# 転職面接で「採る人」と「見送る人」を分ける面接官の基準 - Show: 限界突破ライフハック - Published: 2026-10-19 - Canonical: https://podc.club/en/episode/1866434136_ca1e7a21c367.html - Audio: https://anchor.fm/s/10d527f20/podcast/play/127521889/https%3A%2F%2Fd3ctxlq1ktw2nl.cloudfront.net%2Fstaging%2F2026-9-15%2F434234383-44100-2-ed8b90fa1df21.mp3 - Source: Edited from a machine transcript of the first 30 minutes of the episode. Not an official summary. Tanaka Kei (田中圭), a recruiter with more than a decade of experience, breaks down exactly what separates a hire from a pass in mid-career interviews. He covers how he detects genuine self-starters, the reality of reference checks and SNS screening, the pitfalls of putting numbers on a resume, and the hidden traps of industry jargon. Every technique comes with the concrete methods and real examples he has used across hundreds of individual interviews. ## What this episode answers ### How does an interviewer detect a candidate's ability to self-direct? He asks about the boss's most recent feedback, the work that excited the candidate most, and then breaks down their role phase by phase; differences in narrative resolution and vocal tone reveal when someone is skating over tasks they didn't personally own. He also deliberately challenges their trophy story and throws out unfamiliar topics to see how they respond. ### What mistakes do people most commonly make when putting numbers on a resume? Numbers must always be paired with both an absolute figure and a ratio such as year-on-year, so the reader can judge context — 'grew revenue by ¥300M' means nothing if the reader doesn't know it occurred at a Recruit-scale operation. ### How do job-change frequency and SNS activity affect the screening process? High job-change frequency triggers thorough reference checks; SNS accounts are checked in full after every interview, and multiple candidates have been passed over based on what those accounts revealed. ### What does an interviewer not care about, and what are they actually watching closely? For mid-career hires, education, nerves, and certifications are ignored entirely; job-change frequency, references, and SNS are taken seriously. ### How do industry-specific protocols and jargon function in an interview? At publishing companies 'Who would you most want to meet right now?' implicitly requires a specific editorial pitch — not knowing that code gets you screened out; 'LTV' means lifetime value in startup contexts but loan-to-value in finance, and some interviewers deliberately saturate the conversation with jargon to test real comprehension. ## Article ## New Graduates vs. Mid-Career: Completely Different Criteria Tanaka Kei has spent more than a decade in recruiting, screening roughly 100 candidates per year in group formats and 200–300 new graduates in one-on-one sessions, plus 50–100 mid-career candidates annually. For new-graduate hiring he prioritizes potential, grit, and culture fit, treating each person as a blank slate. Mid-career hiring flips the equation: the first question is always whether a specific, proven skill can make an immediate contribution. Culture fit still matters in both cases, but Tanaka is unequivocal — anyone who needs to be told what to do, rather than rolling the ball themselves and running the project independently, will end up miserable, and so will the company. A mindset of "I'm here to learn and grow" leaks through in interviews even when candidates never say it out loud, and drawing it out is exactly what the interview is for. ## Detecting Self-Direction and Actual Role Tanaka uses several techniques. **Ask about the boss's most recent feedback.** Someone told "please share more information" probably really does need to share more; someone told "feel free to handle it however you like" probably lacks initiative. Recent feedback is hard to fabricate. **Ask about the work that excited them most — not their biggest win.** When people are genuinely fired up, they stretch to fill their role, and that's when the resolution of their story reveals the truth. Small projects or personal anecdotes are fine answers and often more honest. **Decompose the role phase by phase.** A sales deal might run: lead generation → proposal writing → board approval → execution. Tanaka asks exactly which phase the candidate owned. Most people describe everything as though they handled it all, but there is a world of difference between drafting a rough version of a proposal and seeing a finished version through to approval. When candidates drift into phases they didn't own, a kind of surface-skimming appears in how they talk. It is the difference between someone who says they love cooking but only handles pre-prepped ingredients and someone who can fillet a whole fish — the resolution of the explanation gives them away. **Push back deliberately.** He challenges a candidate's trophy story with something like "but isn't that really just…?" — not to intimidate, but to observe stress tolerance and counter-argument patterns in a short window. Some candidates go completely flat within seconds. **Introduce an unfamiliar topic.** He watches whether "I don't know" becomes a dead end or a starting point: can they hypothesize from adjacent knowledge and keep the conversation going? Casual lunch or dinner interviews are also common for mid-career screening; with the armor off, genuine strengths surface — but so do boundary issues. Notably, Tanaka ensures that interviewers on successive panels do not share evaluations with each other beforehand: one person saying "great candidate coming your way" is enough to skew the next interviewer's perception. ## What He Cares About — and What He Doesn't **Doesn't care about:** education (for mid-career hires he often doesn't even know which school they attended), nerves, certifications. **Does care about:** job-change frequency. Someone on their eighth company at age 33 is treated as a potential red flag and triggers intensive reference checks. Issues such as interpersonal problems, a pattern of quitting out of boredom, or coordinated group departures can surface through references. Those references need to come from people who will be genuinely honest — a bland "I think they're probably fine" tells you nothing. **SNS:** Tanaka checked the accounts of two staff members present at the recording session after it ended. People who believe they are anonymous often are not; IDs can be matched across platforms by determined searchers. Content that seems trivial may still lead to a pass; content that looks positive may conceal other problems. Multiple candidates have been turned down based on their social media. ## How to Write a Resume and Messages Overly long resumes and DMs are "violence in the name of thoroughness." The intent is appreciated, but the reader's burden must come first. The single most important rule: **pair every number with its context**. Always give both the absolute figure and a ratio such as year-on-year. "Grew revenue by ¥300M" is unreadable without knowing whether that happened inside a Recruit-scale operation. "Attended 53 job interviews" only sounds like a lot if the reader knows the average is 3–4. Calibrating the comparison to the reader's frame of reference also helps — analogies such as "equivalent to Switzerland's GDP" or "a sub-3-hour marathon finish is the golf equivalent of a single-digit handicap" can make scale land immediately. ## The Trap of Industry Protocols and Jargon At a publishing company, "Who would you most want to meet right now?" is really asking "and what editorial pitch would you bring them?" Answering "my grandfather back home" ends the conversation. Candidates who don't know the industry's implicit Code A fail the litmus test before they realize one was running. Jargon is another minefield. "LTV" means lifetime value in a startup context and loan-to-value in finance. Some interviewers deliberately pile on jargon to gauge real comprehension. Faking expertise invites a deep dive that exposes the bluff. The preparation that works is being able to explain the same concept two ways: one version for insiders, one for outsiders. ## Takeaways - Asking 'what work excited you most' and 'what's your boss's latest feedback' makes it hard to bluff, and walking through each project phase one by one exposes candidates who are glossing over work they didn't actually do. - Resume numbers must always pair an absolute value with a ratio and enough context for the reader to judge impact — a figure that looks impressive at one company may be routine at another. - SNS activity, job-change frequency, and references are taken seriously in mid-career screening; educational background is almost never a factor.