When an interviewer answers your end-of-round question, what tells you it was a hedge rather than a real answer?

## What a hedge is A hedge is an answer you cannot check. Everything in it is true of almost any employer, so it moves you no closer to knowing what a Tuesday on this team looks like. The useful test is mechanical: **after the answer ends, is there anything in it I could ask a second question about?** A date, a decision, a person's role, a thing that broke, a piece of work that was stopped. If there is nothing to grab, you were hedged — regardless of how pleasant the answer felt. ## The four shapes | Shape | What it sounds like | What it dodged | |---|---|---| | Abstraction | Principles, values, adjectives | Any single instance | | Tense drift | What the team is *going* to do | What it actually did | | Scope swap | The company answered instead of the team | The team you would join | | Volume | Long, warm, unsummarisable | The question's actual noun | Volume is the one candidates miss most. A senior person who talks fluently for two minutes creates a strong impression of having answered. Try to write the answer down in one line afterwards; if you cannot, it was volume. ## Why hedges happen (and why one is not a verdict) Most hedges are not concealment. Common, boring causes: the interviewer does not know — a peer on the panel often genuinely cannot describe how the roadmap is set; the interviewer is repeating recruiting language they were given; the subject is a person's departure and, as is common practice at many employers, nobody says much about an individual leaving; or the question was broad enough that abstraction was the only honest answer available. That is exactly why a single hedge earns a follow-up rather than a conclusion. The signal you are actually collecting is what happens on the *second* pass. ## Worked example Final round for a role on a data platform team. The candidate asks the skip-level leader how the team decides what it works on next quarter. > *We're very outcome-driven here. Roadmaps are collaborative — we want the engineers close to the business, not taking tickets.* No episode, no actor, no moment: abstraction plus tense drift. Nothing in it distinguishes this team from any other. The candidate spends one follow-up: > *What is the most recent thing the team stopped working on, and who made that call?* Now one of two things happens. Either the answer lands — the self-serve reporting work was cut partway through the last cycle, and the leader names who agreed to cut it — and the candidate has a real fact plus a thread to pull. Or the second answer is abstract too, and *that* is the data point: the hedge survived a direct request for a single instance, from the person most likely to know. ## Write it down before it decays The habit that makes all of this usable is a notes column: one row per question you asked, one column per interviewer, filled in within a few minutes of the round while the wording is still fresh. Record what you got, not what you inferred — 'no example, moved to company values' is a more useful note than 'seemed evasive'. Across a loop, those cells are what let you compare one interviewer's answer with another's instead of relying on a mood you cannot reconstruct a week later. ## What a substantive answer sounds like Specific, recent, and slightly unflattering. Real answers usually contain a cost: something that was traded away, an argument that took two attempts to settle, a piece of work nobody enjoyed. An answer with no cost in it anywhere is either a hedge or a very well-rehearsed pitch, and both deserve the same follow-up. ## The failure mode The common mistake is not being fooled — it is being polite. The candidate hears the non-answer, registers vaguely that it was thin, nods, and moves straight to the next prepared question because the prepared list feels like the thing to get through. The list is not the point; the answers are. A prepared question that produced nothing is unfinished work, and finishing it costs one sentence.