Ask any leader when their new hires leave and they will say around the one-month mark. Ask when those hires decided to leave and most cannot answer. I can. It was the second week.
Week one is adrenaline
The first week takes care of itself. Everything is new, there is a class to sit in, the leader is paying attention, and the hire has not had a bad day yet. Nobody quits on adrenaline.
Week two is the real onboarding
In the second week the attention drops, the first bad shift happens, and the hire looks around to see whether anyone noticed. That look is the whole hire. If someone noticed, they stay. If nobody did, they start telling themselves the story that ends with a two-week notice in week five.
So week two is where I put the structure, not week one. A named person who checks in daily, in person or by phone. One debrief after the first bad shift, on the same day. A visible early win, however small, that the leader points at in front of the team. And a direct conversation on day ten: how is it going, what surprised you, what do you need.
The number to watch
Do not measure onboarding by week-one attendance. Everyone shows up in week one. Measure it by week-two activity. A hire whose activity holds steady or climbs across the second week almost always makes it to ninety days. A hire whose activity drops in week two has already decided, and the only question is whether you find out before they tell you.
Nobody quits a job in the second week. They quit a feeling: that the attention they got on day one was a sales pitch. Prove it was not.