J. Paul is a London based designer and researcher with expertise in Speculative Design, Service Design, Design Research, and Strategy.
NKD_6 — We'll never know what will happen tomorrow, until tomorrow.
This is the principle that most directly motivates the existence of speculative design as a discipline. If we could forecast tomorrow accurately, speculative design would be a curiosity. We cannot, and so it is not.
The misconception it corrects
The conventional logic of strategy is to predict the future and prepare for that prediction. Better data produces better predictions; better predictions produce better preparations; preparations executed well produce competitive advantage. This is the implicit story of forecasting, trend analysis, scenario planning, and most strategic consulting.
The story has a hole in it. Tomorrow is not knowable today. Not in the cases that matter — the inflection points, the discontinuities, the genuinely new — and these are the cases strategy is supposed to address. In the cases where the future is predictable (linear extrapolation, well-understood markets, mature categories), strategy is mostly a matter of execution and does not need forecasting at all.
The places where strategy creates real value are exactly the places where forecasting is least reliable. NKD_6 names this paradox.
What it changes in practice
Strategy optimises for breadth of preparedness, not depth of bet. The right question is not what is most likely to happen? but what is the smallest set of preparations that covers the widest set of plausible tomorrows? This is the logic of speculative design, of scenario planning done well, and of any strategic posture that takes its own uncertainty seriously.
Forecasts are owned as accountable artefacts. When forecasts are made, they are signed, dated, and revisited. Most forecasts in most organisations are never re-examined; once they have served their purpose in justifying a decision, they are quietly retired. NKD_6-trained organisations track forecasts as falsifiable claims, and review their own forecast accuracy over time. This is a humbling exercise. It also, very quickly, produces better forecasters and — more importantly — less reliance on forecasting.
Strategy meetings ask "what would tell us we're wrong?" before "what do we do?" Inverting the conventional order of strategic conversation. The triggers for changing course are defined before the course is set. This is harder than it sounds; it tends to surface the fact that many organisational strategies have no concrete falsification conditions, which is to say no real decision logic.
A working method: the preparedness matrix
The technique we use most often for NKD_6 in client work is the preparedness matrix.
Construct a 4-by-N grid. The columns are the four to six plausible futures that emerged from your speculative or scenario work. The rows are the strategic moves on the table — a hire, a partnership, an investment, a product, a market entry, a policy position.
For each cell, mark one of three values:
- •+ This move helps in this future.
- •− This move hurts in this future.
- •· This move is neutral in this future.
A move with mostly + and · is robust — it pays off across the considered futures and rarely hurts. These moves should usually be funded first.
A move with mostly + in one column and − in others is specific — it is a bet on one particular future. These moves require explicit confidence in the underlying forecast, and an explicit unwinding plan for the other columns.
A move with mostly − is rejected.
The matrix is not new — variants of it appear in the futures-studies and corporate-strategy literature. The NKD_6 emphasis is that robust moves are systematically undervalued by organisations who treat strategy as a sequence of bets on predicted futures. The matrix makes the trade-off visible, and the act of writing it tends to surface the fact that the supposed bets often have weaker arguments than the boring-looking robust moves.
A note on what this principle does not say
NKD_6 does not say that forecasting is useless. Short-horizon forecasts, in well-understood domains, are valuable and should be done. Forecasting models in mature industries are part of competent management. What NKD_6 says is that the further out, the more inflectional, and the more strategic the question, the less useful forecasting becomes as the primary input to the decision. At those distances, breadth of preparedness beats depth of prediction.
It also does not say that prediction is bad and possibility is good. The discipline is not to abandon estimating the future; it is to estimate it honestly, with the regret-budget exercises and surprise logs and reversibility-first design that earlier principles in this series have argued for.
Where this principle lives in our work
NKD_6 is the principle that connects most directly to the day job of speculative design. Every framing workshop we run starts, implicitly, from this principle: we cannot know tomorrow, so we will instead build several plausible tomorrows and use them to interrogate present decisions. If you want a single working method that takes NKD_6 seriously, the entire speculative design methodology is it.
Key takeaway: at the strategic distances that matter, forecast accuracy is unreliable and gets worse the further out you go. Optimise instead for breadth of preparedness across plausible tomorrows, and treat moves that pay off across many futures as more valuable than they look.