Every engineering organization has a hero. The person who saves the launch at the last minute, who fixes the outage nobody else understands, who works the weekend so the deadline holds. We celebrate them, name them in all-hands, lean on them a little more each quarter. And in celebrating them, we very often reward the exact dysfunction that is slowly hurting the team and, eventually, the hero too.
Heroism is a symptom
The recurring need for heroics almost always means something upstream is broken. Unrealistic timelines that only superhuman effort can hit. A critical system that only one person understands. Processes that fail the moment they meet normal pressure. The hero is not the cure for these underlying problems; the hero is the bandage that lets the organization avoid looking at them for one more cycle.
If your system requires a hero to survive Friday, you do not have a hero problem. You have a system problem that happens to be wearing a cape.
Every time the hero saves the day, the organization receives a small, dangerous reward for not fixing the thing that made the rescue necessary. The dysfunction gets reinforced precisely because the heroics keep masking its consequences.
The compounding costs
Hero culture is expensive in ways that rarely show up on any dashboard until it is too late.
- Bus factor. Knowledge concentrates in the hero, who becomes a single point of failure your business genuinely cannot afford to lose.
- Burnout. Heroes burn out, reliably, and when they leave they take irreplaceable context out the door with them.
- Stunted teammates. Everyone else learns to wait for the hero rather than building their own capability, so the dependency deepens.
- Distorted incentives. You end up rewarding firefighting over fire prevention, so, rationally, you get more fires.
What to celebrate instead
I work hard, and it is genuinely hard, to redirect recognition toward the unglamorous work that prevents crises in the first place. The engineer who wrote the documentation that meant no 2am call was needed. The one who automated the painful deploy so nobody has to babysit it. The one who built the monitoring that caught the issue before a single customer noticed.
Enjoying this article?
Get more like it in your inbox — practical engineering leadership, fintech, and AI. No spam, unsubscribe anytime.
This work is invisible by design; its whole purpose is that nothing dramatic happens. Which is exactly why leadership has to make it visible deliberately, because the organization's natural attention flows to the dramatic save and away from the quiet prevention that made saves unnecessary.
Helping a hero let go
Often the hardest part of dismantling hero culture is the hero themselves, whose professional identity has become wrapped up in being indispensable. Being the person who saves the day feels good, and being asked to stop can feel like being asked to become less valuable.
The kindest and most strategic thing I can do is reframe the value. I insist they spread their knowledge, pair, document, mentor, and I make that the new definition of their contribution. A senior engineer who makes themselves replaceable is far more valuable to me than one who cannot take a two-week holiday without the on-call rotation quietly panicking.
Building the resilient alternative
The replacement for hero culture is not a culture with no excellence; it is a culture where excellence is systemic rather than personal. Shared ownership of critical systems so no single absence is a crisis. Realistic planning that does not secretly depend on someone's weekend. Processes that hold under pressure rather than requiring a rescue. Documentation and automation as first-class work, not the thing you do if there is time left over, which there never is. This is less dramatic and far more durable, and durability is the entire point.

Conclusion: a quieter, stronger kind of team
The goal is a team where no single person's absence is an emergency, where the weekend reliably stays the weekend, and where reliability comes from the design of the system rather than from the exhaustion of one indispensable person. That is a quieter kind of strength than the dramatic last-minute save, and it is a vastly more durable one. Stop celebrating the rescue. Start celebrating the engineering that made the rescue unnecessary. Your best people, and your sleep schedule, will both last much longer.
Get new posts in your inbox
Occasional, practical notes on engineering leadership, fintech, and building with AI. No spam, unsubscribe anytime.
Comments (10)
Leave a Comment
Salisu Adamu
July 29, 2026
Question on "Heroism is a symptom" — how do you actually apply this with someone you personally hired? Struggling with that specific case on my engagements right now.
Wale Ogunleye
July 21, 2026
Would add: the incentives inside the calibration room matter as much as the actual conversation with the engineer.
Toby Sutherland
July 7, 2026
Independent consultant, 5 due-diligence engagements a year across UK/EU fintechs. Reading this on the walk in — the "The compounding costs" bit lands, because I spent this week trying not to fumble monolith-to-services trap on my engagements. Honestly the framing would have saved me at least a promo conversation.
David Lee
June 30, 2026
EM for 3 years, was IC for 6 before that. Reading this after a rough sprint — the "Heroism is a symptom" bit lands, because I spent this week navigating calibration season on my org. Honestly the framing would have saved me at least a headcount ask I lost.
Christopher Clark
June 29, 2026
24 months into my first eng job at a mid-size fintech, so a lot of this is above me, but the "Heroism is a symptom" bit made a concept click that I had been nodding along to in code review for months. Thanks for writing at a level that doesn't gatekeep newer engineers out.
Kofi Mensah
June 29, 2026
One more thing worth naming: the incentives inside the calibration room matter as much as the actual conversation with the engineer.
Abiola Ogunbanjo
June 25, 2026
Does the "The compounding costs" still hold on a 3-engineer team? We're at the smaller end of that and some of these patterns feel like they need a dedicated ops person to run properly.
Ikechukwu Okafor
June 24, 2026
The "Conclusion: a quieter, stronger kind of team" section is doing a lot of work.
Kevin Perez
June 24, 2026
Not fully sold — the "Heroism is a symptom" advice assumes the manager has real hire/fire authority. In a heavily political org the sequencing has to change.
Adeola Olatunji
June 22, 2026
Refreshing to read this framed for our market rather than lifted from a Silicon Valley playbook. Specifically the "Conclusion: a quieter, stronger kind of team" piece — NIBSS write it into the audit questions, and that changes the design constraints in ways the US-centric literature never touches.
