Tell me about a time your analysis turned out to be wrong. What happened, and what did you change?

The shape of a strong answer matters more than the specific incident, because the interviewer is testing intellectual honesty and whether you learn systematically.

A good answer has five parts. First, a specific and real example — an assessment made, a judgement that turned out wrong, with enough detail to be checkable. Vagueness here reads as either no experience or no honesty.

Second, the actual mechanism. Not "I did not have enough information", which is almost always available as an excuse, but the reasoning failure: anchoring on the first plausible explanation and interpreting everything after it through that lens; confirmation bias in the collection, where I stopped looking once the favoured hypothesis was supported; mirror imaging, assuming the adversary valued what I would have valued; or over-weighting one source because I trusted the brand.

Third, how it was handled. Reissuing the assessment plainly and telling the people who acted on it. Analysts who correct themselves openly build more credibility than they lose — the ones who quietly let a wrong assessment stand are the ones who become untrustworthy.

Fourth, a concrete process change. "I am more careful" is not a change. Adding an explicit alternative hypothesis to every assessment is. So is a key assumptions check before writing, or having a colleague argue the opposite case, or a rule that any judgement resting on a single source says so in the text.

Fifth, the framing. This is a probabilistic discipline. A team that is never wrong is a team that only publishes the obvious. What matters is calibration — that things assessed as likely happen about as often as claimed — and that errors produce process improvements rather than defensiveness.

Blaming the sources or the stakeholder for the error is the answer that fails.