Rolling Out On-Call Without Losing Your Team (or Your Mind)

The On-Call Dilemma

Rolling out an on-call rotation is one of those leadership rites of passage that feels equal parts necessary and terrifying. Necessary, because someone has to be around when things inevitably go sideways. Terrifying, because done poorly, it can torpedo team morale, erode trust, and burn people out faster than a buggy deploy on a Friday afternoon.

I’ve been on-call in various forms for decades. I’ve done the late-night wakeups, the 3 a.m. brain fog debugging, the “why am I doing this to myself” introspection. So when it came time to roll out a formal on-call rotation for my current team, I wanted to get it right—not just for the business, but for the humans on the other side of the pager.

Step One: Involve the Team Early

This cannot be overstated—bring your team in from the very beginning. On-call changes how people live their lives. It’s not just a schedule shift; it affects dinners, weekends, sleep, mental health.

We asked:

  • What would make on-call suck less?
  • What boundaries feel non-negotiable?
  • How can we make this sustainable, not just survivable?

Some folks were wary, others cautiously optimistic. Everyone had thoughts. Listening was our first win.

Step Two: Design Around People, Not Just Problems

We started with the ideal: a follow-the-sun model. If you’re lucky enough to have a globally distributed team, this is absolutely the gold standard. People on-call during normal working hours are more alert, more effective, and far less grumpy. You’re not dragging someone out of REM sleep to fix a broken dashboard—they’re already at their desk with coffee in hand.

If you’re not there yet (we weren’t either), you can still design better. For example:

  • Cap after-hours expectations. No 24/7 zombie mode unless it’s truly urgent.
  • Give something back. One team I knew gave the following Friday off after a weekend shift—worked beautifully.
  • Create clear escalation paths. The on-call should never feel alone or unsupported.

Step Three: Make the Pager a Last Resort

The most critical thing we did? We invested heavily in our observability and alerting. We declared war on noisy alerts. If something pages you more than once a week and it’s not urgent, it either gets fixed or turned off.

The pager is sacred. If it buzzes, it better matter.

Step Four: Set Realistic Expectations

Here’s what we told the team:

  • If you’re on-call, that’s your only job that week.
    If it’s quiet and you want to knock out a bug fix? Awesome. If it’s chaos, we’ve already cleared your plate.

  • You must have the tools, access, and support to fix the problem.
    No heroics. No digging through six-year-old wiki pages. If the rotation expects action, we give people what they need.

  • Leadership is on-call too (in spirit).
    We read every alert, every postmortem. We don’t jump in unless asked, but we’re there—aware, empathetic, accountable.

Step Five: Measure What Matters

We tracked two things:

  • Mean time to recovery (MTTR)
  • Team happiness

If MTTR dropped but people were miserable, we weren’t succeeding. On-call has to be sustainable. People have to feel trusted, supported, and valued. Otherwise, they’ll just burn out—or worse, quietly disengage.

Final Thoughts

Done right, on-call doesn’t have to be the boogeyman. I’ve been part of teams where it was just… normal. Predictable. Even boring. And boring is great.

What made the difference?

  • A process that respected people’s time.
  • A team that trusted each other.
  • A leadership stance of accountability, not micromanagement.

Roll it out with empathy. Listen more than you talk. Fix what breaks. Protect your people.

If you do that, you won’t just survive on-call. You’ll thrive through it.