I have a theory about engineering teams: the boring ones win.
Not boring in the sense of uninspired. Boring in the sense of predictable. When something breaks, they know who handles it. When someone leaves, the work continues. When priorities shift, they adapt without a crisis meeting. Boring means the system works without heroics.
This year put that theory to the test. Layoffs, budget cuts, reorgs, shifting priorities – 2022 threw a lot at engineering teams. The ones that stayed productive through it all shared some common traits. None of them were flashy.
Heroics Are a Smell
When you hear “Sarah stayed up until 3am and saved the release,” that isn’t a success story. That’s a failure story with a happy ending. It means the system depended on one person being awake and willing to sacrifice sleep. That works once. It doesn’t work twice.
At Dropbyke, we had a period where our deployment process was held together by one engineer’s knowledge and willpower. It felt fine until that engineer took a vacation and nobody knew how to handle a rollback. We fixed it by making the process boring: documented, automated, executable by anyone on the team.
Resilience is the opposite of heroics. It spreads capability so that no single failure – human or technical – creates a crisis.
Psychological Safety Isn’t a Buzzword
I know “psychological safety” sounds like something from an HR slideshow. But it’s the most practical thing I can recommend for team resilience.
If people on your team can’t say “I don’t understand this,” “I made a mistake,” or “I think this plan is wrong,” you have a team that hides problems until they become incidents. I’ve seen this play out at multiple organizations this year. The teams that recovered fastest from shocks were the ones where bad news traveled quickly because people felt safe delivering it.
How to tell if your team has it:
- People ask questions in meetings instead of nodding along.
- Risks get raised before they become outages.
- Postmortems discuss systems and processes, not blame.
Leaders set this tone through small repeated actions, not speeches. How you react to the first mistake someone reports to you determines whether they will report the next one.
Kill Single Points of Failure (In Your Org Chart)
Every team has them. The person who’s the only one who understands the billing system. The engineer who always handles the deploys. The manager who’s the only one with context on why the architecture looks the way it does.
These aren’t star performers. They’re organizational risks.
Rotate on-call. Pair on critical systems. Document the “why” behind decisions, not just the “what.” Make sure at least two people can deploy, debug, and recover every service that matters. This isn’t exciting work. It’s the work that keeps the team functional when someone is sick, leaves, or gets pulled onto a different project.
At a large consumer platform, the infrastructure team made a deliberate effort to rotate ownership of critical services every quarter. It was slower in the short term. It made the team dramatically more resilient in the medium term.
On-Call That Doesn’t Burn People Out
On-call is where team resilience gets stress-tested. A sustainable rotation has a few non-negotiable properties:
Alert quality matters. If your pager fires for things that aren’t actually urgent, people stop taking it seriously. Tune your alerts. Remove the noise. Every false alarm erodes the team’s willingness to respond.
Runbooks must work under stress. A runbook that requires calm reading comprehension and background knowledge isn’t a runbook. It’s documentation. Real runbooks work at 2am when you’re tired and anxious: clear steps, decision trees, escalation paths.
Recovery time is real time. If someone gets paged at 3am, they shouldn’t be expected to perform at full capacity the next morning. Build recovery time into the rotation. Burnout is a team problem, not an individual weakness.
Two Habits That Compound
After working with a lot of teams, I think two practices matter more than most:
Retrospectives that produce action items. Not retrospectives that produce feelings. Not retrospectives that produce a long list of complaints. Short, focused retrospectives that produce two or three concrete follow-ups, with owners and deadlines. Then actually do them.
Celebrating progress. Teams under pressure tend to only talk about what’s wrong. Acknowledging what went well – even small things – keeps momentum and morale from decaying. This sounds soft. It works.
The Boring Truth
Resilient teams aren’t built during a crisis. They’re built in the quiet periods before one. Shared knowledge, clear processes, honest communication, sustainable pace – none of it’s exciting. All of it compounds.
The teams that came through 2022 in good shape didn’t have a secret. They had discipline. And discipline, unlike heroics, scales.