Why Interviews Go Wrong

Interviewers talk about difficult subjects. Most bad interviews are not caused by difficult subjects. They are caused by preventable decisions, and the pattern of failure is consistent enough to be worth setting out.

Over-preparation is more common than under

A well-prepared interviewer knows the material thoroughly and has a list of questions. The failure mode is treating the list as a schedule to complete rather than a resource to draw on.

This produces the recognisable pattern where a subject says something genuinely interesting and the interviewer moves to the next prepared question. The moment is gone and cannot be recovered, because the subject has learned that unusual answers do not get followed.

The corrective is holding the preparation loosely. Know enough to recognise when something surprising has been said, and be willing to abandon the plan when it happens.

Question order carries more weight than question quality

The first ten minutes establish what kind of conversation this will be. Opening with the most difficult question tends to close a subject down for the duration.

That does not mean avoiding difficulty. It means sequencing so that difficulty arrives after some rapport exists, and after the subject has experienced being listened to accurately.

The corresponding error is leaving the hard question so late it gets rushed. Experienced interviewers tend to place it around two-thirds through — late enough for trust, early enough to follow up properly.

The subject has done this before

Anyone promoting something has answered similar questions many times and has developed reliable responses. These are polished, quotable and empty.

A question phrased in the expected way will reliably retrieve the prepared answer. The only escape is asking something the subject has not been asked, which requires either unusual research or unusual attention.

The most effective technique is frequently the least sophisticated: asking about process rather than meaning. People are rehearsed on what their work is about and rarely on how a specific decision got made, and process answers tend to be both more specific and more revealing.

Silence is under-used

The most common technical fault is filling pauses. A subject who stops talking has often not finished thinking, and an interviewer who waits will frequently receive a second, better answer.

This is uncomfortable and it works consistently. Interviewers who train themselves to wait three or four seconds past comfort report it as the single largest improvement they made.

Sympathy and its limits

Interviewers who make subjects comfortable get more material. This is true and it creates an obvious hazard: a subject who feels warmly towards the interviewer may say things they would not have said, and the interviewer then holds material obtained through a relationship the subject may have misread.

There is no clean resolution. The defensible position is that the subject knows they are being interviewed, the terms were clear, and warmth is not deception. But the discomfort is real and interviewers who claim otherwise are usually not thinking about it carefully.

What to do when it is going badly

Several things help. Naming it sometimes works — acknowledging that the conversation is not going anywhere occasionally resets it, because the subject is usually aware too.

Changing physical activity helps more than it should. Standing up, moving, doing something with hands changes register measurably.

And abandoning the topic entirely is frequently better than persisting. A question that is not working will usually not start working, and the material may be reachable from an entirely different direction later.

The interviews that fail irrecoverably are almost always the ones where the interviewer kept pushing the same door.