Attrition is usually discussed as a recruiting number. The cost, however, does not sit in hiring – it sits on the ramp-up curve.
Attrition is almost always discussed as a recruiting figure: how many left, how many need replacing. That describes the effort, not the cost. The cost sits somewhere else – on the curve between the first day and full productivity.
The ramp-up curve is the real cost
Someone new handles fewer cases at first, escalates more often, and needs support from colleagues who then get less done themselves. That second effect is rarely counted, yet it is often larger than the first.
So the question is not what a hire costs, but how long it takes before someone works independently – and how many people reach that point at all.
People leave earlier than the statistics suggest
Annual attrition hides the fact that a large share of departures happens in the first few weeks. Someone who leaves after ten months has worked. Someone who leaves after five weeks has only cost.
The reasons in that early phase are remarkably consistent: it is unclear who to turn to when unsure; there is no feedback on whether you are doing it right; and there is a sense that asking questions counts as weakness.
What helps
A named point of contact for the first weeks does more than any general training course. What matters is that the role is named and scheduled – “just ask” is not an assignment of responsibility.
Early, reachable wins work just as well: a clearly bounded set of cases someone can handle confidently within days, rather than taking on the full range from the start.
And finally, feedback that does not wait for the end of probation. Six weeks of working without any assessment sends people in a direction that is laborious to correct later.
What helps little
Bonuses retain people who intended to stay anyway. They rarely change the mind of someone who concluded in week three that they are not up to the job.
Nor does a longer initial training course help if the handover into live work is left unprotected. The difficult moment is not the classroom; it is the first conversation without support.
Measure it better
More telling than annual attrition is the 90-day retention rate, split by starting month and team. That figure responds quickly to changes in onboarding and reveals differences between teams that an overall rate blurs.
Pair it with time-to-independence. If that time shortens while quality holds, onboarding is working – regardless of how many people were hired.
