Don’t Reward the Best Firefighter

Every engineering team has that one “hero.” You know the one: the 3 AM incident response god. They dive into a production dumpster fire, pull a rabbit out of a hat, and somehow bring the system back online.

We love them for it. We give them the shout-outs in Slack, the fat bonuses, and the “Senior Staff” titles because they’re the “go-to” person when everything breaks.

But by making them the main character, we’re lowkey ensuring our culture stays a mess.

For every loud firefighter, there’s an invisible fire preventer. This is the engineer who spends a month refactoring a messy legacy service that no one else wants to touch.

Their work doesn’t show up as a shiny new feature on the roadmap. Their success is silent—it’s the catastrophic outage that didn’t happen six months from now.

And their reward? Often, it’s getting passed over for a promo because their “impact” wasn’t as visible as the person who “saved the day.”

This is a total incentive fail, and honestly, it’s on us as managers. Our performance reviews are fundamentally biased toward reactive work.

We’re great at measuring things we can see on a dashboard:

• Features shipped
• Tickets closed
• Incidents resolved

We don’t have a column on our spreadsheet for “disasters averted.” As a result, we’ve built a career ladder that basically encourages engineers to let things smolder, knowing they’ll get more credit for putting out a blaze than for making sure there’s no fire in the first place.

It’s time to stop treating “impact” as a synonym for “loud activity.” Real impact is the verifiable elimination of future risk.

• The Automation Win: The engineer who fixes a flaky, manual deployment isn’t just “closing a ticket.” They’re giving every dev on the team their time back, forever. That’s massive, compounding impact.
• The Refactor Win: The engineer who cleans up a bug-prone module isn’t just “tidying up.” They are measurably lowering the failure rate for the entire business. That is direct risk reduction.

We need to start hyping up the architects of fireproof buildings, not just the people who are good with a hose.

This takes a conscious effort to hunt for the “invisible” work. We need to use data to quantify risk before it fails and treat the reduction of that risk as a top-tier contribution.

Next time you’re sitting in a performance calibration, ask yourself the hard question: Are we promoting the people who are best at navigating a broken system, or the ones who are actually fixing it?