Interview Notes Template: Compare Candidates Fairly (2026)

Interviewing three candidates is easy to keep straight. Interviewing fifteen in a week — phone screens stacked between panel loops — is how "the Tuesday backend person, the one with the migration story" becomes a blur by Friday's debrief. The fix is a consistent interview notes template: the same sections for every candidate, evidence instead of impressions, and a rating you can defend. This post gives you the template, variants for phone screens and panels, and the AI-assisted capture workflow that lets you stay in the conversation instead of typing through it. (Looking for general meeting documentation instead? See the meeting notes template.)

Why a template beats a blank page

• Fair comparison: if every candidate is assessed in the same structure against the same criteria, Friday's debrief compares evidence — not whoever interviewed most recently or told the best story.
• Defensible decisions: notes that record job-related facts against published criteria protect everyone. Notes that record vibes ("great energy!", "not sure about fit") help no one and age badly.
• Less interviewer drift: a template with the competencies pre-filled is a quiet script — it pulls the conversation back to what the loop agreed to assess.

The interview notes template

Copy this and fill the brackets before the interview starts:

Candidate: [name]  |  Role: [role + level]  |  Stage: [screen / technical / panel / final]
Interviewer: [you]  |  Date: [date]  |  Duration: [length]

Competency 1: [e.g. role-specific skills]
Evidence: [what they said/did — specifics, numbers, quotes]
Rating: [1–4 on your scale] — [one line on why]

Competency 2: [e.g. problem-solving]
Evidence: / Rating:

Competency 3: [e.g. collaboration]
Evidence: / Rating:

Candidate's questions: [what they asked — it shows what they care about]
Flags to verify: [claims to check in references or later rounds — not accusations, just open items]
Recommendation: [strong yes / yes / no / strong no] — [the one-sentence case]
Next step: [advance to X / decline with reason / needs another signal on Y]

Three rules make it work: evidence before rating (force yourself to write what happened before scoring it), quotes over paraphrase when something is load-bearing, and job-related lines only — write every note as if the candidate might read it, because in a dispute they might.

Variants by interview stage

• Phone screen (15–30 min): collapse to logistics (notice period, location, compensation range), one or two competencies, motivation, and a clear advance/decline. Speed matters more than depth here.
• Technical interview: one competency block per exercise; capture the approach, not just the result — "asked about edge cases before coding" is stronger evidence than "solved it."
• Panel loop: same template per interviewer, filled independently before the debrief — independent ratings first, discussion second, or the loudest voice sets everyone's score. If your team uses a shared interview scorecard or evaluation form, the competency blocks above drop straight into it.

The capture problem: 15 candidates, one keyboard

The template solves structure; it doesn't solve the fact that typing through an interview costs you the interview. Candidates hear the keyboard, you miss the follow-up cue, and the notes still end up thinner than the conversation deserved. The AI-era fix is separating capture from judgment:

1. Ask consent, then record. "I'd like to transcribe this so I can focus on the conversation rather than my keyboard — is that okay?" Recording laws vary and candidates deserve the choice; almost everyone says yes to that framing.
2. Let the interview transcribe itself. Vowen captures the call audio on your own computer — no notetaker bot appearing in the candidate's participant list — and transcribes it on-device with speaker labels, which also means an interview full of personal data stays on your machine rather than a third-party cloud. It works the same for Zoom, Meet, or Teams interviews, and for in-person interviews via the microphone.
3. Fill the template from the summary. After the call you get the transcript plus a summary with the key points and action items; your job shrinks to dropping evidence into the competency blocks and making the judgment calls — the part that was never automatable anyway.
4. Debrief from evidence. When someone asks "did she actually say she led the migration, or joined it?", the transcript answers in seconds — per candidate, in the candidate's own words.

The judgment stays yours; the AI just guarantees the raw material is complete. For how this compares across tools, see the AI note taker roundup — and if you're transcribing research interviews rather than hiring ones, the interview transcription page covers that workflow.

The bottom line

Fifteen candidates a week is only unmanageable when every interview starts from a blank page and ends with reconstruction from memory. A fixed template makes candidates comparable; evidence-first notes make decisions defensible; and consent-based, on-device capture means you can actually be present in the conversation. Vowen's free tier is enough to try the capture workflow on your next screen — your Friday debrief will read differently.