Ask an owner about onboarding and you’ll hear about week one. Paperwork, the tour, the buddy, the lunch. It’s usually well run, because it’s the part everyone can see.
Then ask what happens in week seven. The answers get vague fast. They’re on the floor. Their manager’s got them. They’re picking it up.
That stretch — roughly day 31 to day 90 — is where the people you eventually lose decide to go. Not consciously, and not because anything went wrong. Because nothing did, and nothing else happened either.
What the drift looks like
In month one, a new person is allowed not to know things. Questions are expected. Someone checks in daily. By month two the check-ins stop, and the permission to be new quietly expires — while the actual competence is still weeks behind.
So the questions stop too. They start guessing, then covering. They learn the workaround instead of the process, because the workaround is what the person next to them does. And when it turns out they’ve been doing something wrong for six weeks, it lands as a correction rather than as training — which is the moment most people privately conclude they’re not going to make it here.
Most first-year turnover isn’t a hiring mistake. It’s a month-two support failure that took nine months to show up on a report.
Four things that close the gap
The number worth tracking
Most businesses measure 90-day turnover, which tells you about your hiring. Track twelve-month turnover for people who cleared 90 days — that tells you about your onboarding, and it’s the one that moves when you fix month two.
One healthcare client we worked with cut voluntary turnover by fifteen percent without touching recruiting, compensation, or benefits. The whole intervention was a scheduled conversation on day 45 and a written definition of independent. It cost a manager twenty minutes per hire.