Confidence has to decay with distance, and a forecast that reads the same at one quarter and three years is not tracking anything.
Why it matters when the plan changes
Organisational judgements are routinely made without a stated horizon, which means the reader supplies one. A view about whether a team can carry a plan over two quarters is a different claim from the same view over three years, and the evidence supporting the first rarely supports the second. A judgement is bounded by definition, applying to a defined arrangement over a defined horizon rather than to a person's future in general, and without that boundary it cannot later be compared against what actually happened.
The tension is that buyers want longer horizons and evidence supports shorter ones. A three-year assessment of leadership capability is more valuable if true and far less likely to be. Stating a short horizon is commercially weaker and is the honest position, and it is what makes a review point meaningful rather than decorative. Scoring such a claim against what followed only works if the claim and its period were fixed in advance.
In practice
An assessment concludes that a leadership team can deliver a transformation. Asked over what period, the answer is that nobody specified. The evidence was gathered against the first eighteen months of the plan; the conclusion is being used to support a five-year commitment that no part of the analysis examined.
Evidence
Forecasts are predictions that are later compared with actual outcomes, which requires a defined period over which they apply.
Forecasting, Wikipedia (2026)Probabilistic forecasts are scored for how well stated confidence matched what happened, which requires that both the claim and its period were fixed beforehand.
Brier score, Wikipedia (2026)
What it cannot tell you
A stated horizon tells you the period a judgement claims to cover, not whether the judgement is correct within it. It cannot compensate for weak evidence, and a short horizon attached to an unsound method is still unsound; horizon length says nothing about the quality of the reasoning behind the forecast.
Questions
By how far the evidence actually reaches and how quickly the situation changes. A plan whose assumptions will be revisited in two quarters cannot support a judgement over three years, because the thing being forecast will have been replaced before the period ends.
Because the number of things that can change grows with time, and each one can invalidate the conditions the judgement rested on. A confidence that stays flat across different horizons is not modelling uncertainty at all; it is expressing a disposition.
Usually the period over which the plan's current assumptions hold, which for most transformations is two to four quarters. Longer claims are possible with wider uncertainty attached, and stating that widening explicitly is what distinguishes a forecast from an aspiration.
The judgement expires rather than being quietly extended. Either a fresh reading is taken against the current plan or the earlier conclusion stops being used. Extending a judgement past its stated horizon is the most common way a stale finding keeps informing decisions.
Yes, and it should be when something material changes. Any of the six triggers means the conditions the judgement rested on have moved, which effectively ends the horizon early, regardless of what date was originally attached to the judgement when it was made.