There is a specific moment, usually sometime in the second week after the close, when the question arrives not as a plan but as a verdict: can I actually do this again? For some founders it arrives as a flat statement of fact, delivered in a voice that has already decided. That voice is not giving you information. It is giving you a classification, and the classification is almost certainly wrong.
The classification runs like this: the company failed because of something that is fixed in you, something that was always there and will be there next time, which means trying again is an exercise in repeating the same outcome. This is not a finding. It is a specific cognitive pattern that researchers have measured, named, and traced to its consequences.
What the stability dimension actually is
Martin Seligman's research on explanatory style identifies three dimensions in how people explain bad events to themselves. One of them is stability: whether the cause of a bad event is seen as permanent and unchanging, or as temporary and specific to this situation. The stable version of the shutdown story sounds like "I lack the judgment this takes." The unstable version sounds like "the market contracted six months before I had runway to survive it."
Both can follow the same shutdown. The shutdown does not choose which version you believe. You do, usually fast, usually under duress, and usually based on the reading that formed in the worst 72 hours. Seligman's work shows that stable attribution after failure is one of the strongest predictors of what he terms learned helplessness: the state where a person stops trying because they expect the outcome to be the same regardless of their action.
In founders this does not always look like visible paralysis. It looks like a persistent inability to generate genuine interest in a next move, a sense that looking at an opportunity is somehow dishonest, or a clean certainty that the next attempt would expose the same flaw. The activity stops before it starts, not because of fear exactly, but because the stable verdict has already delivered its answer.
What most shutdowns actually are
Here is where the failure research is useful. Amy Edmondson's taxonomy of organizational failures draws a line between three types. Preventable failures come from deviating from a known good process. Intelligent failures are the expected outcome of a reasonable test in novel territory. Complex failures, which are the most common category in startup shutdowns, arise from multiple factors interacting in an uncertain environment, with causes that are partly in the external conditions, partly in the sequence of decisions made, and very few of which are stable personal traits.
A market that moved, a capital environment that closed faster than the next round, a sequence of individually plausible decisions that compounded badly: none of these is a stable property of the person who ran the company. When you assign a complex failure to a single stable internal cause, you are collapsing what Edmondson describes as a multi-factor systemic event into a single permanent verdict. That collapse is both inaccurate and, in practical terms, more expensive than the failure itself, because it forecloses future decisions based on a wrong input.
The two founders who do not check
There are two predictable failure modes in how founders carry stability attribution into what comes next.
The first founder assigns the failure entirely to external, unstable causes and moves too fast into the next bet without identifying what in their own patterns is worth changing. The cause was unstable in their account, which means nothing to update. This is the error of pure externalization: it carries the unchanged inputs forward and calls it resilience.
The second founder assigns the failure to a stable internal cause, treats it as a permanent verdict, and either stays frozen or avoids the kind of move that would test the belief. The cause is permanent in their account, which means nothing to try. This is the stability trap: the verdict forecloses the test that would correct it.
Neither founder is checking what the cause actually was. One moves too fast, one stays too still, and both are operating from a label they applied under the worst cognitive conditions of the past several years.
How to check the label
What separates the stability trap from honest self-assessment is not optimism. It is specificity. "I lack the judgment this takes" is a stable global trait. "My judgment on enterprise sales cycles was wrong in this market, in this capital environment, in a rate cycle that compressed timelines" is specific, situational, and at least partially unstable.
Kahneman's work on System 1 thinking explains why the global version forms first: under stress the mind reaches for fast, pattern-matched verdicts, not careful analysis. The stable attribution is the System 1 answer. It requires an effortful second pass to ask whether the cause that is now being treated as a fixed personal trait was actually situational, contextual, or part of a complex failure that had more moving parts than a single verdict can hold.
The question to work with is not whether you could have done better, but whether the specific thing that caused the failure is something that is fixed in you or something that was specific to that context. Those are different questions, and most founders never get to the second one because the first one arrives so fast and so loud.
The cause that closed your company is probably not a permanent feature of who you are. Most complex failures are not. But you will not know that until you look at what the cause actually was, with enough precision to tell the difference between what was situational and what, if anything, genuinely needs to change before the next move.