How does a candidate claim their part of a team win without over-claiming a colleague's work?

## Two failures, one boundary There are two ways to get attribution wrong and they look like opposites: - **Hiding.** The whole win narrated in the plural. Nothing individually assessable survives, and the interviewer either asks the attribution follow-up or scores conservatively. - **Over-claiming.** A team's work narrated as one person's. This is the more expensive failure, because it is not a scoring gap — it is a credibility judgment, and a panel that suspects it will discount your other answers too. Both are solved by the same move: **state the boundary explicitly**, rather than leaving the listener to infer it. ## How to draw it For each story, decide in advance which of three buckets each element falls in. | Bucket | Pronoun | Example | |---|---|---| | Genuinely shared | plural | The team owned the retraining pipeline behind the fraud-scoring model. | | Mine | first person, concrete verb | I traced the feature-freshness step exiting early without failing. | | Someone else's, decisively | named by role | The on-call engineer caught the alert and made the rollback-first call. | The third row is the one candidates skip, and it is the one that does the work. Attributing a real contribution to a named role costs you nothing you were entitled to and buys the listener a reason to believe row two. Interviewers are calibrated on this: they hear dozens of stories where everything good was the candidate's idea, and a story with a visible boundary stands out as the credible one. ## Signals that you have crossed into over-claiming - Every good decision in the story is yours, and every problem came from outside the team. - You use scope verbs — 'owned', 'drove', 'led' — without a single artefact or decision underneath them. - The story has no other people in it who did anything, only people who needed convincing. - You cannot say what you would have gotten wrong without a specific colleague. Any one of these is survivable; together they read as a candidate who does not know where their own contribution ends, which is a real risk to hire at senior level. ## Signals that you are still hiding - The interviewer has to ask what your specific contribution was. - Your first-person sentences are all hedges — helped, supported, was involved in. - The most interesting decision in the story has no owner. ## Claiming influence honestly At more senior scope, much of a real contribution is judgment, sequencing and persuasion rather than artefacts, and that is genuinely harder to attribute. The technique is to name the act, not the influence: not 'I influenced the direction', but 'I argued for rolling back before debugging, against the preference in the room, because stale scores were already live — and that is the call we went with.' The act is checkable in the same way writing code is: it happened at a moment, it had an alternative, it changed what the team did next. When you led through others, keep both halves visible: the engineers did the engineering, and your acts were the review you insisted on, the scope you cut, the risk you escalated. A senior candidate who says 'my team built it; what I did was refuse the first design and pay for a second week' has given the interviewer something to grade without taking anything from anybody. ## Rehearsing the boundary The boundary is easiest to place while editing rather than while speaking. Take the transcript of the story as you actually tell it, mark every plural pronoun, and sort each into shared context (keep) or an act of yours (rewrite in the first person). Then do the reverse sweep for over-claiming: for every first-person verb, ask whether a teammate who was in the room would agree that you did that. Anything that fails that test gets re-attributed by role. A story that survives both sweeps is safe to tell to anyone who was there — which is the only durable standard, since some interviewers will have worked with people who were.