Why does dropping a company's own vocabulary into an unchanged interview story backfire?

## The failure in one line Vocabulary mirroring fails when the words change and the evidence does not. The candidate reads the posting, extracts its distinctive phrases, and sprinkles them over stories that are word-for-word what they told the previous three companies. What reaches the listener is their own language with nothing new underneath it. ## Why it is so easy to spot Three reasons, and they compound. **The listener knows the source text.** A recruiter has read that posting many times and has heard candidates quote it all week. A phrase from it arrives pre-flagged. **The register breaks.** Borrowed phrases sit at a different altitude than the rest of your speech. You spend two minutes in concrete operational detail and then produce one polished marketing sentence. The seam is audible. **The story does not corroborate.** This is the decisive one. The claim is *I ship small and often*; the story that follows describes one big launch at the end of a quarter. Nothing was mirrored except the noun phrase. ## What honest mirroring looks like Mirror the **priority**, and let your own words carry it. Take the line `we ship small and often and expect engineers to talk to users` and a frontend redesign you actually did. - **Flattery version:** *I really align with shipping small and often, and I'm very user-focused.* Followed by the untailored story. - **Evidence version:** *We had a redesign that would have been a three-month branch, so we cut it into slices that each went out behind a toggle. The second slice went out to a small share of traffic first. I sat in on three user sessions after it landed, and what I watched changed how we bounded the next slice.* The second version never uses the posting's phrasing. It does not need to: it demonstrates the thing the phrasing points at, which is what the phrasing was screening for. ## The one legitimate use of their words Their vocabulary may **introduce** your evidence; it may never **replace** it. *You mentioned shipping in small slices — the closest thing I've done to that is this frontend redesign.* That is orientation. It tells the listener which of your stories is about to arrive and why you chose it, then immediately hands over to your own material. What it does not do is assert a trait and stop. ## The value you cannot evidence Sometimes the working-style paragraph names something your record does not reach — say the posting is built around direct user contact and every product you have worked on had a research team in between. The wrong move is to claim it anyway with a stretched anecdote. It fails on the second question, and it costs you the credibility of the stories that were true. The right move is a two-part answer: the nearest real thing, and an honest statement of the gap. *I have not sat with users directly. What I have done is take session recordings and support tickets as the input to what I built next — here is one case. Direct contact is one of the reasons this role interests me.* That answer is safe under drill-down, and it converts a mismatch into a reason you are applying. ## Why this matters beyond one answer The cost is not the sentence. It is the discount rate the listener applies afterwards. Once a claim has been caught floating free of evidence, every subsequent claim gets weighed against the possibility that it too is decoration. Candidates lose loops this way while believing they were building rapport. ## A quick self-check Before a conversation, take each promoted story and ask: *if I deleted every word I took from the posting, would the story still show what they said they care about?* If yes, the tailoring is real. If the story goes generic the moment their words come out, the tailoring was only vocabulary, and the fix is to change which beats you expand, not which adjectives you use.