The less an organisation knows about how an initiative will unfold, the more it needs a plan — because the plan's real function is not prediction but detection.
Two arguments are usually made in the same room, by the same executives, about the same initiative.
The first: we cannot plan this properly, because too much is unknown. The second: we cannot fund this properly, because no one has planned it. Both are stated with conviction. Both sound like prudence. Together they produce a familiar organisational deadlock in which the most consequential work an enterprise undertakes receives the least disciplined thinking, precisely because it is the most consequential.
The deadlock is not caused by uncertainty. It is caused by an assumption about what a plan is for — an assumption so widely shared that it is rarely examined. If a plan is a forecast, then genuine uncertainty does make planning futile, and the first argument wins. If a plan is something else, the deadlock dissolves.
Stanley Portny, writing for a scientific readership in 2002, put the question sharply: when should project management begin on work whose outcome cannot be known? His answer was blunt. At the outset. And his reasoning was the inversion that matters here — uncertainty is not the obstacle to planning, it is the argument for it. [FACT CHECK REQUIRED: the Eisenhower formulation Portny opens with — "plans are useless, but planning is indispensable" — is widely circulated with varying wording and attribution.]
That inversion deserves an executive reading, because it changes how capital should be committed to work the organisation does not yet understand.
The Strategic Context
Most enterprise planning discipline was designed for a world of estimable work. Its instruments — the business case, the baseline, the variance report — assume that a competent team can describe a future state, cost it, and then be held to the description. Where that assumption holds, the machinery works well. Where it does not, the machinery does something worse than fail: it produces confident-looking documents that encode guesses as commitments, and then measures performance against the guesses.
Executives know this. The response, in most organisations, is not to redesign the machinery but to exempt certain work from it. Research is exempted. Early-stage ventures are exempted. Anything labelled innovation is exempted. The exemption is granted in good faith — it seems unreasonable to hold exploratory work to a baseline it cannot honour.
But the exemption is granted at the level of the whole discipline, when it should be granted only to the forecasting part of it. An organisation that exempts uncertain work from planning does not thereby free that work; it strips it of the one mechanism that would let leaders tell, early and cheaply, whether it is going anywhere.
Consider, hypothetically, a pharmaceutical group carrying a portfolio of discovery programs. No one can say which compounds will progress. That genuine unknowability is used, in many organisations, to justify light governance at the discovery stage. Yet discovery is exactly where the cost of a wrong belief compounds longest, and where the difference between a program that is failing and a program that is merely early is hardest to see without deliberate instrumentation.
What Leaders Commonly Misread
Three misreadings recur, and each has a distinct cost.
The first is treating a plan as a promise. When the plan is understood as a commitment to a specific future, every revision becomes an admission of failure. Teams then defend plans past the point of usefulness, because revising one carries reputational cost. The organisation loses the very adaptability the plan was supposed to enable, and it loses it through a governance convention rather than a technical constraint.
The second is confusing detail with rigour. A hundred-line schedule for work nobody understands is not more rigorous than a ten-line one; it is less honest, because it distributes false precision across more decimal places. Rigour under uncertainty means being explicit about what is known, what is assumed, and what is being watched — not being granular about what is guessed.
The third is deferring the decision about what constitutes success. This is the most expensive of the three. Organisations routinely commit funds to initiatives whose definition of a good outcome is left to be settled later, on the reasonable-sounding grounds that it will become clearer with time. It rarely does. What actually happens is that the definition is settled by whoever is most persuasive at the moment the question can no longer be avoided — usually well after the money is spent and the options are gone.
Reframing the Issue
The reframe is straightforward to state and difficult to live with: a plan under uncertainty is an instrument of detection, not an instrument of prediction.
Its value is not that it tells you what will happen. Its value is that it makes explicit what you currently believe, so that when reality diverges from that belief, the divergence is visible — early, and to everyone, rather than late, and only to the people closest to the work.
This changes what a good plan looks like. A predictive plan is judged by how closely outcomes match it. A detective plan is judged by how quickly it surfaces the fact that they do not. Those are different design criteria, and they produce different documents.
It also changes what planning is for commercially. Portny's observation is that the less certain you are, the more it matters that everyone involved shares an understanding of the current situation, the intended direction, and the reasoning behind both. That is not a project-administration point. It is a statement about organisational coordination under ambiguity — and coordination under ambiguity is one of the few capabilities that reliably separates enterprises that can act on early information from enterprises that cannot.
What the Front End Actually Buys
There is a hard commercial reason to do this work before commitment rather than after: optionality is cheapest at the front end and becomes progressively unaffordable.
Early in an initiative, changing direction costs a conversation. Later, it costs contracts, hiring decisions, integration work and credibility. The window in which choices remain genuinely open is short, and it closes whether or not the organisation used it. Deferring definition does not preserve flexibility; it spends flexibility without buying anything with it.
An infrastructure example makes the asymmetry concrete. Once a route is chosen, land is acquired and geotechnical work is commissioned, the practical option set has narrowed to variations on a theme. The decisions that determined whether the asset would be worth building were made months earlier, in a room where the cost of thinking hard was measured in weeks of professional time rather than years of capital.
The front-end questions worth answering before commitment are not complicated. Portny sets out seven, covering intended result, whose goals must be taken into account, the work required, who is responsible and whether they have the expertise, timing, additional resources, and what might go wrong. Restated for an enterprise investment, they compress to about five. What result would justify this investment, stated specifically enough that we could recognise it? Whose view of that result governs, and whose merely informs? What work does this require, who would do it, and do they have the capability? What else must we hold or acquire for this to be possible? What could reasonably turn out differently, and what would it cost us if it did?
None of these require knowing the future. All of them require deciding what we currently think — which is a different and far more achievable thing.
Decision Framework
A practical test for whether an initiative is ready to receive money. Each question has a threshold, and the thresholds are deliberately unforgiving.
| Test | The question | Threshold for funding |
|---|---|---|
| Recognisability | Could we recognise the intended result if it arrived? | Stated specifically enough that two independent executives would agree whether it had occurred |
| Authority | Who is permitted to change what this initiative is for? | The question has been asked and answered in writing |
| Belief | What must be true for this to work? | Assumptions written down and separated from facts |
| Detection | Has anyone established how we would learn we are wrong? | A detection plan exists and someone owns it |
| Reversibility | What does this decision foreclose? | Explicit list of options this commitment removes |
| Exit | What would make us stop? | Conditions defined before funding, not after |
An initiative that fails the Recognisability or Exit test should not receive full funding. It may well deserve a small, bounded allocation to resolve the ambiguity — but that is a different decision, made on different grounds, and it should be described as such rather than dressed as a delivery commitment.
From Strategy to Execution
Immediate. Take the three largest initiatives currently in flight whose outcomes are genuinely uncertain, and ask what would have to be observed for leadership to conclude they are not working. If nobody can answer within the meeting, the organisation is funding those initiatives blind, regardless of how much reporting they generate.
Medium term. Separate the two things a business case currently does. One is an investment argument — why this deserves capital against alternatives. The other is a delivery estimate — what it will take and when. Fusing them forces teams to fabricate precision in the second to win approval on the first. Splitting them lets an initiative be approved on the strength of its logic while being explicit that its estimates are provisional.
Long term. Build the organisational tolerance for revised plans that this all depends on. A governance culture that treats revision as failure will produce stable plans and unstable outcomes. The capability to be built is not better forecasting; it is faster, less defensive re-decision.
Signals to Monitor
- Plans that never change. In genuinely uncertain work, a stable plan is evidence of a plan nobody is using.
- Late-arriving bad news. If problems reach the executive committee already fully formed, the detection layer is not working.
- Assumption drift. Conditions that were assumed at approval and have since changed, without anyone revisiting the approval.
- Definition creep. The stated purpose of an initiative shifting gradually without a decision point — usually the first observable sign that no one holds authority over it.
- Estimate convergence. Teams producing suspiciously similar confidence levels across dissimilar work, which suggests estimates are being written to satisfy a governance format rather than to inform a decision.
Questions for the Leadership Team
- For our largest uncertain investment, what specific observation in the next ninety days would cause us to change course — and who is responsible for making that observation?
- Where have we exempted work from planning discipline, and did we intend to exempt it from forecasting or from thinking?
- When a team revises a plan here, is that treated as good governance or as a performance problem? What does the honest answer tell us?
- Which of our current commitments have foreclosed options we did not consciously decide to give up?
- What would we need to see to stop a major initiative — and has anyone written it down while stopping is still cheap?
Closing Perspective
The organisations that handle uncertainty well are not the ones with better forecasts. They are the ones that decided, before committing capital, what they believed and what would prove them wrong — and then built the discipline to notice.
That is not an argument for more planning. In many enterprises it is an argument for less: fewer documents, produced earlier, containing sharper claims. The test of a plan under uncertainty is not whether events matched it. It is how long it took the organisation to find out they had not.
Related article: Risk Is Not the Chance That Things Go Wrong
Related article: Measuring an Outcome You Cannot Predict
Related article: What Must Be True: The Assumptions Register as a Strategy Instrument
About the author
Kevin Jogin is Founder & Principal Advisor at EraNorth. Meet the Founder.