How does a candidate turn a hiring manager's spoken answers into a confirmed 90-day expectation?

## The problem this technique solves A hiring manager will tell you, out loud, what your first 90 days are for. Two things then happen. Your memory of it degrades into a paraphrase within about a day, and the manager's own version is never written anywhere either. Weeks later, when you are deciding between roles or negotiating start conditions or actually doing the job, both of you are working from remembered impressions of the same sentence. The fix is small and almost nobody does it: capture, read back, restate in writing. ## Step one: capture the words, not the gist While the manager is answering, write down the phrases they used. Not "wants me to own releases" but the actual construction — for a mobile app team, something like *"own the release train, two releases out with no rollback, and get the manual regression pass out of the critical path."* Their nouns matter. "Own" and "help with" are different mandates. "The release train" and "the release process" may be different scopes at that company. Taking notes in a conversation is normal and reads as attentive, not clerical, provided you keep looking up. If notes would be awkward, write the sentence down in the first two minutes afterwards, before anything else. ## Step two: pair the expectation with the challenge An expectation without its context is half a picture. The companion question is: *"What is the biggest challenge facing this team over the next two quarters?"* Asked of the hiring manager, it produces the constraint that your 90-day expectation has to survive. The pair is diagnostic. If the manager's stated first-quarter deliverable has nothing to do with the challenge they just named, one of two things is true: the deliverable is not actually the priority, or the challenge belongs to someone else. Either is worth one more question in the room, phrased as curiosity rather than cross-examination — how the two relate, or where the new hire fits against the challenge. ## Step three: the read-back Before the conversation ends, compress both answers into one sentence and hand it back: *"So the first 90 days are about owning the release train and getting the regression pass off the critical path, and the thing that could make that hard is the pipeline migration — have I got that right?"* Three things happen at once. - **Correction is cheap.** If you misheard the scope, the manager fixes it in ten seconds. Discovering the same gap in month two costs a quarter. - **You demonstrate the habit.** A read-back is the operating behaviour of someone who closes ambiguity early. Managers notice it, and it is one of the few candidate behaviours that is simultaneously useful to you and legible as a signal. - **The expectation becomes confirmed rather than merely stated.** The manager has now said yes to a specific sentence. Keep it to one sentence. A read-back that recites six bullet points stops being a check and starts being a performance. ## Step four: put the same sentence in writing When you next exchange anything written with the manager or the process, use the same sentence, in the same words. Not as a demand and not as a clause — simply as the framing of what you understood the role to be. That gives both of you a shared artifact to point at later, which is exactly what does not exist by default. Be honest with yourself about what this is and is not. A confirmed sentence is not a contract, and treating it as one is a mistake in both directions: role expectations legitimately shift with priorities, and whatever any of this means formally varies by employer and by where you are, so do not describe it to anyone as a commitment you can hold them to. Its value is clarity and shared memory, not enforcement. ## Using divergence as data Because the sentence is now fixed, you can compare it against everything else you hear. If the manager's 90-day expectation shifts between the first conversation and the last, or if the team describes a different first quarter than the manager did, you have learned something you could not otherwise have detected — remembered paraphrases are too soft to disagree with each other. The right move on a divergence is to ask, not to conclude. "Earlier I understood the first quarter as X — has that changed, or was I hearing two parts of the same thing?" is a question a manager can answer easily, and the answer distinguishes a role that is still being shaped from a simple difference in emphasis. ## When to skip it If the manager's answer was genuinely abstract, there is nothing to read back and pretending otherwise manufactures a precision that does not exist. In that case the honest read-back is the absence: you understood the first quarter to be open, and you can ask directly whether that is deliberate.