind tunnel

Strategy testing with your own agents: Operation Zukunft builds and runs AI agents for innovation and foresight.

Build agents for your foresight work →

Published: 16 November 2026 · Last updated: 16 November 2026