What does asking three interviewers the same question reveal that one answer cannot?

## The problem with a single answer Everything you learn as a candidate arrives through one person at a time, and each of them sees a different slice: a peer engineer knows what this quarter felt like, a lead knows what was promised, a skip-level leader knows what was intended and what is coming. Ask one of them how work gets prioritised and you learn their vantage point, not the team's reality. Triangulation is the cheapest correction available: ask **one** question — the one that most affects your decision — of **three** interviewers across the loop, and compare. ## The artefact The mechanism is a notes column: one row for the repeated question, one column per interviewer, each cell filled in immediately after that round with what the person actually said rather than what you concluded. Doing it in the moment matters, because after a full loop the answers blur into a single impression and the differences — which are the whole point — are the first thing to disappear. Record phrasing, whether an example appeared, and whether anything was declined. ## Reading the row **Consistent and each specific.** Three people describe the same mechanism and each contributes a different concrete instance. This is the strongest signal available to an outsider, because independent detail is hard to coordinate. Treat it as credible and move your attention to a different question. **Consistent and identically worded.** The same sentence three times, with no examples underneath, is a talking point, not a practice. It is weak evidence either way: the practice may be real and well-communicated, or the phrasing may be all there is. Probe by asking the third person for an example rather than for the description again. **Contradictory on facts.** The skip-level leader says the platform roadmap is set for the coming quarters; the engineer who would be your peer says priorities get re-cut whenever a stakeholder escalates. Both are usually being honest. The leader is describing intent from where intent is formed; the peer is describing what reached the desk. The gap between those two is often the most useful thing you learn in the whole loop, and it is invisible unless you asked both. **One hedge among clear answers.** Two people answer easily and the third produces a general non-answer. That is worth noting without over-reading: they may be outside the subject, newer than they appear, or unwilling. Which of the three hedged is the informative part. ## What to do with a contradiction Raise it once, and raise it well. Usually the right moment is late in the loop with the person best placed to reconcile it — often the skip-level leader — and the right form names the gap without naming anyone against anyone. Something like: *I have heard two pictures of how priorities get set — one where the quarter is planned, one where it re-cuts when a stakeholder escalates. Can you help me square those?* A leader who can explain the mechanism that produces both — planning at one horizon, interrupts at another, with a stated limit on how much gets re-cut — has answered you. A leader who denies that the second picture happens, when a person doing the work just described it, has told you something about how information travels upward. Do the same for the timing. Asking the same question twice in one round is odd; asking it once per round across a loop is normal and reads as thoroughness. ## Costs and limits You will not have room for many repeated questions — the end-of-round window is short and each interviewer deserves at least one question aimed at what they specifically know. In practice one repeated question per loop is the realistic budget, so choose the one whose answer would actually change your decision. And remember that divergence is not automatically a defect: different vantage points *should* produce different emphases, and only a factual contradiction, or a hedge where others were free, is worth spending your reconciliation question on. ## The failure mode The pattern fails in the same way every hedge does. A candidate notices in the fourth round that the answer does not match the second round's, decides not to make it awkward, and moves on to the next prepared question — and then reconstructs the discrepancy a week later with no notes and no chance to ask. The gap is only worth collecting if you are willing to spend one question on it while you still can.