You Can't Improve What You Don't Measure - How to Know If Your Onboarding Is Working

General
1

Here’s a uncomfortable question: How do you know your onboarding is effective?

Most organizations answer:

  • “People seem to figure it out”
  • “We haven’t had complaints”
  • “Our retention is okay”

That’s not measurement. That’s assumption.

The Metrics That Matter

Leading Indicators (Early Signals)

These tell you if onboarding is on track while it’s happening:

Metric Target Red Flag
Time to first commit Day 3-5 Day 8+
Time to first PR merged Week 1-2 Week 3+
Daily questions asked 5-10 (week 1) <3 or declining rapidly
Buddy meeting attendance 100% Missed sessions
Environment setup time <1 day >2 days

Coincident Indicators (Progress Markers)

These track trajectory through the onboarding period:

Metric Target Red Flag
30-day satisfaction score 8+/10 <6/10
Manager confidence score 7+/10 <5/10
First feature ownership Week 2-3 Week 5+
Code review participation Week 2 Not reviewing by week 4

Lagging Indicators (Outcomes)

These tell you if onboarding worked after it’s done:

Metric Target Industry Average
90-day retention 95%+ ~85%
Time to full productivity 3-4 months 6-9 months
1-year retention 90%+ 77%
New hire NPS 50+ Unknown (most don’t measure)

The Feedback Loops

1. Day 7 Check-in (New Hire → Manager)

Questions:

  • What’s been better than expected?
  • What’s been harder than expected?
  • What information did you need that you couldn’t find?
  • How supported do you feel?

2. Day 30 Survey (New Hire → Onboarding Team)

Questions (1-10 scale):

  • I felt prepared to do my job on day 1
  • I understood what was expected of me
  • My buddy was helpful
  • My manager was available and supportive
  • I would recommend this onboarding to future hires

3. Day 60 Assessment (Manager → Onboarding Team)

Questions:

  • Is this person on track for their role?
  • What would have made their ramp faster?
  • Did our onboarding prepare them adequately?

4. Day 90 Exit Review (If Someone Leaves Early)

Questions:

  • How did onboarding contribute to your decision to leave?
  • What could we have done differently?
  • At what point did you decide this wasn’t the right fit?

Making It Actionable

Data without action is just trivia. Here’s what we do:

Weekly: Review leading indicators for current new hires. Intervene if red flags appear.

Monthly: Review 30-day survey results. Identify patterns and quick fixes.

Quarterly: Analyze lagging indicators. Make structural changes.

Annually: Full onboarding program review. Major investments and redesigns.

You can’t improve what you don’t measure. Start measuring.

2

Time to first commit as a health signal is the most actionable metric I’ve found.

Why this metric works:

  1. It’s objective: Either they committed code or they didn’t. No interpretation needed.

  2. It’s early: You get the signal in days, not months. Enough time to intervene.

  3. It reflects systemic health: A delayed first commit usually means upstream problems.

What delayed first commit tells you:

If commit is delayed by… Then investigate…
Environment issues Golden path broken, documentation outdated
Access problems IT provisioning failed, pre-boarding incomplete
No appropriate task Task queue empty, manager unprepared
Mentor unavailable Buddy capacity issue, poor assignment
Overwhelming complexity Codebase needs better entry points

How we track it:

GitHub webhook fires when anyone commits. If the committer is on our “new hire” list and it’s their first commit:

  • Slack notification to #onboarding-wins channel
  • Timestamp recorded against their profile
  • Dashboard updated

If day 5 passes without a first commit:

  • Alert to manager + buddy
  • Escalation to onboarding lead
  • Required check-in within 24 hours

The data we’ve collected:

Over 2 years, 80+ new hires:

First Commit Timing 1-Year Retention Time to Full Productivity
Day 1-3 94% 2.8 months
Day 4-5 88% 3.2 months
Day 6-10 71% 4.5 months
Day 10+ 52% 6+ months

The correlation is strong. Engineers who commit early tend to stay and ramp fast. Those who don’t are often struggling in ways we can address - if we catch it early.

First commit isn’t the only metric. But it’s the first signal that something might be wrong.

3

The new hire survey questions that actually matter - and what to do with the answers.

Day 7 Survey (5 questions)

  1. “On a scale of 1-10, how prepared did you feel to contribute on day 1?”

    • <7: Investigate pre-boarding gaps
    • Pattern of low scores: Systemic issue

  2. “What information did you need in week 1 that was hard to find?”

    • Free text → feeds documentation improvements
    • Common answers become FAQ additions

  3. “How many hours did you spend with your buddy this week?”

    • <3 hours: Flag to buddy’s manager
    • Pattern: Buddy program capacity issue

  4. “Have you had a 1:1 with your manager?”

    • No: Immediate escalation
    • Every “no” is a manager accountability issue

  5. “One thing that would make next week better:”

    • Actionable immediate feedback
    • Manager receives and responds within 48 hours

Day 30 Survey (10 questions, scaled 1-10)

  1. I understand what success looks like in my role
  2. I have the tools and access I need to do my job
  3. My buddy has been helpful and available
  4. My manager has provided clear expectations
  5. I feel comfortable asking questions
  6. The team culture matches what I expected
  7. I understand how my work contributes to team goals
  8. I feel like a valued member of the team
  9. I would recommend joining this team to a friend
  10. Overall, how would you rate your onboarding experience?

What we do with scores:

Average Score Action
9-10 Celebrate, understand what worked
7-8 On track, minor improvements
5-6 Intervention needed, manager+HR meeting
<5 Critical, skip-level involvement, retention risk

The key insight:

Most companies survey and file. We survey and act. Every <7 score triggers a specific follow-up. Every common complaint becomes a process improvement.

The survey isn’t measurement for measurement’s sake. It’s an early warning system with built-in responses.

4

Connecting onboarding metrics to business outcomes is how you get sustained investment.

The metrics engineering tracks:

  • Time to first commit
  • Time to first feature
  • New hire satisfaction scores
  • 90-day retention

The metrics business cares about:

  • Cost per hire
  • Time to fill positions
  • Team velocity
  • Roadmap predictability
  • Revenue impact

Building the bridge:

1. Translate ramp time to velocity impact

If average ramp time is 6 months and you improve it to 4 months:

  • Each new hire contributes 2 additional months of productivity in year 1
  • For a $150K engineer, that’s ~$25K in recovered value per hire
  • For 40 hires/year, that’s $1M in velocity gained

2. Translate retention to recruiting cost

If onboarding improves 90-day retention from 85% to 95%:

  • 10% fewer early departures
  • Each prevented departure saves ~$90K (recruiting + ramp replacement)
  • For 40 hires/year, that’s 4 prevented departures = $360K saved

3. Translate predictability to planning accuracy

When new hires ramp faster and stay longer:

  • Q2 capacity planning is actually accurate
  • Feature commitments can be made confidently
  • Fewer “surprise” delays from ramp issues

This is harder to quantify but very real in executive discussions.

The dashboard I review monthly:

Metric Current Target Business Impact
Avg ramp time 4.2 mo 3.5 mo +$625K velocity
90-day retention 93% 95% +$180K savings
New hire NPS 62 70 Recruiting advantage
Time to first feature 3.1 wk 2.5 wk Faster roadmap execution

When I present this to the CEO, it’s not “onboarding is important.” It’s “here’s how onboarding affects revenue and costs.”

That’s the language that gets budget.