Leading through uncertainty in a VUCA world

VUCA uncertainty is the letter that has generated more writing than any of the other three, and almost all of it points in the same direction. Gather better information, run better scenarios, reduce the unknown until a decision becomes safe. I have sat in enough Australian executive meetings to tell you what that advice produces in practice, which is a longer wait and a decision made anyway.

The trouble is that the advice answers the wrong question.

Uncertainty gets treated as a knowledge problem, so the response is always more knowledge. That works right up to the point where the remaining unknown is genuinely unknowable, and past that point every additional week of analysis buys you nothing except the illusion of diligence and a later start.

Leading through uncertainty turns on a variable most leadership teams never examine. What decides whether you survive being wrong is how cheaply you can undo what you committed to. I call the discipline built on that idea the reversibility ledger, and it is the spine of this article.

This piece continues my series on the frame. It follows VUCA and the management of volatility, and the reference page for the whole cluster is my pillar on VUCA leadership and where decisions are made.

The problem	Why the usual answer stalls	The question that unlocks it
Leading through uncertainty, definedLeading through uncertainty means directing an organisation when the nature, scale and duration of change cannot be known in advance. VUCA uncertainty delays decisions, widens the spread of legitimate opinions about the future, and makes any single point forecast unreliable.	More analysis has a floorGathering information reduces the part of the unknown that was knowable and leaves the rest untouched. Past that floor, further study buys delay rather than clarity, and planning systems that demand a single forecast push leaders to bury the remaining uncertainty inside the numbers.	Grade commitments by how cheaply they undoConfidence is a poor guide to whether a commitment is safe, while reversibility can be estimated before anything is signed. Sorting decisions by the cost of undoing them separates the ones that deserve months of analysis from the ones that should simply be tried this week.

What VUCA uncertainty actually means

VUCA uncertainty is the inability to know everything about a situation, combined with the difficulty of predicting the nature and effect of the changes under way. It sits at the intersection with volatility. It delays decision processes, widens disagreement about the future, and calls for deliberate risk management and hedging.

That is close to the working definition used in the strategic leadership material the acronym came from, and it repays a slow reading, because it contains two separate claims that get merged everywhere else.

The two halves of the definition

The first half concerns what you cannot know now. The second concerns what you cannot predict later. They are different failures and they need different responses, which is why treating uncertainty as one undifferentiated fog produces one undifferentiated answer, usually a request for more data.

The first half is partly fixable. Some of what you do not know is knowable, and effort will surface it. Market structure, competitor capacity, customer behaviour in your existing base and regulatory timetables all reward investigation.

The second half is not fixable at all. No amount of work tells you the size and duration of a change before it has run, and that is the part where the reversibility question earns its place.

The Strategic Leadership Primer treats hedging as part of the definition itself, which tells you the original authors already understood that the answer was structural rather than analytical.

How uncertainty differs from volatility

Volatility describes how fast and how far things move. Uncertainty describes how little you can say about what comes next. A market can be highly volatile and entirely predictable, like seasonal demand, and it can be perfectly stable while nobody knows what happens after a pending decision.

Keeping the two apart matters operationally, because they call for opposite investments. Volatility rewards speed of response and the willingness to retire what no longer holds, which I cover in the volatility article.

Uncertainty rewards optionality. You want several viable paths open at once, and you want each of them cheap enough to abandon without an argument.

Confuse the two and you get the classic error, which is an organisation that has invested heavily in moving fast down a single road it can no longer leave.

Residual uncertainty, the part that never goes away

Residual uncertainty is what remains after the best analysis you can afford, once every accessible figure has been gathered. It is a floor, and recognising that it exists is the difference between a leadership team that decides and one that keeps commissioning work.

I ask executive teams a single question to locate that floor. What would we need to know to be certain, and is that knowable by anyone at all today.

When the answer is that nobody on earth could know it yet, the analysis phase is over, whatever the calendar says. Everything after that moment is either a decision or a delay dressed as rigour.

Residual uncertainty is also where the opportunities live. Anything fully knowable is already known by your competitors, so the only ground where you can differentiate is the ground nobody can survey.

Why more analysis stops working

Analysis stops reducing uncertainty for two structural reasons. Subjects have become too large and too fast-moving for any individual to master, and the information available about them has become harder to verify. Neither problem is solved by assigning more analyst hours, because both concern the supply of reliable knowledge rather than the effort applied to it.

Australian organisations feel this acutely, and the national evidence supports the feeling rather than dismissing it as a mood.

It has become impossible to know a subject completely

The share of what a professional needs to hold in their own head, against what they must look up, has collapsed over a generation. Expertise has moved from possessing knowledge to knowing where knowledge lives and how quickly it can be assembled. This is the practical end of the expert as a self-contained authority.

The consequence for uncertainty is direct. A leadership team that expects an expert to remove the unknown is expecting something no individual can now supply in a field of any size.

The Productivity Commission’s work on why good practice fails to spread through the Australian economy touches the same nerve, finding that managerial capability varies but generally lags other countries, in a volume built around diffusion rather than invention.

Diffusion is the transmission of knowledge that already exists. When that transmission is slow, individual expertise stays trapped where it was created, and uncertainty stays high across the rest of the organisation for no good reason.

It has become harder to trust what you find

The second problem is verification. Information is abundant and its provenance is often untraceable, so the effort of establishing whether something is true now rivals the effort of finding it. That shifts uncertainty from a shortage into a filtering problem.

Be precise about the mechanism here, because it gets muddled with complexity. An overload of information belongs to complexity, which I cover separately.

What belongs to uncertainty is narrower and nastier. You cannot always tell whether what you are reading is accurate, which means the confidence attached to your inputs is itself unknown.

This is why I push clients towards primary sources with something close to obsession. A statistic from the agency that produced it carries a knowable error. The same statistic quoted third hand carries an unknowable one.

Watching the wrong competitor

A third failure sits underneath both. Organisations monitor the rivals they already recognise, which means their information gathering is precisely aimed at the part of the future least likely to hurt them. The threat that arrives usually comes from a category nobody had on the list.

The pattern repeats across decades and industries. A handset maker watches another handset maker while a computer company arrives. A hotel group watches another hotel group while a marketplace arrives. A navigation device maker watches another device maker while a search company gives the function away.

As Australian surfers put it, worrying about the shark you can see is a poor use of energy, because the one that gets you is the one you never spotted.

The practical fix is to assign someone the explicit job of watching the categories you do not consider competitors. It costs a few hours a month and it is the cheapest uncertainty reduction available to any organisation.

The two ways leadership teams get uncertainty wrong

Leadership teams fail on uncertainty in two opposite directions. They underestimate it and build strategies that neither defend against the threat nor capture the opportunity, or they overestimate it, conclude that the world is unknowable, and replace analysis with instinct. Both errors come from treating uncertainty as a single switch.

Understanding both is what makes the reversibility approach in the next chapter workable rather than glib.

Underestimating uncertainty

Underestimation produces strategies built on a single expected future. The organisation commits its capital, its headcount and its brand to one path, and the plan reads beautifully because it never has to accommodate an alternative. It fails the moment the world picks a different branch.

The business history of dismissed technologies is a long one, and every entry follows the same shape. A senior figure declares that a market will never exist, and the declaration is defensible on the day it is made and wrong within a decade.

What is interesting is that these people were rarely stupid or uninformed. They were reasoning correctly from the demand that existed rather than from the demand a lower price would create.

So the tell is not arrogance. It is a forecast that extends current behaviour forward without asking what would have to change for the behaviour to change.

Overestimating uncertainty

Overestimation looks like the opposite and does comparable damage. A leadership team decides the environment is unknowable, abandons the analytical rigour of its planning, and starts making strategic bets on instinct. The write-offs that follow are usually larger than the ones caused by excessive caution.

I see this most often after a shock. An organisation that was blindsided once concludes that forecasting is worthless, and swings from over-planning to no planning inside a single quarter.

Managers almost never know nothing. Even in a genuinely opaque situation, most teams can identify a range of plausible outcomes or a discrete set of scenarios, and that alone is enormously powerful.

The right question is therefore never whether the future is knowable. It is which level of uncertainty you are actually facing, because the level determines both the strategy and the process you should use to build it.

The binary trap inside your planning cycle

Conventional planning forces a binary view by demanding one number. A budget process that requires a single point forecast obliges leaders to bury the underlying uncertainty inside their cash flows, which systematically pushes the organisation to understate its own exposure in order to make the strategy defensible.

Watch what this does to behaviour rather than to numbers. Someone presenting a range gets asked for the number. Someone presenting a number gets approved.

So the process trains your most careful thinkers to present false precision, and it does so without anyone intending it.

The remedy is small and structural. Require every forecast above a set value to arrive with a stated range, the assumption that would break it, and the date by which that assumption can be tested.

The reversibility ledger, a better question than how sure are we

The reversibility ledger grades every commitment by what it would cost to undo, rather than by how confident anyone feels about it. Confidence is a state of mind and it cannot be audited. Reversibility is a property of the commitment itself, it can be estimated before anything is signed, and it tells you how much analysis the decision actually deserves.

This is the practical core of the article, and it is the shift I ask leadership teams to make when uncertainty has them stuck.

Why reversibility beats confidence as a decision test

Confidence and correctness are only loosely related, and the relationship weakens exactly when uncertainty rises. Reversibility behaves differently. A commitment that can be unwound in a fortnight for a known cost is safe whether or not it turns out to be right, which removes the need to resolve the argument before acting.

Notice what this does to a meeting. A confidence argument has no natural end, because two capable people can hold different readings of the same evidence indefinitely.

A reversibility question closes in about four minutes. What does it cost to stop, who can stop it, and how quickly would we know we should.

I have watched decisions that had circled for three months settle in a single session once the room was asked to price the exit rather than defend the entry.

Sorting your commitments into three grades

Three grades cover almost everything. Cheap to reverse, expensive to reverse, and effectively permanent. Each grade earns a different amount of analysis, a different approval level and a different review rhythm, and getting a decision into the wrong grade is the most common source of both paralysis and disaster.

1. Grade one, cheap to reverse. A pilot, a trial supplier, a new channel, a changed process in one team. Decide inside the team, this week, with a stated review date.
2. Grade two, expensive to reverse. A system, a partnership, a physical site, a restructure. Analyse properly, and buy an option first wherever the market allows one.
3. Grade three, effectively permanent. An acquisition, a long lease, a public commitment, an exit from a category. Take the full analysis, take the delay, and take it to the top of the organisation.

Two rules make the grading honest. Grade the decision on the cost of undoing it rather than on the size of the initial spend, because a cheap purchase that locks your data into a format you cannot leave is a grade two dressed as a grade one.

And grade before the debate rather than during it. A team that grades after arguing will grade to justify the position it has already taken.

Turning assumptions into hypotheses

An assumption is something you believe and never test. A hypothesis is a provisional conclusion drawn from limited evidence, held until something better arrives, with a stated way of checking it. Converting the first into the second is the daily practice that makes a reversibility ledger work.

Two questions do most of the work, and I ask them constantly in my own teams. What is the assumption here, and how will we know whether it is true.

The relief that follows is the part people do not expect. Framing a belief as a hypothesis removes the obligation to already know something that nobody can yet know, which lowers the temperature of the conversation considerably.

It also changes what a wrong answer means. A failed hypothesis is information that arrived on schedule, and a failed assumption is an embarrassment, which is why organisations that trade in assumptions learn so slowly.

Many small bets, each with a kill threshold

Hedging under VUCA uncertainty means holding several small positions rather than one large one, and pre-committing to the condition under which each gets stopped. The pre-commitment matters more than the size, because the failure mode of small bets is not losing them. It is refusing to close them.

The bias at work is the sunk cost fallacy, the tendency to continue investing because of what has already been spent. If you have ever sat through the second half of a bad film because the ticket was expensive, you have run the experiment on yourself.

The defence is to write the kill threshold before the bet starts, in the same document that authorises it. A threshold agreed in advance survives the emotional attachment that develops afterwards.

And shorten your windows. Plan delivery in shorter increments with more feedback loops, and build contracts that can be re-examined at agreed intervals rather than contracts that attempt to anticipate every contingency and take nine months to sign.

What VUCA uncertainty does inside the organisation

Uncertainty lands on six fronts, competition, customer demand, talent, technology, stakeholders and the economics of the model. Each front generates a different kind of unknown, and treating them as one general anxiety is why so many responses feel busy and change nothing measurable.

Take them in turn, because the useful move is to identify which front is actually generating your uncertainty.

Competition and customer demand

Competitive uncertainty comes from not knowing which players will act, when and how. Demand uncertainty comes from customers who do not always know what they want until they see it. The second is far larger than the first and receives a fraction of the attention.

Small operations can now threaten established ones, and even a startup can find itself copied before it has finished launching. Watching named rivals gives you a comfortable picture of a threat landscape that has stopped being organised by name.

Demand uncertainty scales with novelty. When I began offering recruitment based on shared values rather than on prior experience, nobody anywhere was selling that service, so no research existed and the uncertainty was total.

Contrast that with selling a replacement vehicle to an existing competition client who already wanted one. Same business owner, same year, radically different uncertainty, and only one of the two rewarded analysis.

Talent, technology and the economics of the model

Skills uncertainty is the sharpest of the three, because it is personal. Ask yourself which of your current skills will become obsolete first and when, and notice that you cannot answer. Neither can your organisation, which is why workforce planning built on named future roles keeps failing.

The scale is documented rather than speculative. Employers surveyed for the World Economic Forum expect 39 per cent of core skills to change by 2030, with curiosity and lifelong learning among the fastest rising.

Technological uncertainty comes from not knowing which technologies will emerge or combine. Several clean energy technologies compete to power vehicles and cities, and several medical approaches compete to treat the same conditions, with no way to pick the winner in advance.

Economic uncertainty then decides whether a model lives. My recruitment business grew because uncertainty ran in its favour, and my competition vehicle business closed because uncertainty ran against it, and I made comparable decisions in both.

Uncertainty, your people, and the duty that now attaches in Australia

Sustained uncertainty is a workplace hazard with a regulatory framework attached in Australia, rather than a mood to be managed with encouragement. Job insecurity, low job control and unclear expectations sit among the psychosocial hazards that employers must identify and control under work health and safety law. That changes what a leader is permitted to leave unaddressed.

This is the point where the Australian version of this topic diverges sharply from the international literature, and it deserves its own treatment.

Low job control is a regulated hazard in its own right

Safe Work Australia’s model code of practice on managing psychosocial hazards names low job control among the hazards a business must address, alongside high job demands, poor support and low role clarity. Low job control means having little say over how, when or where the work gets done.

Read that against the argument running through this whole series. The decision rights question stopped being a philosophical preference about empowerment the moment a regulator listed its absence as a hazard.

The code is explicit that businesses must eliminate or minimise psychosocial risks so far as is reasonably practicable, following the same risk management process used for any other hazard.

It also notes that these hazards interact. High demands alone may be manageable, and high demands combined with low control and weak support produce a risk considerably larger than any single factor, which is precisely the combination that prolonged uncertainty creates.

What leading through uncertainty owes your team

Three things carry most of the weight. Say what you know and what you do not, name the date by which more will be known, and protect the areas where the team still has genuine control. None of that requires you to have answers you do not have.

The instinct in senior teams is to withhold until certainty arrives, on the reasoning that half a story unsettles people. In my experience it does the opposite, because a team without information does not wait patiently. It fills the space with the worst available reading.

The naming of a review date matters more than the content. A person can tolerate a great deal of unknown when they know when the next instalment arrives.

And expand control wherever you can, because it is the lever most directly in your hands. Someone who cannot influence the outcome can usually still be given authority over the method.

What should never be outsourced to a resilience program

Resilience training helps people recover from load. It does not reduce load, and Australian regulators have been clear that training cannot be the dominant control where the underlying work design is the problem. Offering a course to a team whose real issue is structural reads as an instruction to cope better.

I am not against these programs, and I recommend them regularly. The sequence is what matters, since a resilience offer that arrives before any structural change signals that no structural change is coming.

So run the order the other way. Reduce what you can reduce, widen the control people hold over their own work, and then offer support for what genuinely remains.

That sequence also happens to be the one the risk management process in the code already requires, which makes it easier to fund than most leadership initiatives.

Scenario planning, small bets or hedging, which approach fits your decisions

Four approaches to VUCA uncertainty get sold to leadership teams, and each one suits a different level of residual uncertainty. Scenario planning suits a small set of identifiable futures, small bets suit a wide range of possibles, hedging suits exposure you can offset, and more analysis suits the case where the unknown is genuinely knowable.

Match the approach to the level of uncertainty you actually face rather than to the one your planning calendar assumes.

Comparing the four approaches on explicit criteria

Start from the middle column and find the situation that matches yours, then read across. The last column is the one that catches people out, because every approach here fails in a specific and predictable way, and choosing well means choosing the failure you can live with.

Approach	What it does	Situation it fits	Cost and maturity required	How it fails
More analysis	Converts knowable unknowns into known facts	The information exists somewhere and nobody has gathered it	Low, mostly analyst time and access	Continues past the point where the remaining unknown is unknowable
Scenario planning	Builds distinct futures and the triggers that signal each one	A small number of identifiable alternative outcomes	Medium, needs facilitation and executive attention	Produces documents nobody revisits once the trigger fires
Small bets with kill thresholds	Buys information by acting, at a survivable scale	A wide range of possibles with no obvious scenarios	Medium, needs delegated authority and tolerance for closure	Bets get started and never stopped, so the portfolio silts up
Hedging	Offsets an exposure by taking a counterbalancing position	Exposure that can be priced and matched, such as currency or supply	High, needs genuine financial or commercial expertise	Costs real money every year to protect against something that may not happen

One caution about the first row. Analysis is the only approach on this list that feels productive while achieving nothing, which is exactly why it is the default in organisations where being wrong carries a heavier penalty than being slow.

The signals that tell you which to fund

Three signals settle it quickly. If someone in the market already knows the answer, fund analysis. If you can name three or four distinct futures, fund scenario planning. And if the honest answer is that the range of outcomes cannot be listed at all, fund small bets, because acting is now the only way to generate information.

A fourth signal cuts across all three, and it is the one I watch for first. Find out who is allowed to close an initiative without escalating the decision.

When the answer is nobody, small bets will not work for you yet, whatever the situation calls for. A portfolio approach without the authority to close things becomes an expensive collection of orphans within about eighteen months.

Fix the closing capability first. It is unglamorous, it is cheap, and it determines whether every other approach on that table works.

How I can help you manage VUCA uncertainty

I work with executive teams, boards and manager communities on deciding under uncertainty, in English and in French. The formats below cover most situations, and each one is built to end with commitments graded, thresholds written and owners named, rather than with a shared feeling that everyone is now more comfortable with ambiguity.

Keynotes on deciding without full information

I speak on VUCA uncertainty at conferences, leadership offsites and industry summits, using the reversibility ledger as the through line because it gives an audience something to do on Monday. The session works best when it closes with each table grading one real decision they are currently sitting on.

Formats run from a short keynote to a half day with facilitated work between segments, and booking details sit on my page on managing in a VUCA world.

Workshops on experimentation and small bets

Over a half day or two days we build your reversibility ledger, set kill thresholds on the bets you already have running, and design the smallest testable version of the commitment you have been circling. Teams leave with one experiment live and a closing date in the calendar.

The curiosity work usually runs alongside it, since seeing early is what makes a small bet timely rather than reactive, and that thread continues in my work on curiosity in management.

Diagnostics and support for manager communities

For organisations inside a prolonged unknown, I run a short diagnostic on where uncertainty is being absorbed by individuals rather than by the system, and where job control has quietly narrowed. The output is a written picture with an owner against each item.

I then work with the manager community itself, because the people holding the line during a long unknown are the people least likely to ask for help. Building cross-functional range helps here, which is the argument behind the T-shaped expert generalist.

Conclusion: leading through uncertainty rewards cheap exits

Everything in this article comes down to one substitution. Stop asking how confident the room is, and start asking what it would cost to undo what the room is about to commit to.

That substitution does three things at once. It ends debates that could otherwise run for months, it routes each decision to the right level of analysis, and it makes being wrong survivable, which is the only condition under which an organisation will genuinely try anything.

It also protects the people doing the work, and in Australia that has stopped being optional. Prolonged uncertainty combined with narrow job control is a documented hazard, and widening what a team may decide is both the fastest response to the unknown and the clearest way to meet the duty.

So take one decision you have been circling. Grade it, price the exit, write the threshold that would stop it, and run the smallest version this month.

You will learn more in six weeks than in six months of analysis, and you will have found the only durable answer available when leading through uncertainty, which is an organisation that can afford to be wrong.

Frequently asked questions about leading through uncertainty