Build one to see the machinery. An automaker's PESTEL scan surfaces battery costs, fuel policy, urban congestion rules, income growth, charging infrastructure, ride-hailing economics. The impact–uncertainty screen sorts them: income growth and emission-norm calendars are high-impact but largely predetermined, so they're planning assumptions, not scenario material. Two survive as critical uncertainties: how fast EV adoption actually runs (policy resolve × battery economics × charging build-out), and whether personal ownership endures or shared mobility wins the city. Cross them and four 2030s appear: the Long Today (ownership endures, EVs crawl, so ICE upgrade cycles rule), the Private EV Boom (ownership endures, EVs sprint, so home charging is the battleground), the Fleet World (shared wins, EVs crawl, so durable ICE fleets and operators squeezing cost-per-kilometre), and the Robo-Fleet Future (shared wins, EVs sprint, so cars become a service business). None is a forecast. Each is a coherent world your strategy will either survive or not.
Narration is what makes scenarios usable rather than academic. Each future gets a name, a short story of how the world got there (what happened to policy, prices, and behaviour, in sequence), and its winners and losers, because executives don't stress-test against spreadsheet cells, they stress-test against worlds they can imagine operating in. The internal-consistency test does the intellectual work: in the Robo-Fleet Future, what happened to fuel retailing, parking real estate, and car finance? If the story contradicts itself, the scenario is a mood board, not a tool.
The wind-tunnel is the payoff. Fly each element of current strategy (the product portfolio, the capex plan, the dealer network) through all four worlds and sort the results. No-regrets moves win everywhere (platform flexibility that serves both ICE and EV, a service business that thrives under ownership or fleets): fund these without hedging. Contingent moves win in specific worlds: pair each with a tripwire, a named and observable indicator that its world is arriving (EV share of new sales crossing X%, a megacity capping private registrations), plus a pre-agreed response with an owner. Doomed moves fail everywhere and were usually sacred: retiring them is the exercise's least popular, most valuable output.
Scenarios also license a different investment style: real options. When a world is possible but unproven, buy the option, not the factory: a pilot fleet, a technology partnership, a convertible capacity plan. Then if the tripwire fires you can scale from a standing start instead of a cold one. This reframes 'wasted' hedges: the pilot that never scaled wasn't waste, it was the insurance premium on a future that happened not to arrive. Companies that only fund certainties are structurally last to every new world.