Companies treat hiring as the expensive part and onboarding as an administrative afterthought. The recruiting spend, the interview loop, the compensation negotiation — all of it is scrutinised, budgeted and measured. Then the person arrives, receives a laptop, a set of logins and a calendar of introductions, and is left to work out the rest.
The departures that follow eighteen months later get recorded as bad hires. Some of them are. A great many were competent people who never got established, and the point at which that was decided was much earlier than anybody looks.
What a new person actually needs is a map of who decides
The formal onboarding usually covers policy, systems and org structure, all of which is available in writing and none of which is what determines effectiveness. What a new hire needs is the informal layer: who actually makes a decision as opposed to who is listed as owning it, which relationships have history, what has already been tried and failed, and where the real constraints sit.
That knowledge transfers through people, not documents, and it transfers fastest when somebody deliberately makes the introductions. Where nobody does, a capable person spends six months reconstructing it by trial and error, and every error costs credibility they had not yet banked.
The manager is the variable that matters most, and this is where the failure is usually structural rather than personal. Onboarding well takes several hours a week for a couple of months — deliberate context-setting, early feedback, brokered introductions — and a manager carrying too many direct reports does not have those hours. It is a predictable consequence of stretching the span of control past what anyone can hold, and it lands on exactly the middle tier that was cut and is being rebuilt.
Remote and hybrid hiring sharpened all of this, because the incidental transfer that used to happen by proximity no longer happens at all. Nothing about being in a building was magic; what it did was generate unplanned contact, and organisations that removed it without replacing it deliberately have a new-joiner problem they usually attribute to something else.
There is a measurement point worth making. Companies track time-to-productivity, which is a manager's estimate and therefore mostly noise. A better question, asked at ninety days, is whether the person can name the five people they need in order to do the job, and whether they have spoken to all five. That is answerable, it is comparable across hires, and it predicts what the estimate does not.
The internal comparison makes the case on its own. An internal move outperforms an external hire at the same level, consistently, and the reason is not that internal candidates are better — it is that they arrive with the map already in their heads. That is most of why internal mobility keeps turning out to be the cheapest recruiting channel nobody uses, and it is the same asymmetry that makes an external hire's first ninety days worth designing rather than scheduling.
