Every team has a "that one thing we forgot" story. The server nobody restarted, the invoice nobody sent, the backup nobody tested. The fix isn't more meetings — it's a written recovery drill. Here's the 30-minute version we use, free to copy.
The 30-minute drill
- Pick your worst failure (5 min). Not the most dramatic — the most likely. Usually it's "the person who does X is unavailable."
- Write the 10-line recovery card (10 min). One line per step, in the order a stranger would need them. Include credentials locations (not the credentials), the one command that fixes 80% of cases, and who to call if line 7 fails.
- Hand it to someone else (10 min). They execute the card literally — no explanations allowed. Every question they ask is a gap in the card; patch it on the spot.
- Date it and store it where people look (5 min). A drill that lives in a forgotten wiki tab is theater. Pin it next to the thing it protects.
The rule that makes it stick
If the drill hasn't been run in 90 days, it doesn't exist. Recurring calendar invite, 30 minutes, quarterly. Cheap insurance beats expensive archaeology every time.
What you get
- A recovery card your future self (or a new hire) can follow at 2 a.m. without heroics.
- A list of your actual single points of failure — they surface embarrassingly fast.
- Documentation that stays true, because it's re-verified every quarter instead of rotting.
If you'd rather not build the card templates from scratch, we packaged them — recovery cards, handoff checklists, and the drill calendar — in the Ops Starter Kit and the broader Automation Starter Pack.
What's your team's "that one thing we forgot" story? Drop it in the comments — I'll reply with the first line of its recovery card.
Top comments (0)