On-call compensation calculator

On-call compensation — what you should actually pay

On-call is the single biggest quality-of-life complaint in engineering orgs, and one of the most commonly under-compensated. Most teams pay zero for on-call — it's baked into salary — and wonder why senior engineers burn out. This calculator models a proper on-call comp structure (passive stipend + active pay) so you can set fair rates.

The two costs of on-call

1. Passive burden (168 hrs/week): Even with zero pages, being on-call means no drinking, no phone-off activities, no being far from a laptop, sleeping with ringer on. Life quality suffers regardless of incident volume.
2. Active response (minutes per page): Actually handling an incident. Nights and weekends double-up on life disruption. A 45-min page at 3 AM costs the engineer their next-day productivity too.

Compensation structures

Most sophisticated orgs pay both:

• Passive stipend: Fixed payment for being on-call. Typically $150-500/week. Flat rate regardless of incidents. Acknowledges the 168-hour burden.
• Active pay: Hourly rate during actual incident response. Usually 1.5-2x the engineer's base hourly rate for off-hours pages. Business hours usually don't pay extra (already at work).
• Comp time: Alternative to money — hours paged = hours off later. Popular with some orgs. Trades budget pressure for scheduling pressure.

Benchmark comp

Industry data (PagerDuty, Google SRE book, various tech company disclosures):

• Senior engineer, 8 weeks/yr rotation, 4 pages/week avg: Total on-call comp typically $5-15k/year on top of base.
• Staff / principal, 6 weeks/yr, 3 pages/ week: $8-20k/year. Higher hourly due to senior base rate.
• Heavy on-call (12+ weeks, 10+ pages/week): $15-40k/year. Usually signal that system health needs serious investment.

Total on-call comp of 3-8% of base salary is typical. Outliers above 10% usually indicate either generous comp or systems needing reliability investment.

When salary is "enough"

Some orgs argue senior salaries already include on-call — "you're highly paid, deal with it." This works if:

• Base comp is > top-of-market (25-40% above median for role + region)
• On-call volume is very low (< 1 page/week on average)
• Rotation is short (< 4 weeks/year) and team is large (> 8 engineers rotating)

Most orgs don't meet these thresholds. Without explicit on-call comp, you get: (1) resentment, (2) turnover in on-call-heavy roles, (3) passive- aggressive refusal to carry pager, (4) senior engineers negotiating no-on-call clauses.

The real fix: reduce pages

The cheapest on-call program is the one that rarely fires. Reducing page volume 50% saves more than any comp adjustment:

1. Error budgets / SLOs: Google SRE approach. Define acceptable failure rate; only page when budget is being burned.
2. Auto-remediation: Known failure modes get scripted auto-recovery. Page only on failures of recovery itself.
3. Alert hygiene review: Every alert that fires without requiring action is noise. Weekly review to kill false positives.
4. Business-hours-only non-critical: Reserve off-hours pages for customer-facing incidents. Defer everything else to morning.
5. Post-incident culture: Every page becomes a question: "Can we make this not page again?" Tracked and prioritized.

Rotation structure

Good rotation design prevents burnout:

• 6+ engineers per rotation: Fewer means each person is on too often.
• 1-week rotations: Longer rotations accumulate fatigue. Shorter rotations cause context-switching overhead.
• Primary + secondary: Primary pages first, secondary backs up if no response. Allows primary to occasionally miss a page without incident.
• Follow-the-sun: For global teams, EU/US/APAC rotations can cover 24/7 without off-hours pages for anyone.
• Swap-friendly: Easy swap process for scheduled conflicts. Resentment builds when swaps are painful.

Legal considerations

• Exempt vs non-exempt: In the US, non-exempt employees must be paid for all work time including on-call. Exempt employees (most engineers) have more flexibility but fairness matters.
• EU rules: "On-call time" can count as working time under EWTD (European Working Time Directive), especially if response is required. Check local laws.
• California: "Reporting time pay" requires minimum compensation when called in, even briefly.
• Union considerations: In unionized environments, on-call comp is often codified in contract.

Red flags that signal under-compensation

• Engineers negotiating no-on-call at hire
• Tenure-weighted preference to skip on-call (seniors pushing it to juniors)
• Complaints about "who has the pager this week" in standups
• Burnout patterns (health issues, reduced performance) correlating with on-call weeks
• Turnover specifically citing on-call as reason

If you see these, raise comp, raise hiring bar for team capacity, or invest in reliability to cut page volume. Usually all three.