The Near-Miss Problem in Accounting
In 1976, NASA created a system that rewarded people for admitting mistakes.
The Aviation Safety Reporting System invites pilots, crew, and air traffic controllers to report near-misses: moments where something could have gone wrong but didn’t. A miscommunication that almost caused a runway incursion. A warning light briefly ignored. A checklist step skipped under time pressure.
The reports are confidential. There is no punishment. Report a near-miss within ten days, and you are protected from disciplinary action for the incident itself.
This sounds backwards. Why protect people who made errors?
Because the industry had learned something expensive: disasters rarely come from nowhere. They are the last domino in a chain of smaller failures that nobody talked about. The near-miss that got ignored. The shortcut that worked once and became a habit. The question someone thought about asking but didn’t.
Since 1976, the system has collected over 1.7 million voluntary reports. Aviation is now one of the safest industries on earth. More reported problems. Fewer actual disasters.
The mechanism is simple: surface the small failures early, before they compound into large ones. The only thing standing between a near-miss and a pattern is whether anyone said something.
The same chain exists in accounting
The errors that hurt clients are rarely first-time events. A misclassification that persisted across quarters because no one flagged it in month one. A revenue recognition approach that was slightly off from the start and only surfaced during due diligence. A cash flow projection built on an assumption that was never challenged.
By the time these problems become visible, they have usually been sitting in the books for a while. The cost is not just fixing the error. It is the restatements, the delayed close, the investor conversation that should have been straightforward, the trust that takes longer to rebuild than the books did.
What accounting has not historically done well is create the conditions where people surface the small thing before it becomes a large one. The culture in most firms runs in the opposite direction. Errors are embarrassing. Questions signal uncertainty. Raising a doubt about someone else’s work feels like an accusation. So the doubt goes unspoken, the shortcut stays in place, and the near-miss never gets logged.
The thirty-second fork
There is a specific moment worth recognizing, because anyone who has worked in accounting has been there.
You are reviewing something. A journal entry, a reconciliation, a client deliverable. Something feels slightly off. The number does not match your mental model. The approach is one you have not seen before. You cannot quite articulate it, but something is pulling at you.
One path: you tell yourself it is probably fine. The person who prepared it knows the account better than you do. You are already behind. Asking will slow things down. You move on.
The other path: you pause and ask. Can you walk me through this? I want to make sure I understand it. Does this look right to you?
The gap between those two paths is about thirty seconds of mild discomfort. The gap in outcomes can be significant.
Aviation did not become safe because pilots stopped making mistakes. It became safe because the industry built systems that made it easy to talk about them. The near-miss got reported. The pattern got spotted. The procedure got updated before anyone got hurt.
What this means for founders and business leaders
If you lead a finance team, the question worth sitting with is not whether your people make errors. They do, because everyone does. The question is whether the environment makes it easier to surface them or easier to bury them.
If someone flags an error in their own work, that is quality control working. The failure is when the error travels all the way to the client, or to the audit, or to the cap table, because nobody felt safe raising it at the thirty-second fork.
For founders working with an external accounting team, the same principle applies. The firms that serve you best are the ones that will tell you when something looks wrong, even when the answer might be that the business is harder than the model suggested. That conversation, the one that feels like bad news, is usually the one that saves you six months of going in the wrong direction.
High standards require honest communication. You cannot fix what you cannot see, and you cannot see what people are afraid to show you.
The next time someone on your team pauses and asks a question that turns out to have an obvious answer: that is a good sign. They asked. They checked. The near-miss got logged.
That is how you build something that holds.