Employee Onboarding Metrics: 10 Numbers That Show New Hires Are Ramping

Why Checklist Completion Is Not a Metric

An onboarding checklist measures the organisation's effort, not the new hire's progress. A person can complete every module and still spend their fourth week not knowing which tool to open for which task. The metrics that matter describe the employee's trajectory: how quickly they reach normal output, how well they have plugged into the team's tools and cadence, and whether their early patterns look like someone who will stay.

The ten below are chosen because each can be measured from data you already have or can collect passively, and each changes a decision: extend the ramp plan, pair with a mentor, fix a broken process, or intervene before a resignation. The 90-day onboarding framework shows how to schedule the interventions; this article is about the numbers that trigger them.

Ramp Speed: Metrics 1 to 4

1. Time to productivity

The number of weeks until the new hire's output reaches the team median. "Output" is role-specific: tickets closed, deals progressed, code merged, cases handled. Benchmark: eight to twelve weeks for individual contributors in knowledge roles, longer for senior and specialist hires. If you have no output measure for a role, defining one is the first job.

2. Time to first independent deliverable

The day the hire ships something without a reviewer rewriting it. This lands earlier than full productivity and is the first real signal that training took. Benchmark: two to four weeks. A hire past week six with no independent deliverable needs a conversation about blockers, not more training.

3. Active hours trajectory

Weekly active hours on work applications, plotted across the first twelve weeks. Healthy ramps climb steadily and level off near the team norm. Two shapes are warning signs: a flat line well below the team (the hire is waiting for work or access) and a line that overshoots and stays high (the hire is compensating with hours, a burnout precursor). Automatic time tracking gives this without anyone filling in a form.

4. Ratio of training time to doing time

Hours in training material and shadowing versus hours in real work, by week. Should cross over by week three or four. If training still dominates in week six, either the program is too long or the team has not handed over real tasks.

Integration: Metrics 5 to 7

5. Core tool adoption

Of the applications the role uses daily, how many has the hire used in the last five working days? List the five to eight tools that define the role, then track first use and regular use of each. A hire still not in the ticketing system or CRM by week two has an access problem or a training gap. App and website tracking answers this directly.

6. Focus time per day

Uninterrupted stretches of thirty minutes or more on work applications. New hires often have almost none in weeks one and two, which is normal, then it should rise toward the team level. If it never rises, the hire is being interrupted or is drowning in meetings; if it is far above the team level, they may be isolated. Compare with the meeting load for the same weeks.

7. Collaboration touchpoints

Count of distinct colleagues the hire interacts with through work tools each week: shared documents, ticket handoffs, code reviews, meetings. Healthy integration shows the count widening over the first month. A hire whose touchpoints are still only their manager and their buddy at week six is not yet part of the team.

Early Risk: Metrics 8 to 10

8. Attendance pattern stability

Variance in start time and daily active hours. New hires typically settle into a stable pattern within three weeks. Growing variance, particularly late starts and shortened days, is one of the earliest signals of disengagement. Attendance tracking surfaces this without anyone watching the clock.

9. After-hours activity

Work activity outside scheduled hours in the first ninety days. Some is normal in week one. Sustained after-hours work from a new hire usually means the workload was set for an experienced person, and it predicts both burnout and early exit. Treat it as a workload signal, not a commitment signal.

10. Thirty-day sentiment check

The one metric on this list that is not passive. A short survey at day thirty and day ninety, three questions, one being "do you expect to be here in a year". Combine with metrics eight and nine: a hire with rising attendance variance who also answers that question with hesitation is the one to sit down with this week. The retention guide covers what that conversation should be.

Putting the Metrics on One Dashboard

Track the ten per hire, by onboarding week, against the team median for the same week. Three views cover most needs: a per-hire ramp chart (metrics one to four), an integration panel (five to seven), and a risk flag that lights when two or more of eight to ten move the wrong way in the same fortnight.

Review the dashboard in the weekly one-to-one rather than in a separate meeting; the one-to-one guide has a five-minute agenda slot for it.

A Worked Example: One Hire, Twelve Weeks

A support engineer joined a fourteen-person team in March. The checklist was complete by day three. The metrics told a different story over the following weeks.

By week two, core tool adoption showed the ticketing system and the knowledge base in daily use, but the internal escalation tool had never been opened. Nobody had granted access. That was a ten-minute fix that would otherwise have surfaced as "slow to escalate" in a week-eight review. By week four, the training-to-doing ratio had crossed over on schedule and the first independent deliverable, a resolved customer case with no reviewer edits, landed in week three.

Week six is where the risk flag lit. Active hours climbed to eleven percent above the team norm and stayed there, after-hours activity appeared on four evenings in one week, and attendance variance widened as start times drifted later. Two of the three early-risk metrics moved in the same fortnight. The manager raised it in the next one-to-one, sharing the screen rather than reading the numbers out. The hire had been assigned the same queue volume as a five-year veteran and was working late to keep up.

Queue allocation was reset to a ramp schedule. By week nine, after-hours activity was gone, attendance had stabilised and focus time had risen to the team level. Time to productivity landed at week eleven, inside the benchmark. The ninety-day sentiment check came back positive. Without the metrics, the most likely version of that story ends with a resignation in month four, recorded as a hiring mistake.

Benchmarks by Role Type

The healthy-by-week figures above describe a typical individual contributor in a knowledge role. Adjust the ramp expectations by role type before comparing hires against them; a senior engineer and a sales development representative should not share a benchmark.

What the Numbers Are Not For

Onboarding metrics are for spotting where the organisation is failing the new hire: missing access, absent handover, unrealistic workload, an isolated seat. They are not a scorecard for the person. A hire who is slow to ramp is a signal to look at the ramp, and the honest finding is usually a process problem rather than a hiring mistake.

Tell new hires which of these numbers you look at and why, and show them their own. The trust guide explains why that disclosure changes how the data lands. A first-month experience in which someone is measured secretly is the fastest way to turn a promising hire into a ninety-day exit.

Frequently Asked Questions