On-Call Management for Engineering Managers

The Manager’s Role in On-Call Success

Engineering managers determine whether on-call programs sustain team health or accelerate burnout. While individual engineers respond to alerts and resolve incidents, managers own the strategic decisions: rotation design, team sizing, tool selection, and organizational advocacy. These decisions compound over months and years, creating either sustainable operations or mounting attrition risk.

Google’s SRE practices recommend that on-call duty consume no more than 25 percent of an engineer’s time, with at least 50 percent dedicated to engineering work. Achieving this balance requires deliberate management decisions, not accidental outcomes. Managers who treat on-call as an afterthought inherit the consequences through turnover, declining response quality, and team resentment.

This guide focuses specifically on the leadership responsibilities that make on-call programs work, covering strategy selection, health measurement, resource advocacy, and sustainable practices.

Selecting the Right Rotation Strategy

Rotation strategy determines how on-call burden distributes across your team. The wrong choice creates unfairness that erodes trust. The right choice depends on your team’s size, coverage needs, and preferences.

Sequential rotation assigns shifts in exact roster order. User A follows User B follows User C. This approach provides maximum predictability since engineers know precisely when future shifts occur. However, it creates uneven weekend distribution. If User A always covers the first weekend of the month, that pattern persists indefinitely.

Weekly rotation advances each user’s position by one shift per week. Over a complete rotation cycle, everyone experiences different weekdays and weekends. This algorithm solves the fairness problem inherent in sequential rotation while maintaining reasonable predictability.

Fair distribution maximizes time between each engineer’s shifts, providing the longest possible recovery periods. Rather than following a fixed sequence, assignments space out to give each person maximum downtime. This works well for teams where recovery time matters more than predictable patterns.

As a manager, your job is understanding your team’s priorities. Some engineers value predictability for personal planning. Others prioritize maximum recovery time. Discuss options with your team rather than imposing a choice. For detailed algorithm comparisons, see Fair On-Call Rotation Design.

Building the Right Team Size

Team sizing directly determines rotation sustainability. Too few engineers means excessive on-call frequency. Too many creates coordination overhead and knowledge diffusion.

For weekly rotation schedules, target four to five engineers minimum for 24/7 coverage. This ensures each person is on-call roughly one week per month, which research suggests is the maximum sustainable frequency for most engineers. Fewer than four engineers signals a capacity problem requiring attention, not acceptance.

When calculating team size, account for realistic availability. Vacations, sick time, training, and personal leave reduce effective capacity. A four-person team with typical time-off patterns might only have three engineers available in any given month.

If your team is too small, advocate for additional headcount rather than accepting unsustainable rotation frequency. Frame this as reliability investment: burned-out engineers make more mistakes during incidents, take longer to resolve issues, and eventually leave, creating a vicious cycle of increasing burden on remaining team members.

Measuring On-Call Health

Managers need metrics beyond incident counts and response times. Those operational metrics matter, but they miss the sustainability picture.

Alert volume per shift indicates whether engineers face reasonable workload. More than a few pages per week suggests alerting problems requiring remediation. Track this over time to identify trends before they become crises.

After-hours page frequency specifically measures sleep disruption. Occasional overnight alerts are inevitable, but regular 3 AM pages indicate systems that need stability investment or monitoring refinement.

Rotation frequency measures how often each engineer carries the pager. Monthly rotation works for most teams. More frequent rotation signals capacity problems. Less frequent might indicate coverage gaps or over-reliance on specific individuals.

False positive rates directly impact engineer experience. When alerts fire for non-problems, they train engineers to ignore notifications and fragment attention without providing value. Target near-zero false positives for critical alerts.

Beyond quantitative metrics, qualitative feedback matters equally. Regular one-on-ones with on-call engineers reveal issues that numbers miss: unclear escalation paths, inadequate documentation, insufficient training, or unsustainable rotation patterns. The people carrying pagers are best positioned to identify what is and is not working.

Protecting Time Off and Recovery

Sustainable on-call requires genuine time away from pager responsibility. Half-measures do not provide recovery.

Configure roster-wide exclusions for company holidays. No engineer should receive alerts during organization-wide time off. Implement user-specific exclusions that automatically advance rotation when individuals take vacation or personal leave.

Support flexible substitutions where team members can arrange coverage swaps without requiring manager approval for every change. This self-service approach reduces administrative burden while enabling engineers to manage their own schedules around personal commitments.

After particularly difficult on-call shifts with multiple severe incidents or extended overnight engagement, encourage affected engineers to take recovery time. This is not weakness; it is strategic sustainability. Engineers who recover properly perform better on subsequent shifts.

Separating On-Call from Development Work

Attempting both development and on-call duties simultaneously results in fragmented development work, increased stress, and lower quality incident response. Development requires focused blocks of uninterrupted time. On-call demands immediate attention to alerts. These requirements conflict fundamentally.

When engineers are on-call, reduce their development expectations accordingly. Some organizations designate on-call periods as interrupt-driven work focused on operational improvements, documentation updates, and system reliability rather than feature development.

This separation protects both responsibilities. Feature work gets the focus it deserves during non-on-call periods. Incident response gets full attention during on-call shifts. Neither suffers from constant context switching.

Advocating for Resources

Engineering managers must translate on-call burden into language leadership understands. “The team is stressed” resonates less than “alert volume increased 40 percent this quarter while team size remained constant, creating attrition risk that threatens our reliability goals.”

Build business cases using concrete data. Track alert volume trends over time. Document incidents where response time suffered due to inadequate coverage. Calculate the cost of engineer attrition in recruiting, onboarding, and lost institutional knowledge.

Frame on-call investment as reliability investment. Better tooling reduces resolution time. Larger teams enable sustainable rotation frequency. Improved monitoring decreases false positive rates. These investments protect customer experience and business revenue.

When requesting additional headcount, present specific scenarios. “With four engineers, we achieve monthly rotation. With three after the upcoming departure, we move to bi-weekly rotation, which increases burnout risk and likely drives further attrition.” Concrete projections are more compelling than abstract concerns.

Supporting Team Development

On-call provides unique learning opportunities when managed well. New engineers develop system understanding through incident response exposure. Experienced engineers deepen expertise by mentoring others during complex issues.

Implement shadow programs where new team members observe on-call shifts before carrying primary responsibility. This reduces anxiety about first shifts while building confidence through exposure. Pair junior engineers with senior mentors for knowledge transfer during incidents.

After incidents, conduct blameless retrospectives focused on system improvement rather than individual fault. Ask “what circumstances created this situation” rather than “who caused this problem.” This approach surfaces genuine root causes while maintaining psychological safety.

Track on-call performance as part of broader development conversations, but avoid creating perverse incentives. Counting resolved incidents rewards quantity over quality. Tracking response time without acknowledging shift difficulty penalizes engineers who inherit complex problems. Balance metrics with context.

Building Long-Term Sustainability

Short-term heroics are not sustainable operations. Engineers who sacrifice sleep and personal time to maintain unreliable systems eventually leave. Those who remain become exhausted and cynical.

Your role as a manager is building systems and practices that work without requiring heroics. This means investing in alert quality so pages represent genuine problems. It means maintaining team size so rotation frequency stays reasonable. It means enforcing time-off policies so engineers actually recover.

Strong on-call culture transforms operational duty from dreaded burden into sustainable practice. Teams that feel supported, fairly compensated, and psychologically safe maintain reliability while preserving well-being. Teams that feel exploited and unsupported produce the opposite outcomes.

The strategic decisions you make about rotation design, team sizing, and resource advocacy compound over time. Invest in getting them right, and your team will reward you with stable operations and sustained engagement.