Succession is usually treated as a search problem to be solved when a departure is announced. The work that determines the outcome happens years earlier, in how decisions were distributed.
Readiness is built through decisions, not titles
A potential successor becomes credible by having made judgements of comparable weight and lived with the consequences.
Where senior decisions have all been made centrally, the layer below has managed execution rather than exercised judgement, however capable they are individually.
Promoting from that layer produces a first year spent learning what the previous leader already knew, at a moment when the organisation can least afford it.
The strongest leaders create the weakest benches
A highly effective senior leader resolves problems quickly and personally, which is efficient and removes exactly the situations their team would have learned from.
Their department performs well while they are present and struggles immediately when they are not, and the two facts have the same cause.
This is why succession risk is often highest around the people whose performance looks best.
An external hire resets the institutional memory
Bringing in an outsider imports experience the organisation lacks and discards the accumulated knowledge of why things are arranged as they are.
The new leader spends the first period discovering constraints that were obvious internally, and some of the changes they make will rediscover why the previous arrangement existed.
That cost can be worth paying where a genuine change of direction is wanted, and it is a poor trade where continuity was the objective.
Naming a successor changes the people not named
Identifying one candidate publicly tends to prompt the others to consider their own prospects, which can cost several senior people rather than one.
Naming nobody preserves ambiguity and leaves each candidate uncertain whether to stay, producing a slower version of the same attrition.
Organisations generally handle this by developing several people with genuinely different roles ahead of them, so the outcome is not framed as a single contest.
Unplanned departures are the case that matters
Planned retirements allow a handover; illness, resignation and recruitment by a competitor do not, and they are considerably more common.
The practical test is what would happen if a senior person did not return next week, and the answer is usually known without any analysis being required.
Where the answer is uncomfortable, the remedy is distributing decisions now, since that is the only part of succession that cannot be arranged at short notice.