
The Slack message arrives at 3:47 PM on Friday: "Revenue forecast showing -23% for Q4. Board deck goes out Monday. Can you take a look?"
You recognize the breed of panic. The model that guided three quarters of planning now claims your bestselling product will lose money next month. The formulas worked yesterday. Today they generate impossible numbers that would submarine everything if anyone believed them.
By 6:30 PM, you’ve traced the error to a data pipeline that’s been quietly corrupting records for weeks. A quick patch gets the numbers back to reality. The board deck ships Monday morning. Messages of gratitude fill your notifications. You feel indispensable, until the same pipeline breaks again next quarter.
This story repeats everywhere. The monthly report that always crashes at deadline or the integration that fails when one person is out. There's the forecast that needs manual edits because systems can’t agree on what a customer is. Each crisis spawns a hero, and each rescue teaches the company that this level of failure is acceptable.
But there’s another kind of work entirely — the kind that makes heroes unnecessary.
The difference between heroic work and sustainable work shows up most clearly on Monday mornings.
Monday #1: The quarterly forecast collapses hours before the board meeting. You spend six frantic hours debugging spreadsheets and chat words of reassurance to everyone else asking when it will be ready. The meeting happens. The quarter proceeds. The sun rises. People see you holding up the ceiling, and that visibility becomes proof of your value.
Monday #2: The forecast runs on time. Slides load cleanly. The CFO clears her throat, clicks forward, and the room shifts into strategy without hesitation. Nobody thanks you. Nobody notices. Which is the point, as it often should be.
This is the strange economy of attention: the broken thing fixed is always louder than the thing that never breaks. Heroics earn gratitude. Prevention earns silence.
Recurring cracks develop caretakers. The Excel formula wizard who can nurse a report through 47 manual steps. The engineer who knows which field breaks if you run the pipeline past midnight. The manager who becomes essential during every monthly fire drill.
These people aren’t scheming; their expertise is real. But their value depends on the dysfunction persisting just enough to need them. That’s why “quick fix now, proper solution later” wins so often. Later rarely comes. Temporary patches harden into permanent scaffolding, until the building is more workaround than design.
The alternative is quieter: building while serving. You still patch what breaks; you have to. But you slip in changes that last. Automate the manual step that always fails. Standardize the definition that keeps drifting. Add monitoring that catches errors before they cascade.
Document what doesn’t break, not just what you fix. Track uptime, not just outages. Not for bragging, but to remind people that absence has real value too.
And when a “new” crisis is really the fourth instance of the same root cause, map it out together. Build memory for dysfunction so the next fix addresses the foundation, not just the symptoms.
Here’s what success looks like: forgetting.
Three months after you stabilize the pipeline, nobody remembers corrupted data. Six months after you automate the forecast, nobody recalls the late-night debugging. A year later, the fact that reports used to collapse on deadline has faded into folklore.
The systems you build become invisible infrastructure. Processes run. Reports generate. Meetings start on time. The crises that once defined your weeks simply stop arriving.
The most valuable work you do is the work nobody notices. The problems you solve are the ones that stop happening. The crises you prevent are the ones that vanish from memory until people can’t imagine why they were problems at all.
The trap is becoming indispensable. The goal is something larger: to build walls that hold themselves, so that on some future Friday afternoon the only thing anyone discusses is weekend plans.
Related reading
Latest entries
Like this? Subscribe via email here.