A company will run a six-week search, involve eight people in interviews, negotiate hard, and pay a premium to land someone — and then invest almost nothing in the first ninety days that determine whether any of it works.
I've never understood the math on this. The expensive part is already sunk. Onboarding is the cheapest point of leverage in the entire employment relationship, and it's the one most consistently left to improvisation.
What actually gets decided in ninety days
Three things get settled early and are very hard to change later.
Whether they can be effective here. Not whether they're capable — you established that in hiring. Whether they can convert capability into results in this environment, with its particular politics, systems, and unwritten rules. Someone brilliant who can't figure out how things get done will underperform someone average who can.
Who they trust and who trusts them. The relationships formed in the first few months tend to persist. A new hire who lands in a corner of the org with a narrow set of contacts will still be operating from that map two years later.
Whether the job matches what they were sold. Every hiring process oversells slightly. The gap between the pitch and the reality gets discovered early, and how it's handled determines whether it becomes a shrug or a slow-burning resentment.
None of those are addressed by a laptop, a login, and a benefits enrollment link.
The three failure modes
Sink or swim. The company hires a senior person specifically because they're senior, then provides nothing on the theory that a senior person shouldn't need help. But seniority transfers judgment, not context — and context is exactly what's missing. This is why experienced hires sometimes fail spectacularly at companies where less experienced people thrive.
Death by orientation. The opposite: three weeks of scheduled sessions, systems training, and meet-and-greets, with no real work. It looks thorough and it's disorienting in a different way. People form their sense of competence by doing things, and a month of being talked at leaves someone feeling less capable than when they arrived.
Hired for one job, handed another. The role shifts between offer and start — priorities changed, someone left, a fire started. Nobody names the change explicitly, they just start assigning different work. The new hire is now quietly doing a job they didn't accept, and they'll figure that out around month five.
What works
Give them something real to finish in the first two weeks. Small, self-contained, genuinely useful. Not a training exercise — actual work that ships. Nothing establishes footing faster than having contributed something. It also surfaces friction in your systems immediately, while they still have fresh eyes.
Name the unwritten rules out loud. Every company has them: who actually decides, which meetings matter, what "urgent" means here, whose objection will sink a project. Insiders don't perceive these as rules — they're just how things are. Writing down even ten of them and handing them over on day one is one of the highest-return things I know, and it costs an afternoon.
Assign a peer, not a manager. People will ask a peer the questions they're embarrassed to ask a boss, and those are the questions that matter. Make it someone's explicit job for ninety days.
Introduce them deliberately across the org, not just within their team. Ten scheduled thirty-minute conversations with people they'll need to work with, with a stated purpose. This is a common shortcut to skip and it shapes their effectiveness for years.
Ask at thirty days what's different from what they expected. This is the single most valuable question in onboarding and it has a short window. At thirty days they can still see your company clearly. By ninety they've acclimated, and the observations are gone forever. Ask it, write the answers down, and act on the ones that are true.
The retention connection
People rarely leave because of one thing. They leave because of an accumulation, and the first entries in that ledger are written early.
The employee who couldn't figure out how to be effective in month two and never fully recovered. The one who realized in month four that the role was different from the pitch and no one acknowledged it. The one who never built relationships outside their immediate team and so never felt like they belonged.
None of those show up in an exit interview as "bad onboarding." They show up as "wasn't the right fit" or "found a better opportunity." The cause is a year upstream of the departure.
The version that costs almost nothing
If you do nothing else, do these four:
- A real, finishable piece of work in the first two weeks.
- A written list of the unwritten rules.
- A named peer whose job is to answer dumb questions for ninety days.
- The thirty-day question, asked seriously, with the answers acted on.
That's not a program. It's about a day of setup and a genuine commitment to two conversations.
Against the cost of a failed senior hire — the search, the salary, the lost year, the disruption to the team, and doing it all again — it's the best return available anywhere in the people budget.