The replacement-cost fallacy
The standard build-versus-buy calculation compares the salary premium of an external hire against the training cost of an internal candidate, and usually concludes that buying is faster. This is the wrong frame, because it prices the wrong thing. What you are actually buying or building is not a person who can do a task; it is the tacit knowledge of how work gets done here—which decisions are reversible, who must be consulted, where the informal authority sits, what has already been tried and failed. That knowledge has no market price because it cannot be transferred in an offer letter. The genuine trade-off is between speed of capability and depth of context, and the two trade against each other almost mechanically: the faster the acquisition, the thinner the contextual roots.
This is why the senior external hire who looked decisive in the interview so often stalls in month four. The capability arrived intact; the context did not, and the organization rarely budgets for the second to catch up to the first.
What buying actually transfers, and what it cannot
Acquired talent transfers reliably along one axis and poorly along another. It is strong at importing capabilities the organization does not yet possess and cannot teach itself—a regulatory specialism, a platform discipline, a market it has never sold into. It is weak at anything that depends on accumulated relationships or institutional trust, because those are not attributes of the person but of the person's position inside a specific web. When you buy, you reset that web to zero and ask the new hire to rebuild it under deadline pressure. The decision rule that follows is sharper than the usual one:
- Buy when the gap is a genuinely novel capability with no internal seed to develop, when the time horizon is short and externally imposed, and when success depends on technical skill more than on internal coordination.
- Build when the capability already exists somewhere in adjacent form, when success turns on trust and cross-functional credibility, and when you can absorb the slower ramp in exchange for retention and contextual depth.
The failure mode is compounding, not one-off
The expensive error is not a single bad hire; it is a buying habit that quietly dismantles the internal development engine. Every senior role filled from outside is a promotion that did not happen inside, and ambitious internal people read that signal precisely. They conclude the ceiling is real and they leave—often the strongest ones first, because they have the most options. The organization then has fewer internal candidates next time, which makes the next role harder to fill internally, which justifies buying again. This is a self-reinforcing loop, and it runs in the direction of dependency: each external hire raises the marginal cost of the build option until building stops looking feasible at all. The talent market becomes a permanent supplier rather than an occasional one, with all the pricing power that implies.
The decision implication: manage the ratio, not the hire
The unit of analysis is wrong if it is the individual requisition. No single decision to buy is irrational; the damage lives in the aggregate. The discipline a People Lens imposes is to set and defend an explicit build-to-buy ratio at the level of the function and the senior cohort, and to treat each external senior hire as drawing down a budget that must be replenished by deliberate internal development. Two practical commitments follow. First, every buy decision should name the internal candidate it bypassed and state honestly why the build option was rejected—this surfaces the compounding cost while it is still cheap to see. Second, capabilities that are core and durable should bias hard toward build even when buying is faster, reserving acquisition for the peripheral, the novel, and the urgent. Speed is a real benefit, but it is borrowed against a stock of internal capability that no hiring budget on its own ever rebuilds.