You can't fight what you don't measure.

Burnout is a distribution, and distributions can be fixed.

On-call burnout gets treated as a culture problem, so the fixes offered are cultural, gratitude, pizza, a retro. But burnout's inputs are measurable events. Every page has a timestamp, a recipient, and an outcome. Measure the distribution, the night share, and the pages that led to nothing, and burnout stops being a feeling someone has to confess and becomes a number you can fix.

When a production system degrades, nobody schedules a retro about how it feels, someone pulls the metrics. On-call burnout is the one degradation we handle the other way around, discussed as morale, addressed with appreciation, while the paging system quietly holds a complete event log of the thing causing it.

To be precise about what gets measured, the pages, not the people. Nobody needs monitoring software pointed at a human. The workload arrives pre-instrumented, because every page already carries who, when, and what came of it.

Four reads of the paging history cover most of what burnout is made of:

Left unmeasured, paging load does not spread, it concentrates. Escalations gravitate to whoever answers, so your most reliable responder accumulates load precisely because they are reliable, and the better they handle it, the more invisibly it grows. Seat-rationed tools make it worse, concentrating pages on the licensed few, but the gravity exists everywhere.

The hero problem is invisible in every view except the distribution. Measured, it is a rebalance, a rotation change made this week. Unmeasured, the first time it becomes visible is usually the resignation letter.

There is a humane point under the arithmetic. "I am burned out" is a confession, and confessions are hard to make and easy to soften into "I'm fine." "I took 47 of our 80 pages this month, and 19 landed after midnight" is a fact, and facts do not need courage to state or interpretation to act on. Measurement moves the conversation from a mood a manager has to detect to a workload a manager can fix, which is fairer to both sides of the table.

Before any wellness initiative, look at the futility rate. The pages that woke someone and required nothing are the purest burnout accelerant in the system, all cost and zero purpose, and unlike rotations or staffing they can simply be removed, tune the condition, stop paging on it, let it wait for morning. Every futile page deleted is subtracted from every rotation, every month, from now on. It is the highest-leverage kindness in on-call, and it looks like signal hygiene.

Signal9 treats paging load as an operational metric, because it is one, escalation analytics show the distribution, the timing, and the outcomes behind every page, and unlimited users means coverage decisions never concentrate load for licensing reasons. The measuring is done. The conversation it enables is yours.

How do you measure on-call burnout? From the paging history, which already records every event. Distribution of pages per responder. The share landing at night and on weekends. The futility rate, pages that required no action. And the conditions that page repeatedly. Together they show who is absorbing the load, what it costs, and how much of it never needed a human.

Why does on-call load concentrate on one person? Because escalations gravitate to whoever reliably answers, so reliability attracts load. The concentration is invisible without distribution numbers, and the person carrying it usually will not raise it themselves. Measured, it is fixed with a rebalance, unmeasured, it tends to surface as a resignation.

Does Signal9 track on-call workload? Yes. Signal9 shows the workload behind every page, who was paged, when, through what channel, and what came of it, with escalation analytics that make distribution across responders visible. Pages that led to nothing are visible too, so the noisy conditions creating them can be tuned instead of endured.