An organisation can look resilient right up until one person is unavailable.
A senior executive takes leave. A founder steps away for a week. A long-serving manager is sick. A key technical person resigns. Someone who normally answers the difficult questions is suddenly not there.
Then the organisation starts to behave differently.
Decisions wait. Meetings are postponed. People check whether something can be held until the person returns. Customers are told that somebody will come back to them. Work continues, but with more caution. Exceptions sit unresolved. Informal workarounds appear.
The absence can look like the problem.
Often it is not.
The absence is the test. The dependency was already there.
When a capable person is present every day, their judgement can conceal how much the organisation has come to rely on them. They answer quickly, connect information across functions, remember the history, resolve ambiguity and know which risks matter. Their capability keeps the system moving.
That makes the dependency difficult to see.
Performance can remain high while organisational independence quietly declines.
Capability can mask condition
A highly capable person can compensate for structural weakness for a long time.
They know who to call. They remember why a past decision was made. They recognise an issue before anyone else does. They can interpret incomplete information and make a call without needing the organisation to formalise every rule.
None of those things is inherently a problem.
The problem emerges when the organisation stops building enough alternative pathways because the person is consistently available to bridge the gaps.
Instead of clarifying authority, people ask them.
Instead of strengthening decision rules, people ask them.
Instead of distributing critical knowledge, people ask them.
Instead of resolving ambiguity closer to the work, people ask them.
Over time, the person’s helpfulness stops being an individual behaviour and becomes part of the operating model.
That is why key-person dependency is often underestimated. Leaders look for a single irreplaceable specialist or founder whose departure would create an obvious crisis. But dependency can be much more ordinary than that.
It can sit in the person who:
– approves exceptions that supposedly belong elsewhere; – carries the relationship history nobody has documented; – knows which stakeholders need to be consulted before a decision is made; – resolves cross-functional conflict because formal ownership is unclear; – translates between technical and commercial teams; – knows when a policy can be flexed and when it cannot; – becomes the default escalation point because people trust their judgement.
The person may not even appear indispensable on the organisation chart.
Operationally, they can still be load-bearing.
Absence changes behaviour before it stops work
A useful mistake to avoid is assuming dependency exists only when work completely stops.
Most organisations are more adaptive than that.
People find ways around the missing person. They defer some decisions. They become more conservative on others. They escalate to somebody more senior. They ask another experienced colleague. They reduce the scope of what they are willing to commit to.
So the first sign of dependency is often not failure.
It is behavioural change.
Ask what happens when the person is unavailable:
– Do decisions take longer? – Does authority move upward? – Do people become less willing to make exceptions? – Are meetings rescheduled because the person “needs to be there”? – Does customer or stakeholder communication become less certain? – Does somebody else inherit a sudden spike in escalations? – Do teams discover that important knowledge exists mainly in one person’s memory?
These are continuity signals, but they are also operating-model signals.
They show where responsibility may be distributed formally while judgement, confidence or authority remains concentrated in practice.
“We can cope without them” is not the same as resilience
Many leadership teams test dependency with a simple question:
Could we keep operating if this person left?
That is useful, but incomplete.
Most organisations can keep operating.
The better question is:
What would change in how the organisation operates?
Would decisions move more slowly?
Would more matters reach the executive team?
Would risk tolerance fall?
Would staff spend more time searching for context?
Would customers experience inconsistency?
Would other senior people absorb the load?
Would teams narrow what they are prepared to decide themselves?
A system can continue functioning while becoming materially more dependent, slower or more fragile.
That is why resilience should not be defined simply as continued activity.
A resilient organisation retains enough distributed judgement, knowledge and authority that the absence of one person does not materially distort how work gets done.
Dependency is rarely only about knowledge
Succession and continuity planning often focus first on knowledge transfer.
Document the process. Write the handover. Build the procedure. Cross-train somebody else.
Those actions matter, but they may not address the deeper dependency.
The organisation may know what the person does and still depend on them for judgement.
That distinction matters.
A procedure can explain the normal case. It may not tell somebody how to respond when two priorities conflict, when a customer request falls outside policy, when a stakeholder is politically sensitive, or when the consequences of a wrong decision are unclear.
Those situations are precisely where decision authority and organisational trust become visible.
If the backup person knows the information but still feels they need permission to act, the dependency has not really moved.
It has only been documented.
The real test is whether judgement can travel
Key-person dependency becomes less dangerous when capability can travel through the organisation rather than remaining attached to an individual.
That does not mean every person should be interchangeable. Nor does it mean removing expertise, hierarchy or senior judgement.
It means making deliberate choices about where different forms of dependence are acceptable and where they are not.
For a genuinely specialist technical decision, concentration may be rational.
For ordinary exceptions, recurring customer issues or routine cross-functional trade-offs, repeated dependence on one senior person may indicate something else: authority has not moved with responsibility, decision rules are weak, or trust has not developed far enough for judgement to sit closer to the work.
The objective is not to eliminate dependence entirely.
It is to prevent unexamined dependence from becoming the organisation’s default way of operating.
A practical absence test
One of the simplest ways to examine the issue is to choose a person who is widely relied on and ask five questions.
1. What routinely waits for this person? Look beyond tasks. Include decisions, approvals, introductions, interpretations and conflict resolution.
2. What moves upward when they are unavailable? This reveals whether their absence redistributes judgement or simply pushes it higher.
3. What do other people know, but still not feel authorised to decide? This separates a knowledge gap from an authority or trust gap.
4. Which relationships depend on the person rather than the role? Customers, stakeholders, partners and internal networks can create dependency even when formal processes look robust.
5. What capability would the organisation need to strengthen so their absence changed less? The answer may be documentation. But it may also be clearer authority, stronger decision rules, broader stakeholder relationships or more confidence in distributed judgement.
Those questions turn absence from a contingency-planning exercise into an organisational design diagnostic.
The warning sign is not that someone matters
Good organisations contain people who matter enormously.
Expertise matters. Trust matters. Relationships matter. Experience matters.
The warning sign is not that losing somebody would hurt.
The warning sign is that ordinary organisational functioning has quietly adapted around their constant availability.
When that happens, the person is no longer simply contributing capability.
They are carrying part of the system.
And the longer they carry it successfully, the easier it is for everyone else to mistake performance for resilience.
The question is therefore not only, “Who could we not afford to lose?”
It is also:
“Whose absence would change the way this organisation makes decisions?”
That is where dependency becomes visible.