Onboarding Metrics: 9 KPIs That Predict New Hire Ramp
Nine onboarding metrics worth tracking: lagging KPIs that prove onboarding worked, leading KPIs that flag a stalled ramp early, and how to benchmark each one.
Most companies measure onboarding by what got completed: forms signed, sessions attended, checklists closed. None of that says whether the new hire is ramping. Meanwhile, 60.8% of HR leaders say early turnover has increased in the past year, and only 36% describe the handoff from recruiting to the hiring manager as seamless. The onboarding metrics below are the ones that would have shown those problems coming, split into the numbers that show whether onboarding is working right now and the numbers that prove it worked.
What Are Onboarding Metrics?
Onboarding metrics are the measurements that tell you whether new hires are ramping, staying, and integrating, rather than whether onboarding tasks got done. The useful ones split into two kinds: leading metrics such as check-in completion and 1:1 frequency, which predict the outcome, and lagging metrics such as retention and time to productivity, which confirm it.
| # | Metric | Type | How to calculate |
|---|---|---|---|
| 1 | Time to first meaningful contribution | Leading | Days from start to the first owned deliverable shipped |
| 2 | 30-60-90 milestone completion | Leading | Milestones hit on time ÷ milestones set |
| 3 | Check-in completion rate | Leading | 30-, 60-, and 90-day check-ins held ÷ check-ins scheduled |
| 4 | New-hire sentiment | Leading | Average pulse score at each mark, and the trend between marks |
| 5 | Manager 1:1 frequency | Leading | Weekly 1:1s held in the first 90 days ÷ 12 |
| 6 | Network integration | Leading | Regular collaborators beyond the immediate team at days 30, 60, 90 |
| 7 | 90-day retention | Lagging | Hires still employed at day 90 ÷ hires who started, by cohort |
| 8 | First-year retention | Lagging | Same calculation at day 365 |
| 9 | Time to productivity | Lagging | Days from start to the role’s “fully ramped” threshold |
Lagging metrics are what leadership asks about. Leading metrics are what let you change the answer.
Leading Metrics: Signals While You Can Still Act
Leading onboarding metrics are observable inside the first 90 days and predict the lagging ones. Most are manager behaviors: whether the 30-60-90 plan is being followed, whether check-ins and 1:1s are happening, and whether the new hire has shipped real work. When these slip, retention follows a few months later.
The reason to track manager behavior directly is that it varies more than anything else in onboarding. In Enboarder’s 2025 survey, 44.8% of organizations provide only general 30-60-90 guidelines and leave execution to manager discretion, and 28.8% of HR leaders have seen a manager give a new hire no guidance at all.
Five leading metrics catch problems early:
- Time to first meaningful contribution. The day a new hire ships something with their name on it. For most knowledge-work roles this should land inside the first 30 days. If it has not happened by day 45, find out why. Either the manager is not assigning real work or the hire is struggling with the work they have, and both are cheaper to address now than at day 90.
- 30-60-90 milestone completion. Requires the plan to be written. If a manager cannot produce one, that is the finding.
- Check-in completion rate. Whether the 30-, 60-, and 90-day conversations happened. Our 30-60-90 day check-in questions cover what to ask in them. This metric only asks whether they took place.
- Manager 1:1 frequency. Weekly 1:1s from week two are the baseline for a new hire. Our manager check-in cadence guide covers why the cadence matters. For onboarding, the metric is simply how many of the first twelve weeks had one.
- New-hire sentiment. A three-question pulse at each mark. The absolute score matters less than the trend. A hire who drops between day 30 and day 60 is telling you something the completion metrics cannot.
The Metric Most Teams Miss: Network Integration
Network integration measures how quickly a new hire builds working relationships beyond their manager and immediate team. It predicts retention because connection is the second most common reason new hires leave, and it predicts productivity because most knowledge work depends on knowing who to ask. Almost nobody tracks it.
The evidence for it is consistent. BambooHR’s 2025 benchmark found that 31% of new hires said their onboarding lacked meaningful interactions, and that being welcomed by the team was the top driver of belonging, ahead of manager support. Enboarder’s survey puts lack of connection with the team or culture at 19.5% of early exits, second only to misaligned expectations.
The simple version: at days 30, 60, and 90, count the people the new hire works with in a typical week who are not their manager or on their immediate team. A number that is flat or shrinking is a stalled ramp, even when every checklist is green.
The better version uses the collaboration data your tools already have. Organizational network analysis reads signals like shared meetings, messages, code reviews, and documents to map who actually works with whom. Windmill’s Organizational Health report does this across the company, so HR can see whether a new hire’s network is widening or whether they are quietly isolated, without asking anyone to fill in a tracker.
Lagging Metrics: Proof That Onboarding Worked
Lagging onboarding metrics measure outcomes after the fact: whether the hire stayed and how long they took to reach full output. They are the numbers that justify the onboarding budget, and they are useless for rescuing any individual hire, because by the time they move, the decision has already been made.
90-day and first-year retention. Calculate both by cohort, not as a running average, so a bad quarter is visible instead of diluted. Then cut by role and by manager. Early attrition that clusters under one manager is an onboarding problem. Attrition spread evenly across managers usually points at hiring or compensation.
Time to productivity. AIHR defines it as the time it takes a new hire to reach the expected level of performance in their role. The definition only works if “expected level” is written down per role before the hire starts: for an engineer, shipping independently at the team’s review standard; for an account executive, carrying a full quota. Enboarder’s 2025 survey found HR leaders split almost evenly on how long full onboarding takes, 45% saying one to three months and 45% saying four to six, which suggests most companies are estimating rather than measuring.
How to Set Onboarding Benchmarks
Set onboarding benchmarks from your own history, not industry averages. Pull the last two years of hires, calculate each metric by role, and use the median as the baseline and the spread as the target to narrow. A ramp that is consistent across managers is worth more than one that is fast for some hires and slow for others.
Three cuts make the benchmarks useful:
- By role. A support hire and a staff engineer should never share a time-to-productivity target.
- By manager. The widest spread in onboarding outcomes is usually between managers, not between roles. This is the cut that tells HR where to coach.
- By cohort. Hires who started in the same month went through the same version of onboarding. When you change the process, cohorts are how you see whether the change worked.
Watch trends, not single numbers. A 90-day retention rate of 88% means nothing on its own. The same rate falling from 95% over three quarters means onboarding broke somewhere.
Common Onboarding Measurement Mistakes
The most common onboarding measurement mistake is counting completed onboarding tasks instead of the new hire’s outcomes. Forms signed and sessions attended prove the process ran. They say nothing about whether the hire is ramping, and they produce a green dashboard right up until the resignation.
Other patterns that show up repeatedly:
- One survey at day 90. By then, the hire who was struggling at day 30 has either recovered on their own or decided to leave. Ask at every mark.
- Averaging across roles. A blended time-to-productivity figure hides the one function where ramp takes twice as long.
- Filing early attrition under recruiting. If a hire leaves in the first 90 days, look at the onboarding data for that hire before concluding they were a bad fit.
- No named owner. HR owns the metrics, managers own the inputs, and someone has to reconcile the two every month. Our guide to the role of HR in the onboarding process covers that split.
Instrumenting Onboarding Metrics Without a Spreadsheet
Most leading onboarding metrics already exist in the tools a company uses every day: the calendar knows whether 1:1s happened, GitHub and Jira know when real work shipped, and Slack knows who is collaborating with whom. Instrumenting onboarding means reading those signals instead of asking managers to report them.
That is how Windmill approaches it. Pulse Surveys include a New Hire schedule that sends a check-in a set number of days after each start date, so sentiment at 30, 60, and 90 days collects itself. The 1:1 Meetings report lists employees who have not had a recent 1:1 with their manager and ranks managers by how many of their reports they actually meet with, which is manager 1:1 frequency without a tracker. Weekly recaps generated from GitHub, Jira, and 20+ other tools show the week a new hire’s first real work landed. Organizational Health shows whether their network is growing.
Onboarding metrics are only useful if someone sees them before the exit interview. Pick the nine above, or the five you can actually collect this quarter, and put them somewhere a stalled ramp is visible at day 30.
Frequently Asked Questions
What are the most important onboarding metrics to track?
Start with five: 90-day retention, time to productivity, check-in completion rate, new-hire sentiment at 30, 60, and 90 days, and how quickly a new hire builds working relationships beyond their immediate team. Retention and time to productivity tell you whether onboarding worked. The other three tell you whether it is working while there is still time to intervene.
How do you measure onboarding effectiveness?
Compare each new hire's ramp against a baseline for their role: how long it took to ship a first meaningful deliverable, whether they hit their 30-60-90 milestones, and whether they are still there at day 90 and day 365. Pair those outcomes with a short pulse at each 30-day mark so you can see the sentiment that precedes the outcome.
What is time to productivity in onboarding?
Time to productivity is the number of days from a new hire's start date to the point where they deliver at the level expected of a fully ramped person in that role. Because that point differs by role, define it per role in advance, and track time to first meaningful contribution as an earlier proxy.
What is a good 90-day retention rate for new hires?
There is no universal benchmark, because it varies by industry and role, but a 2025 Enboarder survey found one in five HR leaders report losing up to half their new hires within 90 days. Track your own rate by role, manager, and hiring cohort, and treat a downward trend as an onboarding problem before treating it as a hiring problem.
Should onboarding metrics be owned by HR or by managers?
HR owns the metrics and the reporting. Managers own the inputs, because most leading indicators, such as 1:1 frequency, milestone completion, and feedback given, are manager behaviors. The best setups make the manager-level numbers visible to HR so a stalled ramp is noticed at day 30 instead of at the exit interview.