Heroics Hide the Broken System

Category: Organisations and systems The rescue gets rewarded while the rot that made it necessary goes ignored.

There is always a name.

In every team that has been under strain for a while, there is a name that comes up when things go wrong — the person you call at eleven at night when the deployment falls over, the one who knows where the bodies are buried, who can talk the angry customer down, who stays late and fixes the thing nobody else can fix. When the emergency hits, everyone exhales, because they know who to ring. And the person on the other end of the phone picks up, because they always pick up, and they sort it out, because they always sort it out.

It is genuinely impressive, and I want to be clear about that, because what follows can sound like an attack on the people who do this, and it isn’t. The hero is usually skilled, usually generous, usually the most committed person in the room. The problem is not the hero. The problem is what the hero’s existence allows everyone else to stop noticing.

Picture the night it goes wrong. A system that should not have been fragile falls over at the worst possible time, and the hero is summoned. They work into the small hours, find the obscure cause, apply the fix, and by morning the thing is running again. There are messages of thanks. Someone says, not for the first time, “I don’t know what we’d do without you.” There may be a small reward — a mention in the all-hands, a bonus, a quiet word from someone senior. The hero, exhausted and a little proud, accepts it, because it was hard and they did it and the praise is deserved.

And that is the exact moment the real failure escapes unexamined.

Because nobody asks the question that matters. Nobody asks why the system was fragile enough to fall over in the first place. Nobody asks why a piece of infrastructure the whole business depends on had a single point of failure that only one human being understood, or why the fix lived in one person’s head and not in a document, a test, or a design that made the failure impossible. The crisis was survived, the survival was celebrated, and the conditions that manufactured the crisis were left fully intact, ready to manufacture the next one.

So the next one comes. Of course it does. And there is the name again, and the phone rings again, and the hero rides out again, and is thanked again. The organisation has not solved a problem. It has acquired a dependency, and learned to feel grateful for it.

The Principle

When extraordinary individual effort repeatedly rescues a broken system, the rescue masks the breakage instead of fixing it. The save is visible, dramatic, and rewarded; the underlying fault is invisible, boring, and ignored. So the organisation comes to depend on heroes and grows blind to the conditions that keep producing the emergencies the heroes are called in to solve.

The heroics are not the cure for the broken system. They are the thing that lets the broken system survive without ever being repaired. Every successful rescue is also a missed diagnosis — proof that the organisation can absorb the failure, and therefore a reason not to address it. The hero is, without meaning to be, the load-bearing patch over a crack that nobody will now fix, because the patch is holding.

And the better the hero is, the longer the crack goes unrepaired. Competence buys time, and the organisation spends that time on everything except the thing that needed it.

Why It Is Inevitable

It is inevitable because the rescue and the root cause are wildly mismatched in how they present themselves, and we reward what we can see.

A rescue is a story. It has a clear shape — a threat, a struggle, a named person who turned up and won. It happens in compressed time, under pressure, with an audience watching. It is easy to point at, easy to praise, easy to put in a performance review. Gratitude flows toward it naturally, because the relief is real and the effort was visible.

A root cause is the opposite of a story. It is a dull structural fact — a missing test, an undocumented process, a design that concentrated risk in one place, a workload that guaranteed mistakes. Fixing it is slow, unglamorous, and invisible when it works, because a fixed root cause produces the absence of a crisis, and you cannot throw a thank-you at something that didn’t happen. The person who re-architects a system so the 2 a.m. call never comes again gets no 2 a.m. call to be thanked for. Their reward is silence, which feels, organisationally, like nothing happening.

So the incentives split cleanly. The dramatic save is concentrated, immediate, and celebrated. The boring fix is diffuse, delayed, and unnoticed. Faced with that, any rational organisation drifts toward valuing the save — not by deciding to, but by repeatedly reacting to what is in front of it. Each individual act of gratitude is correct. The pattern they add up to is a trap.

It compounds because the hero, too, is caught in it. Being the person who saves the day is a powerful identity. It feels good to be indispensable, to be the one they call. Some heroes — not all, but enough — come to derive their standing from the rescue, and a person whose worth is built on rescues has no incentive to eliminate the conditions that produce them. Not consciously, not cynically — but the part of you that is valued for solving the fire does not lie awake inventing fire prevention. The system that rewards heroics quietly recruits its heroes into preserving the very fragility they keep saving everyone from.

And there is the simple fact of time. The hero is summoned precisely when there is no time to do anything but stop the bleeding. The moment to fix the cause is afterwards, calmly, when the pressure is off — but afterwards the hero is exhausted, the crisis has passed, the urgency has drained away, and there is a fresh queue of work demanding attention. So the post-mortem doesn’t happen, or happens as a box-tick, and the structural fix slides down the list behind everything that feels more pressing. The cause survives not because anyone defended it, but because the only window to fix it closed while everyone was recovering.

None of this requires anyone to be foolish. The people thanking the hero are right to be grateful. The hero is right to be proud. The organisation is right to be relieved. Every reaction in the chain is reasonable, and the chain produces an outcome no one would choose: a business that has confused surviving its failures with not having them.

How It Shows Up

• There is a named person — or a small handful — without whom critical things simply stop working, and everyone knows their name
• Praise, bonuses, and promotions flow toward dramatic saves, while quiet prevention work goes unremarked
• “I don’t know what we’d do without you” is said as a compliment, not heard as an alarm
• The same category of emergency keeps recurring, and each time it is treated as a fresh surprise rather than a pattern
• Post-mortems get scheduled and then quietly skipped once the crisis has passed and the urgency has gone
• The hero is too busy fighting fires to ever document, train, or design out the fires
• When the hero takes leave, the team holds its breath — and sometimes the emergency politely waits for them to come back
• Knowledge lives in people’s heads and war stories, not in systems, tests, or documents
• Stability is described in terms of who is reliable, never in terms of what is robust

Why It Causes Damage

The damage is not the long night the hero works. It is everything that long night allows the organisation to avoid.

The first cost is the unrepaired fault itself. Every rescue that doesn’t end in a fix leaves the same fragility in place, and fragility doesn’t stay still — systems grow, dependencies pile on top of the weak spot, and the failure that was survivable at small scale becomes catastrophic at large scale. The organisation is not getting safer between emergencies. It is getting more exposed, while feeling reassured, because the last one was handled.

The second cost is concentration of risk. A business that depends on heroes has put its critical capabilities inside individual human beings — their memory, their availability, their willingness to pick up the phone. That is the most fragile place you can store anything important. People get sick. People take holidays. People burn out, and heroes burn out faster than anyone, because the role has no off switch — the better you are at being indispensable, the less the organisation can ever afford to let you rest. And people leave. When a hero leaves, they don’t just take a skilled employee; they take the only working copy of how several critical things actually function, and the cracks they were silently patching all open at once. The organisation discovers, far too late, that it never had stable systems — it had one stable person, and they’re gone.

The third cost is the corrosion of judgement. An organisation that keeps surviving its failures learns the wrong lesson from survival. It concludes the failures are tolerable, because look — we always get through them. Each rescue is filed as evidence of resilience when it is actually evidence of recurring breakage. The dashboards look fine between crises. The story leadership tells itself is one of a tough, capable team that handles whatever gets thrown at it. What is actually happening is that a small number of people are absorbing the cost of a problem no one will name, and the absorbing is mistaken for health.

And there is a quieter effect on everyone who isn’t the hero. When rescues are what gets rewarded, you teach the whole team that the path to recognition runs through visible crisis. The rational career move becomes to be present and heroic when things are on fire, not to quietly prevent the fire — prevention earns you nothing and, worse, removes the stage on which heroism is performed. So you breed a culture that is, at the margin, faintly invested in things going wrong, because going wrong is where the glory is. You will rarely see this stated. You will see it in who gets promoted, and in how little anyone wants the boring job of making the emergencies stop.

Worst of all, the whole pattern is self-concealing. A system propped up by heroics looks, from a distance, like a system that works — because it does work, right up until the hero is unavailable at the wrong moment. The success of the rescues is exactly what hides the need to fix anything. The organisation is not told it is fragile. It is told, repeatedly and warmly, that it is lucky to have such good people. And it is — which is precisely why it never asks the question that would save it.

How To Counter It

You cannot counter this by being less grateful to the people who save you. They earned the gratitude, and withdrawing it just punishes the wrong thing. You counter it by refusing to let the rescue close the case — by treating every heroic save as the opening of an investigation, not the end of one.

• Treat “we had to be heroic” as a defect report, not a success story. The need for extraordinary effort is itself the bug. After every rescue, the question is not “who do we thank?” but “what was broken enough that a person had to do that, and how do we make sure no person ever has to again?”
• Make the post-mortem non-negotiable, and protect the time for the fix. The window to address a root cause is after the crisis, when it’s calm — and that is exactly when the urgency evaporates and the fix gets dropped. So schedule it as real, funded work before the relief wears off, and treat skipping it as a serious failure, not an understandable one.
• Reward the prevention, not just the rescue — and say so out loud. The person who quietly re-engineered a system so the 2 a.m. call stopped coming should be visibly more celebrated than the person who heroically answered it. This is hard, because prevention is invisible — so you have to go out of your way to make it visible, naming the crises that didn’t happen and crediting the people who killed them.
• Hear “I don’t know what we’d do without you” as an alarm. Any time a person becomes structurally indispensable, that is a risk to reduce, not a strength to enjoy. Spread the knowledge, document the process, build the redundancy. A healthy organisation should be able to lose any one person without holding its breath.
• Make heroes design themselves out of a job. The highest form of the role is not the person who can always fix it — it’s the person who makes the fix unnecessary. Reframe the hero’s value around that, and free their time to do it, because a hero permanently busy fighting fires can never build the thing that prevents them.
• Count the recurrence. Track how often the same category of emergency comes back. If a problem has been heroically solved three times, it has not been solved once — it has been patched three times, and the recurrence is proof the cause is still there, waiting.

What Good Looks Like

Good looks, frankly, a little boring.

It looks like a team where the critical systems are robust enough that nobody’s phone rings at eleven at night, and where, if it did ring, more than one person could answer it. Where the most respected people are not the ones who perform the most dramatic rescues, but the ones whose corner of the business is so quietly reliable you forget it’s there. Where someone can go on holiday — even the someone — and nothing holds its breath.

It looks like a culture that treats the need for heroics as the failure, and the absence of heroics as the achievement. Where after something goes wrong the loudest conversation is not about who saved it but about what allowed it, and that conversation reliably ends in a structural change, not just a thank-you. Where the people who prevent crises are more celebrated than those who survive them — and where prevention is made visible on purpose, because everyone understands an invisible win is the easiest kind to stop rewarding.

It does not look like an absence of skilled, committed, generous people. You want those people; they are the best part of any team. It looks like an organisation that has them and refuses to depend on them being superhuman — that takes their hard-won knowledge and bakes it into systems, so the brilliance survives the person and the rest of us can sleep. The goal was never to get rid of the hero. It was to build a place that no longer needs one, and to let the hero finally rest.

A Reflective Question

Think of the last time someone in your organisation saved the day. When the relief had passed, did anyone ask why the day needed saving — and did the answer to that question ever actually get fixed, or are you simply waiting, gratefully, for the next time you’ll need them?