Build the runbook before you need it
Nobody hands new managers a plan for this. You get trained on performance reviews and one-on-ones, not on what to do when a teammate's parent dies, or a storm knocks out someone's power for a week, or something genuinely frightening happens to someone on your team. By the time you need that plan, you're already behind.
The fix is boring and unglamorous: build a runbook before the crisis, not during it. Not a policy document nobody reads, just a short, practical reference you can pull up when you're too rattled to think clearly. I break mine into three buckets.
Personal loss and family emergencies. A death in the family, a sick kid, a medical scare. Know your company's bereavement and leave policies cold, before someone needs them, so you're not fumbling through an HR wiki while a person on your team is grieving. Know what to say (short, direct, no scripts: "take the time you need, don't worry about handoffs, I've got it") and what not to say (anything that starts with "let me know if"). People in crisis don't have bandwidth to tell you what they need. Offer something concrete instead.
Collective disruption. This is anything that hits the whole team, or a chunk of it, at once rather than just one person: a hurricane bearing down on your Gulf Coast teammates, a wildfire evacuation order in California, a regional power grid failure after an ice storm, a transit shutdown that strands half your team in a city office. The common thread is that multiple people need support at the same time, and your usual channels might be part of what's down.
Have a way to check on everyone fast that doesn't rely on a single channel. If Slack and email both run through the same internet connection that just went out for a chunk of your team, you need a backup: a group text thread, a WhatsApp group, a shared doc where people mark themselves safe, whatever works for your team. Set this up before the storm, not during it.
Two things worth stealing from bigger companies. PayPal used to issue a physical ID card with a list of phone numbers that connected you to physical security in whatever country you were in, useful if you were traveling and suddenly needed help fast. And most larger companies run an office-wide emergency alert system, the kind that emails, calls, and texts everyone in an affected building or region the moment something happens. If your company has one, know how it works and who's on the other end of it. If it doesn't, that's worth flagging up the chain.
Traumatic and shocking events. This is the one nobody wants to write a runbook for, and it's exactly why you should. One new experience for me as a manager: learning a teammate had been at a large public event when a mass shooting happened nearby. My first instinct was to text. My second was to realize I didn't actually know if I had the right number, or if their emergency contact was current, or who else on the team might have been there too. Luckily, everyone was fine. But it exposed a gap I should have closed months earlier: a simple, current list of how to reach every person on your team outside of work channels.
If that example feels too heavy for a management blog, I get it. But I'd rather write about it plainly than pretend it doesn't happen. These are the moments managers actually get judged on, not the roadmap slides. Handling them well isn't about having the perfect words. It's about having already done the boring logistical work so you can show up as a person instead of scrambling for a phone number.
A few resources worth lining up ahead of time, not after: your company's EAP contact info, local crisis and disaster resources for wherever your team is based, and a clear sense of who on your team has training (first aid, mental health first aid) you could lean on if something happens off-hours.
One easy way to keep this from going stale: bake it into your year-end review cycle. When you're already asking people to update goals or fill out self-reviews, add a line asking them to confirm their emergency contact and how they'd want to be reached if something happened off-hours. It takes two minutes and means the list never goes more than a year out of date.
The runbook itself doesn't need to be elaborate. A single doc with emergency contacts, a communication fallback, relevant policies, and a short "what to say / what not to say" section covers most of it. What matters is that it exists before you're standing in the middle of an emergency trying to write it from scratch. I put together a blank template you can copy: Team Emergency Runbook Template.
My rule with most things in management: if I'd be embarrassed explaining to my team why I didn't have something ready, I build it now. Emergencies don't wait for you to be prepared.