What leaders usually conclude
The instinctive read is that this is simply what deep expertise looks like — a strength the firm should protect and deploy, not a problem to solve. That instinct is not wrong about the value. It is incomplete about the mechanism. A single point of expert dependency behaves differently depending on how much slack exists around it, and slack is exactly what has been consumed by the dependency itself.
What the evidence suggests
Organisation design research offers a useful frame here. Jay Galbraith’s work on organisations as information-processing systems, developed at a time when firms were grappling with rising task uncertainty and coordination demand, describes several structural responses available to a growing organisation — among them building self-contained units and creating lateral relations that don’t route every complex decision through one node. What professional services firms often do instead, especially in their growth years, is the opposite: they route more and more complexity toward the person best equipped to handle it, because in the moment that is the fastest, safest choice for the client in front of them. The cumulative effect of many individually reasonable choices is a single point of failure nobody deliberately built.
The self-reinforcing part of the pattern is what makes it distinct from ordinary specialisation. The busier this person becomes, the less capacity exists to reduce the dependency — because reducing it requires exactly the unstructured, unbillable time that scarcity has already eliminated. Teaching a successor is slow, uncertain, and invisible in the short term compared with taking the next client call personally. Every rational weekly decision points the same direction, and the direction is away from resolving the dependency.
This is also a specific instance of the cross-unit coordination failure Donald Sull, Rebecca Homkes and Charles Sull describe in their research on strategy execution across more than 250 companies. Their finding — that only 9% of managers believed they could reliably rely on colleagues in other functions — describes an organisation-wide pattern of coordination breaking down at boundaries between people and teams. A single-expert dependency is the same failure compressed into one person: rather than the firm failing to coordinate across several colleagues, it has simply never built the coordination in the first place, because one person’s availability has always been sufficient to paper over the gap. The underlying fragility is identical; it has just not yet been exposed, because it has never needed to be.
In practice, this often surfaces first not as a crisis but as a quiet, recurring scheduling problem: the same name appears on every complex proposal, every escalation, every pitch for a certain category of work, long before anyone frames it as organisational risk. Firms that do eventually name it explicitly tend to do so only after a near-miss — a period of illness, an unexpected departure elsewhere in the firm that suddenly reveals how much this person was quietly compensating for — rather than through any routine review of where the firm’s capability genuinely sits.
Competing explanations
A firm might instead attribute the constraint to insufficient hiring, arguing that a second senior specialist would simply relieve the load. This is sometimes true, but hiring alone does not transfer the specific judgement this person holds, which was built through years of exposure to unusual matters rather than through a role description. A firm might also attribute it to the individual’s own reluctance to delegate — and that can be a real, separate factor — but the structural version of the problem exists even with a willing teacher, because willingness does not create the hours that teaching requires.
There is also a client-side dimension worth naming separately. Clients themselves often reinforce the dependency, asking for this individual by name because past experience with her was excellent — a perfectly rational preference from the client’s point of view, and one that makes the firm’s internal task of broadening capability harder, not easier. Every satisfied client who asks for her specifically adds another data point confirming, to the firm, that routing complexity her way is the safe choice in the moment, even as it deepens the long-run constraint. This is not a reason to discourage the preference; it is a reason to notice that the market signal and the organisational risk point in opposite directions.
A further complication is succession timing. Firms often wait for an obvious moment — a planned departure, a sabbatical, a promotion — to finally invest in building a second capability, but by the time that moment arrives, the dependency has usually deepened further, not lessened, because every intervening year has added more client relationships and more matters routed through the same person. The best window to begin reducing a concentration risk of this kind is typically well before any visible trigger for doing so exists, which is precisely when it feels least urgent.
Where this shows up
The pattern described here is most recognisable in professional services firms, part of the wider Professional and Knowledge-Based grouping. The same organisational questions appear, under different names, in other sectors.
Questions worth considering
- What would it take to make this person less necessary — and does anyone in the firm currently have the time to do that work?
- Is the constraint being treated as a strength to protect, a risk to manage, or both at once?
- How many of the firm’s most valuable relationships currently run through a single individual’s calendar?
Your team is facing similar challenges?
Subscribe to The Riddle Ledger
Perspectives are sent as they are published, with no promotional email, and you can unsubscribe at any time.