Onboarding

Onboarding is hired to get a new hire productive and keep them here long enough to become expensive-to-replace. Companies that treat it as an HR formality misread the stakes. The dollars at play: first-year voluntary attrition in the US runs 20–35% depending on industry, and every one of those replacements costs half to two-thirds of annual salary to rehire and ramp. A program that reduces first-year attrition by even five points pays for itself ten times over in any company larger than 500 people. That math is the argument every onboarding leader needs to have with finance, and most don't make loudly enough.

Take a 700-person professional services firm hiring 120 people a year. An ineffective onboarding: the new analyst shows up, picks up a laptop, sits through a 4-hour Zoom, and is staffed on a client engagement the following Monday with no shadow time. By day 45, the new hire knows less about the firm than a well-read candidate did in the interview, and the first 1:1 with their manager happens in week four. An effective onboarding: preboard week with the firm's history and three past case studies; week one focused on craft (the firm's method), team (shadow a partner on a client call), and systems (time tracking, IT, calendar norms); weekly 1:1s for the first 90 days; a milestone review at day 90 against a plan set in week two. Same firm, same hire, two different first-year outcomes.

Most onboarding programs are really week-one programs. Days 1–5 are well-planned; days 30, 60, and 90 are nobody's responsibility. By day 30, the manager is back to steady-state work, the HR team has moved on to the next cohort, and the new hire is navigating by email threads. The quiet failure mode is week 4, not week 1: the new hire feels forgotten, the manager feels overbooked, and the first sign anyone notices is a resignation at month five. Programs that work move budget and attention to days 30–90 specifically because everyone else under-invests there.

In healthcare, onboarding is heavily regulated and uneven by role. An RN joining a hospital system has a credentialing workflow (licenses, certifications, background), a clinical orientation (unit-specific protocols, EMR training), and a social-integration layer (preceptor, shift assignments). The regulatory bit is almost always covered. The social-integration bit is uneven — which is where first-year RN attrition lives. Hospitals running 22% first-year RN attrition and hospitals running 10% have essentially the same regulatory training. They differ on the preceptor match, the first-shift pairing, and whether the new nurse has someone to text at 2 AM on night five.

By 2027, onboarding is partially automated, partially personalized, and entirely measured. AI generates a first-90-day plan based on role, background, and the team's current project load. The manager gets a weekly nudge with coaching prompts. The new hire's onboarding dashboard isn't a static checklist — it's a dynamic journey with inferred gaps ("you haven't met anyone from Product yet — here are three people to connect with"). The differentiator in 2027 onboarding isn't the day-one checklist; it's the quality of the coaching layer that runs for the manager across days 14–90.

• Day-one compression. Trying to fit all of onboarding into week one makes week one overwhelming and weeks 2–13 empty.
• No manager involvement. The manager is the program. Onboarding without the manager is orientation.
• Checkbox completion as success. A new hire who's completed 14 of 14 modules and doesn't know their team's actual priorities is not onboarded.
• No day-60 or day-90 checkpoints. The first 30 days get attention. The next 60 are where attrition risk actually lives.
• One program for every role. Sales onboarding, engineer onboarding, and nurse onboarding share almost no content. A shared template erodes specificity for all three.