How to brief the next-round interviewer: turn first-round notes into exactly what to dig into
On this page
- What does "exactly what to dig into" look like?
- How do you find the dig-into items in your notes?
- A worked example: from raw notes to a brief
- What should you leave out so round two stays independent?
- Where should the brief live?
- How should the second-round interviewer use the brief in the room?
- Can AI pull the dig-into items from a transcript?
- Questions people ask
Reread your first-round notes against the role's must-haves and mark each one evidenced, thin or untested. For every thin or untested item, write one dig-into line for the next interviewer: what the candidate said that raised it, in their own words, the single question to ask, and what a good answer would add. Send three to five of these, ranked, with no score attached, after you have submitted your own scorecard. The second round then spends its time on what nobody knows yet instead of repeating the first.
This page is the method for finding those items in your notes and writing them so they can be used. The full handoff template, with logistics and the "do not repeat" list, is in interview handoff notes between rounds; use this page to fill its "still no evidence" and "claims to verify" sections well.
What does "exactly what to dig into" look like?
A dig-into item has three parts. If any part is missing, the next interviewer either ignores it or asks it badly.
| Part | What it contains | Weak version | Usable version |
|---|---|---|---|
| Trigger | What the candidate said or did not say, quoted, with a time if you have a transcript | "Leadership unclear" | "Described the rollout as 'we' throughout [18:40]; never said what he decided" |
| Question | One open question that tests it, not a list | "Ask about leadership" | "Walk me through one decision on that rollout that was yours alone, and what happened next." |
| Settled when | What answer would close the gap either way | (missing) | "He names a decision, the alternative he rejected and the result." |
The "settled when" line is the part most briefs leave out, and it is what stops the second interviewer from accepting another general answer. It also gives the debrief something to check: was the item settled or not?
How do you find the dig-into items in your notes?
Go through your notes, or the transcript if the candidate agreed to one, in five passes. Each pass looks for a different kind of gap.
- Coverage. List the must-haves for the role. Next to each, write the evidence you got, or "none". Any must-have with no evidence is an item, unless the next round is already assigned to it.
- Claims without a number or an owner. "Grew the account", "improved quality", "led the migration". Each needs a figure, a timeframe or the candidate's own part in it.
- "We" without "I". Answers told entirely in the plural describe the team. Note where you never heard what the candidate did personally.
- Answers that stopped early. The candidate described the situation and the task but never reached the result, or gave the result but not how they got there.
- Contradictions and hedges. A date that does not match the resume, a tool named in one answer and denied in another, "I'd have to check" on something they say they own.
Then sort what you found into three groups: settled (leave it out of the brief), thin (one more question would settle it) and untested (needs real time in the next round). Only the last two go in the brief.
What each kind of gap becomes
| What you see in your notes | Why it matters | Dig-into instruction |
|---|---|---|
| A must-have with no evidence | The debrief will have nothing to score | "Not covered: vendor negotiation. Ask for the last contract she renegotiated and what she gave up." |
| A result with no number | Impossible to compare with other candidates | "Said churn 'dropped a lot' [22:10]. Ask from what to what, over what period, and how it was measured." |
| "We" throughout | You may be hearing the team's work | "Ask which decision on the project was his and what he would change now." |
| Only one example of a skill | One story can be the exception | "Gave one strong escalation example. Ask for a second, different one, ideally one that went badly." |
| A hedge on something central | May be a gap or nerves | "Unsure how month-end close was split. Ask what she owned on day one of close." |
| A resume detail that does not match the call | Needs a neutral check, not an accusation | "Resume says 2021 to 2024; call said 'about two years'. Ask her to walk through the dates." |
More follow-up patterns, by type of answer, are in probing questions for interviewers.
A worked example: from raw notes to a brief
Invented example. The candidate, role and company below are fictional, and the timestamps are examples.
Round-one notes for Priya Raman (invented), Implementation Consultant, from the hiring manager's first interview:
- 6 yrs SaaS onboarding; current book ~25 mid-market clients
- "we got go-live down to about three weeks" [14:05]
- led "a couple" of data migrations, CSV and API [19:30]
- difficult client story: escalated to her manager, client stayed [26:50]
- SQL: "I can read it, I usually ask the data team" [31:15]
- asked about travel %; I said 10 to 20 (confirm with recruiter)Five passes later, the brief for the second interviewer, who owns the technical round:
DIG INTO (ranked), for round 2, Priya Raman (invented)
1. Data migration depth (must-have, thin)
Trigger: "led a couple" of migrations [19:30], no detail.
Ask: "Take the hardest migration you led. What broke,
and what did you do about it?"
Settled when: she names the source system, the volume,
one failure and her own fix.
2. SQL (must-have, untested)
Trigger: "I can read it, I usually ask the data team"
[31:15].
Ask: use the standard SQL exercise for this role.
Settled when: she writes or corrects the join herself.
3. Go-live time claim (thin)
Trigger: "we got go-live down to about three weeks"
[14:05].
Ask: from what, over how many clients, and what she
changed personally.
Settled when: before and after figures, and her part.
Not for you: travel expectations (recruiter will confirm).
Covered, no need to repeat: client escalation story.Notice what is not there: no rating, no "strong communicator", no "I'm not sure about her". The second interviewer knows where to spend the time and what would settle each point, and still forms their own view.
What should you leave out so round two stays independent?
- Your score and your verdict. An interviewer who reads "lean no" walks in looking for the reason. Submit your scorecard, then write the brief; why the order matters is in independent interview feedback before the debrief.
- Adjectives about the person. "Nervous", "arrogant", "polished" carry your impression without your evidence.
- Framing gaps as flaws. Write "untested" or "thin", not "weak". You do not know yet; that is the point of the next round.
- Personal details. Anything about health, family, age or other protected characteristics stays out, even if the candidate raised it.
Where should the brief live?
In the hiring record, where the next interviewer prepares, not in a chat thread. In Greenhouse, the scorecard has a "Note for Other Interviewers" field: Greenhouse says these notes are visible to other users scheduled to interview the candidate and appear under the Interview Prep tab of their interview kit and in the candidate's activity feed (Greenhouse interview kit guide, as of October 2026). Because those notes are visible to the hiring team, write them as you would any record a candidate could one day ask to see. In other systems, a note on the candidate record tagged for the next interviewer does the same job.
Send it the same day, while you can still check a quote, and early enough that the next interviewer reads it before preparing their questions.
How should the second-round interviewer use the brief in the room?
- Plan the time. Give the ranked items a slot in the round, usually the middle, after rapport and before the candidate's questions. Keep your own assigned topics too.
- Ask fresh, not "you said". Open each item with the open question, not "the hiring manager thought your answer was vague". The candidate should not feel cross-examined about the last round.
- Follow up until it is settled or clearly not. The "settled when" line tells you when to stop.
- Close the loop in writing. In your scorecard, add one line per item: "Item 1, migration depth: settled, named the system, the failure and her fix [quote]" or "not settled".
- Pass forward what is still open. If an item is still unsettled, it goes into your own brief for the next round, or to the debrief as an open question.
The debrief then reads item by item: which gaps were closed and by what evidence. The structure for that meeting is in the interview debrief template.
Can AI pull the dig-into items from a transcript?
It can draft them, if you give it the must-haves and the three-part shape, and you check every quote against the transcript before sending. A prompt that works with any approved AI tool:
From this interview transcript and these must-haves: [list],
list up to five items the next interviewer should dig into.
For each: the trigger (an exact quote with timestamp), one open
question, and what answer would settle it. Rank them.
Do not include ratings, opinions about the person, or anything
about age, health, family or other personal characteristics.
If a must-have was never discussed, say "not covered".Some interview tools draft part of this for you. Interview Signal (we make it) exports next-round questions alongside the scorecard; treat them as a draft for your "ask" lines and add the trigger and "settled when" yourself.
The worked example uses invented people and numbers.
Questions people ask
How can I use notes from a first-round interview to tell the second-round interviewer exactly what to dig into?
Reread your notes against the role's must-haves and mark each one as evidenced, thin or untested. For every thin or untested item, write one dig-into line: what the candidate said that raised it, in their words, the one question to ask, and what a good answer would add. Send three to five of these, ranked, with no score, after you have submitted your own scorecard.
How many things should I ask the next interviewer to dig into?
Three to five. A 45-minute round can usually test three open areas properly with follow-up questions. A list of ten turns the next round into a checklist and leaves no time for the interviewer's own assigned topics, so rank the items and say which one matters most.
Should I tell the second-round interviewer what I thought of the candidate?
No. Pass on what was said and what is still unknown, not your rating or adjectives about the person. A rating anchors the next interviewer, who then looks for evidence to match it. Ratings are compared in the debrief, after everyone has scored independently.
What makes a good dig-into item for the next interview?
It points to something specific the candidate said, gives one open question that tests it, and says what answer would settle it, for example the candidate's own decision, a number, or a second example. 'Probe leadership' is not specific enough; 'she said we cut onboarding to two weeks; ask what she personally changed and how she measured it' is.