The shukatsu axis: deriving it from your history instead of picking it from a list

The axis (就活の軸) question is a coherence check on your decision-making: does this student choose companies by a stable criterion, or by brand and momentum? A candidate whose axis genuinely fits the company is a lower attrition risk; a candidate with no axis will accept on vibes and quit on vibes — and interviewers have watched that movie.

It is also a consistency trap by design: your axis must explain this application. An axis of "work that directly reaches consumers" at a B2B infrastructure company creates a contradiction the interviewer will name. Before any interview, run the check: does my stated axis predict me sitting in this chair?

List-site axes — "growth environment", "impact on society", "good people" — fail because they are conclusions without derivations; anyone can hold them, so they distinguish no one and predict nothing. The workable method runs backward from your history: list the moments you were most absorbed (in study, gakuchika, part-time work), the choices you made repeatedly, and what the absorbing moments share. The shared property, named concretely, is an axis with a built-in evidence trail.

"Environments where I can see the numbers move because of my own work" — derived from a sales part-time job episode — is an axis that answers its own "why" and survives follow-ups. Aim for two or three axes: one about the work itself, one about the environment or people, optionally one boundary condition. One axis is thin; five is indecision wearing a costume.

Your real axis set legitimately includes conditions you should not lead with in interviews: salary floor, location, remote policy, stability. The two-layer solution: the full honest set drives your actual application decisions — this is jiko-bunseki paying rent — while the interview set is the subset about work and environment, which is not deception but selection, the same way every answer in an interview is.

If asked directly about conditions ("is compensation part of your axis?"), honesty in proportion works: acknowledge conditions exist as boundary criteria, then return weight to the work axes. Pretending money does not matter reads as either naive or dishonest — interviewers have salaries too.

Three standard probes. "Why that axis?" — this is where derivation pays; the episode behind the axis is the answer. "Then why us, and not competitor X who also fits?" — requires one company-specific fact layered on top of the axis; the axis narrows to an industry, the specific fact picks the company. "What if the work here doesn't match your axis sometimes?" — the mature answer acknowledges every job includes off-axis work and states the ratio or trajectory you care about.

The killer inconsistency is between rounds: an axis that mutates between the ES, first interview, and final reads as answers optimized per audience — the exact opposite of what an axis claims to be. One set, told consistently, evidenced differently.

The underused half of the axis: when multiple naitei arrive, the honest axis set you wrote in month one is the comparison rubric that keeps the decision from collapsing into brand and salary alone.

InterviewTrail AI keeps your axes, episodes, and every company's facts in one place — so both your interview answers and your final decision draw on the same coherent record.