Capability template for recruiting and hiring
Interview process review after a hiring round
After a hiring round closes, the recruiting lead gets a review of the interview process: questions that gave no usable evidence, competencies nobody covered, inconsistent rubric use, and revisions to approve before the next search. It reviews the process only; no candidate is assessed.
Listed under
Add it to a workflow in the Fully Powered Planner, beside the teams and roles that will run it. Book a fit call
The outcome
The next search starts from what this round taught you instead of from the same interview plan. You see which questions gave interviewers nothing they could write down, which questions were asked twice, which parts of the agreed rubric were never put to anyone, and where interviewers used the rubric differently from one another.
Each finding comes with a proposed change, written as a draft you can accept, edit or drop: a reworded question, a question removed, a question added for an uncovered competency, a change to the feedback form. The recruiting lead decides which revisions go into the next plan. Nothing changes until then.
The review is about the process. It does not say who the strongest candidate was, does not revisit any outcome, and does not compare interviewers as people. It reports by question, by competency and by interview.
Who starts the work, and when
The recruiting lead, or a recruiter they ask, starts it once a hiring round has closed and the feedback is in, ideally before the interview plan is reused for a comparable search. It suits any role you will hire for again, such as a team lead you recruit every year or a position you fill repeatedly.
The recruiting lead reads the draft, checks the notes the assistant judged unusable, and approves, edits or drops each proposed revision. The AI assistant sends nothing and changes no plan, form or record.
What they supply
- The agreed rubric for the role: each competency, with what good evidence looks like.
- The interview plan the round used: which question was asked in which interview, and which competency it was meant to cover.
- The panel feedback with candidates and interviewers anonymised: for each question asked, the rating given and the note of what the candidate said or did.
- A short note of any process issues people raised during the round, such as a late start or a missing briefing.
- A few settings: the role, the date the round closed, the rating scale and who reads the review.
What comes out
One review in a fixed order. It opens with where the files came from and a headline count. Then the questions that gave little usable evidence, the questions repeated across interviews, the coverage of every competency in the rubric, and how the rubric was used: ratings given without a usable note, and ratings for the same competency that sit far apart.
It goes on to the questions that were not asked, each process issue people raised checked against the feedback where the files allow, the list of notes judged unusable so you can check the judgement yourself, and the proposed revisions. Each revision names the finding behind it and waits for the recruiting lead's decision. If a header in a file does not match, the review stops and asks rather than guessing.
An invented example
Illustrative example with invented data. The company, the role, the interviews, every question and every word of the feedback below are invented. The company, called Company K here, is an invented 45-person firm that makes booking software for restaurants. It has just closed a round for a Customer Operations Manager, with three candidates interviewed. The recruiting lead asked an operations coordinator to prepare the review. Candidates and interviewers appear only as codes (K1 to K3 and P1, P2, P4).
The inputs come first, in full, so the template can be run on them. The review follows.
The inputs
1. Settings
Role: Customer Operations Manager. Round closed: Friday 6 November 2026.
Review date: Tuesday 10 November 2026.
For: the recruiting lead. Prepared by: operations coordinator.
Rating scale: 1 to 4.
Thresholds: use the template's defaults.
2. Agreed rubric
criterion_id,criterion,what_good_evidence_looks_like
C1,Handles escalations and fixes causes,"A specific escalation the person led, what they did first, and what they changed afterwards"
C2,Plans team cover,"How cover was built from volume, and a case where it fell short"
C3,Coaches individuals,"A kind of coaching given to a named type of person, and what changed"
C4,Uses data to decide,"A metric they tracked, what it showed and the action they took"
C5,Works with product and engineering,"A joint piece of work with dates and who agreed"
C6,Writes clear customer guidance,"A piece of customer-facing writing they produced and what they removed or changed"
3. Interview plan
interview_id,interview_name,question_id,question,criterion_id
I1,Hiring manager conversation,Q1,"Tell us about the last serious escalation you handled. What did you do first, and what changed afterwards?",C1
I1,Hiring manager conversation,Q2,How do you coach people on your team?,C3
I1,Hiring manager conversation,Q3,How do you plan cover for your team?,C2
I2,Operations panel,Q3,How do you plan cover for your team?,C2
I2,Operations panel,Q4,"Which support metric do you watch most closely, and what did you do the last time it moved?",C4
I2,Operations panel,Q5,How comfortable are you with numbers?,C4
I3,Cross-team conversation,Q6,How do you work with product and engineering?,C5
I3,Cross-team conversation,Q7,Describe a repeat complaint you traced to its cause.,C1
4. Panel feedback (anonymised)
candidate_ref,interview_id,interviewer_ref,question_id,rating,evidence_note
K1,I1,P1,Q1,3,"Described a billing outage; paged the on-call engineer within ten minutes, wrote the customer update personally and later got the retry rule changed."
K1,I1,P1,Q2,3,Strong coaching instincts.
K1,I1,P1,Q3,2,"Builds the rota from ticket volume by hour; had a gap on Sundays that was covered personally."
K1,I2,P2,Q3,3,Same rota example as the first interview; nothing new.
K1,I2,P2,Q4,4,"Cut the reopen rate from 14% to 9% by splitting two macros; checks it weekly."
K1,I2,P2,Q5,3,Good with numbers.
K1,I3,P4,Q6,3,"Worked with product to retire a confusing setting; gave the dates and who agreed."
K1,I3,P4,Q7,4,"Traced repeat refund complaints to one unclear policy line and got it reworded within a week."
K2,I1,P1,Q1,3,"Walked through a delayed-shipment complaint; phoned the customer, then asked the warehouse to change the label check."
K2,I1,P1,Q2,3,Seemed keen on developing people.
K2,I1,P1,Q3,3,"Uses last quarter's ticket counts to plan cover; no example of a shortfall."
K2,I2,P2,Q3,,NOT ASKED
K2,I2,P2,Q4,2,"Tracks first-reply time in a spreadsheet; could not say what was done when it rose."
K2,I2,P2,Q5,2,Seemed unsure.
K2,I3,P4,Q6,2,Said engineers are easy to work with.
K2,I3,P4,Q7,3,"Found a repeat complaint came from a policy wording and got it changed."
K3,I1,P1,Q1,4,"Described a data-loss complaint; worked with the customer through the weekend and set up a checklist afterwards."
K3,I1,P1,Q2,,NOT ASKED
K3,I1,P1,Q3,2,"Plans cover with a shared calendar and asks people to swap shifts two weeks ahead."
K3,I2,P2,Q3,4,Gave the same shared-calendar answer; nothing new.
K3,I2,P2,Q4,3,"Compares weekly ticket tags to find the top three causes; named the three found."
K3,I2,P2,Q5,3,Strong analytical thinker.
K3,I3,P4,Q6,3,Said product is a close working partner.
K3,I3,P4,Q7,1,Answer was general; no specific case.
5. Process issues noted during the round
1. Operations coordinator: the second interview for K2 started 20 minutes late and ran short.
2. Recruiting coordinator: the feedback form has a rating box and a comment box but no place for an example.
3. Recruiting coordinator: the cross-team interviewer did not see the level descriptions for C5 before the round.
The review
INTERVIEW PROCESS REVIEW: Customer Operations Manager, round closed 6 November 2026
For: the recruiting lead. Prepared by: operations coordinator. Review date: 10 November 2026. From: rubric, 6 criteria; interview plan, 8 rows, 3 interviews, 7 questions; feedback, 24 rows, 3 candidates; process issues note supplied. Scale: 1 to 4. Columns not read: none. Defaults used: question minimum (2); spread threshold (2); candidate minimum (3).
Coverage: Feedback rows read: 24. Not read: none. Outcome: complete.
Headline: 7 questions in the plan; 4 interview-question pairs gave little usable evidence. Repeated across interviews: 1. Criteria: 3 covered, 1 thin, 1 asked with no usable evidence, 0 planned but never asked, 1 with no question. Rated rows with no usable note: 10 of 22. Ratings 2 or more points apart: 2 of 5 compared. Not asked: 2. Process issues noted: 3.
1. Questions with little usable evidence (asked 2 or more times, fewer than half of the notes usable)
Q5 · Operations panel · C4: 0 usable of 3 asked. Judged not usable: impression only 3.
Q2 · Hiring manager conversation · C3: 0 usable of 2 asked. Judged not usable: impression only 2.
Q3 · Operations panel · C2: 0 usable of 2 asked. Judged not usable: repeat of an earlier answer 2.
Q6 · Cross-team conversation · C5: 1 usable of 3 asked. Judged not usable: general statement 2.
2. Repeated questions
Q3 (C2) is in the Hiring manager conversation and the Operations panel. First: 3 usable of 3 asked. Later: 0 usable of 2 asked; 2 notes repeated the earlier answer.
3. Criteria coverage
C1 Handles escalations and fixes causes: covered. 3 of 3 candidates with usable evidence; Q1, Q7.
C2 Plans team cover: covered. 3 of 3 candidates with usable evidence; Q3.
C3 Coaches individuals: asked, no usable evidence. 0 of 3 candidates with usable evidence; Q2.
C4 Uses data to decide: covered. 3 of 3 candidates with usable evidence; Q4, Q5.
C5 Works with product and engineering: thin. 1 of 3 candidates with usable evidence; Q6.
C6 Writes clear customer guidance: no question in the plan. 0 of 3 candidates with usable evidence; no question.
4. Rubric use
Rated with no usable note, by interview: Hiring manager conversation 2 of 8 rated; Operations panel 5 of 8; Cross-team conversation 3 of 6. All interviews: 10 of 22.
Ratings 2 or more points apart where two or more interviewers rated the same criterion for a candidate: C1, 1 of 3 comparisons, largest gap 3 (K3 · I1 · Q1 and K3 · I3 · Q7); C2, 1 of 2 comparisons, largest gap 2 (K3 · I1 · Q3 and K3 · I2 · Q3).
Outside the scale: none.
These gaps may come from the ratings resting on different questions, so they are a reason to read the rows, not a finding about anyone. Each interview here had one interviewer, so an interview code identifies that person; keep this section to those who need it.
5. Questions not asked
1. K2 · I2 · Q3 (C2, the only question for C2): see process issue 1.
2. K3 · I1 · Q2 (C3, the only question for C3): no process issue noted.
6. Process issues noted
1. Operations coordinator: the second interview for K2 started 20 minutes late and ran short. The files show something related: one question not asked in that interview (K2 · I2 · Q3). Start time and length are not in the files.
2. Recruiting coordinator: the feedback form has a rating box and a comment box but no place for an example. The files show something related: 10 of 22 rated rows with no usable note. The form itself is not in the files.
3. Recruiting coordinator: the cross-team interviewer did not see the level descriptions for C5 before the round. The files show something related: 1 usable note of 3 for Q6, the only question for C5. What the interviewer was given beforehand is not in the files.
7. Notes judged not usable, for the recruiting lead to check
K1 · I1 · Q2 · impression only
K1 · I2 · Q3 · repeat of an earlier answer
K1 · I2 · Q5 · impression only
K2 · I1 · Q2 · impression only
K2 · I2 · Q5 · impression only
K2 · I3 · Q6 · general statement
K3 · I2 · Q3 · repeat of an earlier answer
K3 · I2 · Q5 · impression only
K3 · I3 · Q6 · general statement
K3 · I3 · Q7 · general statement
8. Proposed revisions (all drafts; each waits for the recruiting lead)
R1. Reword Q2 (C3, Hiring manager conversation) to ask for one past case. Because: finding 1, 0 usable of 2 asked. Draft wording: "Tell us about someone you coached. What was the gap, what did you do, and what changed?" Decision: awaiting the recruiting lead.
R2. Remove Q3 from the Operations panel. Because: findings 1 and 2; both notes repeated the first interview, and C2 is covered 3 of 3 there. Decision: awaiting the recruiting lead.
R3. Remove Q5 (C4, Operations panel). Because: finding 1, 0 usable of 3 asked; C4 is covered 3 of 3 by Q4. Decision: awaiting the recruiting lead.
R4. Reword Q6 (C5, Cross-team conversation) to ask for one joint piece of work, and ask the rubric's owners for the C5 level descriptions, which the files do not hold, so that interviewer has them before the round. Because: finding 1, 1 usable of 3; process issue 3. Draft wording: "Tell us about something you changed together with product or engineering. Who was involved, and when?" Decision: awaiting the recruiting lead.
R5. Add a question for C6 to the Operations panel, where R2 and R3 remove two questions. Because: finding 3, C6 has no question. Draft wording: "Describe a reply or help article you wrote that you are proud of. What did you leave out, and why?" Decision: awaiting the recruiting lead.
R6. Add a required "example or quote" field before the rating box on the feedback form. Because: finding 4, 10 of 22 rated rows had no usable note; process issue 2. Decision: awaiting the recruiting lead.
R7. Look with the panel at the rows named in finding 4 before the next search, and hold a short session on the C1 and C2 rubric levels if the gaps come from the levels rather than from different questions. Because: finding 4, ratings 2 or more points apart on 1 of 3 comparisons for C1 and 1 of 2 for C2; the C1 gap compares Q1 with Q7, so it may reflect different evidence. Decision: awaiting the recruiting lead.
R8. Put the only question for a criterion first in its interview, and agree in advance what to drop if time runs short. Because: finding 5, the only question for C3 was not asked for one candidate (K3 · I1 · Q2) and the only question for C2 was not asked for another (K2 · I2 · Q3); process issue 1 concerns the second. Decision: awaiting the recruiting lead.
9. Gaps in the files
None. Repeats are checked by question ID only.
The review of Customer Operations Manager looks at the process only. No candidate has been assessed, scored or compared, and no outcome has been reopened. Nothing has been sent and no record changed; every revision waits for the recruiting lead.
Where it fits
- Recruiting and hiring, after any round that will be run again for a similar role.
- People and HR, where the same interview process is reused for several hires and nobody has time to look back at it.
- Professional services, where a small group of partners and senior staff interview for the same few roles year after year.
- Software technology, where a loop of several interviews can end up asking one question twice and another never.
- Hospitality, where repeated hiring for supervisor and manager roles makes a better interview plan pay back quickly.
- Growth & scaling, when you hire in rounds and each round should make the next one better.
What has to change for your company
Decide what counts as usable evidence for your roles; the template starts with a recorded example, action or figure the candidate gave. Set your rating scale, the number of times a question must be asked before it is judged, and how far apart two ratings must be to count as inconsistent. Agree who sees the review, since it names interviews and questions but never people. Agree who may approve a revision to the plan, and whether the hiring manager also signs off.
The template works from your anonymised feedback export. Remove names and contact details first, and leave out any column that records an outcome or a personal characteristic; the assistant will not read such columns.
What has been checked
Fully Powered AI wrote this capability template. It was editorially reviewed on 10 November 2026. It was tried on 10 November 2026 on an invented example round in Claude, and that record is published with it. It has not yet been tested in a client environment. It has not yet been tested as a packaged version in ChatGPT, Claude, Copilot or Gemini.
Testing in your environment means running it on a past round your recruiting lead already knows well, with cases written in advance: a header that does not match, a question asked twice, a competency with no question, a rating with no note. The recruiting lead checks every count against the files, and reads the notes judged unusable, before anyone relies on the review. Interview plan and scorecard is the capability for the plan before a round, and Hiring pipeline status report covers the pipeline itself.
The full capability template
Everything above is the outline. The complete capability template is the working version: the roles, the inputs, the procedure, the exceptions and stop conditions, the output format, the reviewer checklist, the acceptance tests, the instructions for an AI assistant and what to customize for your company. It comes with a download and the link to its folder in Fully Powered’s public repository.
Free with an account. Sign in by email: no password and no card.
Next step
Put this capability to work for your team: on your plan, or built with you in a sprint.
Build this in a sprint
In a sprint, Fully Powered adapts this capability template to your rules, proves it on your own work and puts it in the hands of the roles that own it.
Two ways to begin: