When the 2011 earthquake hit Japan, Toyota found something more unsettling than the damage inside its own plants: it could not see its own supply chain past the first tier of suppliers. The parts that stopped the production line sat three and four layers removed from any contract Toyota had actually signed. Toyota’s answer was not a better forecast. It was a map of every supplier, and every supplier’s supplier, so that the next shock would be understood in days, not months.
That distinction ran through the panel I shared with BDI at the Techne Summit, on strategy in uncertain times: the organizations that come through a shock in better shape are rarely the ones that called it first. They are the ones that could already see who else in their market system the shock would move, which scenario they were already rehearsing, which of their own borrowed practices never carried the conditions that made them work elsewhere, and which of their dependencies a crisis would expose before anyone had thought to audit them. Strategy in uncertain times is not a forecasting problem. It is a systems-visibility problem.
Weak Signals Are Data, Not Noise
The institutions that move early on a shock are not the ones with the sharpest forecast. They are the ones that treat weak signals as data rather than noise: a shift in customer behavior, a supplier’s change of tone, something a frontline team keeps mentioning that never makes it into a report. The reactive institutions usually saw the identical signal. It simply died somewhere between the person who noticed it and the person with the authority to act on it, because the path from signal to decision was long, and because nobody asked the one systems question that matters before any other: who else in this system will the shock move?
Shipping companies that rerouted around Africa early, as the Red Sea became unsafe for merchant traffic, kept control of their own lead times and their own costs. The lines that waited for certainty paid for both. The effect reached well past shipping: it reached the Suez Canal’s revenue, and through that, the wider Egyptian economy. The shock was never only a shipping problem. It was a systems problem that happened to announce itself first in shipping.
A Scenario That Does Not Change a Decision Is a Story
Most organizations that run scenario-planning exercises are, without quite admitting it, writing stories rather than plans. A scenario earns the name only if it changes a decision: two or three scenarios, built on the two uncertainties that matter most, each with a clear trigger, a first action and a named owner. The useful discipline sits in what a scenario is not. It is not a prediction, and treating it as one is how a ‘small bet’ turns into a bet the whole organization is making, without anyone having chosen to make it. Separate the moves that are right whichever scenario arrives, the ‘no-regret moves’, from the options worth leaving open, review the set every quarter, and the scenario stays a live decision tool rather than a document that was once presented to a board and has not been opened since.
Shell ran exactly this discipline in the early 1970s, modeling a scenario in which oil-producing states took control of price. When the 1973 crisis arrived, Shell was better prepared than its competitors, not because the scenario predicted the crisis correctly, but because it had already changed real decisions rather than produced another report. The practical version of the same discipline is close to home for any enterprise: if an import were disrupted for three months, what buffer stock would be needed, and who the alternative suppliers are. That decision gets made now, while it is still cheap, not in the first week of the disruption, when it is not.
Borrow the Principle, Not the Package
A best practice travels badly when it is copied whole. The useful question is never whether something worked elsewhere; it is why it worked there (which regulation, market structure, culture or infrastructure made the difference) and whether that same condition exists here. Where it does not, the question becomes what the local equivalent would be, tested through a small pilot, adapted, and checked against what local users actually say before anyone talks about scaling it. This is the systemic lens I have applied for years in gender-sensitive finance, turned toward a different question: look at the barriers behind a solution (socio-cultural, regulatory, market), not at the label attached to it.
Walmart carried its American retail model into Germany almost unchanged: the same pricing and service model, the same management approach, into a different employee culture, a different service expectation and different customer behavior. It withdrew from the market after years of losses. The model had not failed; the context had never been asked whether it would accept the model. M-Pesa is the same lesson from the other direction. It succeeded in Kenya because it had an agent network and a community trust that the model depended on entirely, and the idea needed substantial adaptation everywhere it was carried afterwards, precisely because that trust and that network could not simply be assumed to exist.
What a Crisis Teaches an Institution About Itself
No institution sets the date of its own audit; the crisis sets it instead, and what it reveals is rarely where the organization expected to look. It shows where the real decision power sits, which is often not where the organization chart says it sits. It shows which relationships hold under pressure: with clients, with partners, within the team itself. It shows how much of the work depended on a single client, a single funder or a single person, a concentration risk that a calm year lets everyone ignore. And it shows who told the truth early, and what happened to that person afterwards, which is as honest a reading of an institution’s culture as any engagement survey will ever produce. The strongest institutions do not simply survive the crisis; they keep the lesson and turn it into standing practice, rather than letting it fade back into normal operations once the pressure lifts.
Egypt’s dependence on Russian and Ukrainian wheat, laid bare by the war in 2022, is the same lesson at the scale of a national economy. The company-level version of it is identical in shape, only smaller: a single supplier, a single client, a single funder, a single route. The crisis does not create the dependency. It only insists on being noticed.
Move Fast on What Is Reversible
The panel closed on the simplest test of all, and it is the one I use myself when a decision has to be made faster than the evidence is ready: decide quickly to protect people and clients, communicate transparently, and activate the no-regret moves already identified, such as diversifying suppliers and markets. Slow down, deliberately, on the decisions that cannot be undone, such as a large investment, a restructuring or a round of layoffs, and wait for one more signal before closing the option off. Amazon calls this the distinction between ‘one-way’ and ‘two-way’ doors: a decision you can walk back easily is taken quickly; a decision with no way back is given the time it needs.
None of the four readings above is a forecasting technique. Each is a way of seeing the system clearly enough that the right decision becomes obvious once the signal arrives. That is an advantage uncertainty cannot take away.
I speak on strategy in uncertain times, systems thinking and enterprise finance at sector events. My talks and speaker profile list the topics and formats.
Work With Me on This
I work with institutions and enterprises on strategy built on a clear view of their market system: who moves when a shock arrives, which scenarios change a decision, and which dependencies need attention before a crisis finds them. See the Service →