"You Build It, You Run It"
The modern DevOps philosophy, popularized by Amazon, is "You build it, you run it." This means developers don't just write code — they carry the pager for that code in production. The person who wrote the service is also the person who gets woken up when it breaks.
This has a powerful benefit: code quality improves dramatically when the author is also the one debugging a 3 AM outage. Engineers start writing better error messages, better runbooks, and better tests — because they know they'll be the ones reading them at midnight.
The Human Cost
On-call is physically and mentally draining in ways that compound over time. Being woken at 3 AM fragments sleep cycles even if the incident resolves in five minutes. Being tethered to a laptop on a Saturday — unable to hike, travel, or drink a beer — restricts your freedom in ways that a salary number doesn't fully compensate. Chronic sleep deprivation degrades cognitive function, decision-making, and emotional regulation. These are not soft concerns; they are engineering reliability concerns. A sleep-deprived engineer makes mistakes.
On-Call Hygiene is the discipline of designing schedules, compensation structures, and alert policies that treat engineers as human beings with biological limits — not as always-available machines. The goal is to make on-call sustainable for years, not just survivable for months.
Key Point: Burnout is a lagging indicator. By the time engineers are quitting, the damage has been accumulating for months. The leading indicators are pages-per-night, alert noise ratio, and how often engineers say "I'm fine" when asked about on-call. Measure the leading indicators before you lose the people.