On-Call Ramp-Up for New Engineers
Engineers don't fully on-call from day one. Ramp.
The problem
New engineers added to on-call rotation cold are dangerous. They miss context; they call senior staff at 3am for routine fixes.
Senior burnout from carrying junior shifts is a quiet drain. Multiply by team size; the cost is large.
Ramp-up is a workflow, not an ad-hoc thing. Build the path explicitly.
The ramp
Week 1: shadow shifts. New engineer attends incidents but does not act. Reads runbooks; asks questions in retrospect.
Week 2-3: secondary on-call. Pages route through primary first; new engineer is backup. Real exposure, low risk.
Week 4+: primary on-call with explicit escalation rules. Page senior on the third unrunbookable issue, not the first.
What to teach
Service architecture: top 5 services, top 5 dependencies, top 3 third-party integrations.
Incident process: ack flow, comms templates, decision authority during a crisis.
Runbook navigation: how to find a runbook, how to update a runbook, how to flag a stale one.
Checkpoints
End of week 2: simulated incident drill. Senior runs a fake outage; new engineer leads response.
End of week 4: review with manager. Confidence rating, gaps identified, decision to graduate or extend.
Quarterly: refresher drill. Skills decay; periodic practice maintains them.
Apply to your team
Document your ramp-up plan in a single page. New engineers should be able to read it on day 1.
Run one simulated incident per quarter for the whole rotation. Cheap practice; high payoff.
Track time-to-first-solo-shift per new hire. Aim for 4 weeks for senior hires, 8 weeks for junior.