I love asymmetries. Here is one that explains most of what is wrong with large organizations.
Failure is visible. Inaction is invisible.
That's it. That single asymmetry, left unmanaged, will quietly rot an institution from the inside while everyone inside it feels perfectly safe.
When you fail visibly — a bad experiment, a wrong call, an initiative that goes nowhere — you leave a trace. A name, a ticket, a postmortem. Someone is associated with it. The org remembers.
When you fail by not acting — the integration never proposed, the experiment never run, the conversation never had — nothing happens. No artifact, no counterfactual, no cost center. The ledger is silent. Nobody is blamed because there is nothing to blame.
This is not a psychology problem. It is an incentive problem. People respond rationally to what gets recorded. And what gets recorded is only what happened.
Two very different kinds of organizations.
The first kind — call them action-tolerant — treat a well-reasoned failure as tuition. They ask not just "what went wrong" but "what did we decide not to do, and were we right?" They make inaction visible. They mourn the unrun experiment. Google in its early years, Amazon under Bezos, Berkshire in its early decades — organizations where the person with the best judgment eventually won, because judgment was actually being tested.
The second kind — call them inaction-optimized — have learned, through incentive and example, that the safest move is to never be visibly wrong. Approval chains grow longer, not because decisions got harder but because spreading sign-off dilutes individual exposure. Meetings multiply. Alignment becomes the work. Nobody fails because nobody commits.
And it gets worse. In the inaction-optimized org, the scorecard is not even balanced. Make nine customers happy and one angry, and it is the one angry customer that gets remembered, escalated, discussed. This is not malice. It is loss aversion — the well-documented human tendency to feel losses more acutely than equivalent gains. Organizations don't just fail to record inaction. They also systematically overweight the failures they do record. The rational response, if you are paying attention, is to do as little as possible. And most people, eventually, are paying attention.
This is most of the Fortune 500. This is most government. This is most large banks, most legacy telecoms, most healthcare systems. Not because the people are bad. Because the ledger is broken.
It feels safe. The failures are quiet and the feedback loops are slow. You can spend years inside an inaction-optimized organization without feeling the cost — because the cost is denominated in things that never existed.
And by the time it's visible, it looks like a talent problem, or a strategy problem, or a culture problem. It is none of those. It is a ledger problem. The inaction-optimized org built a perfect system for eliminating visible failure, and achieved it by eliminating action.
What remains is not safety. It is a very clean, very legible path toward irrelevance.
Make inaction visible. Name the experiments that didn't run. Ask who talked themselves out of the right call. Reward the person who acted on good judgment and was wrong over the person who never acted at all.
Because one of them still has their judgment intact.
Tarek Amr, April 1, 2026