Executive search
Executive Search Shortlist Feedback: A Template for Evidence and Next Actions
A feedback record and reconciliation agenda that links each client comment to agreed criteria and evidence, separating preference, missing evidence, changed criteria and decisions.
When shortlist feedback conflicts, do not merge it into a single view. Record who made each comment, the agreed criterion and the evidence behind it. Distinguish an evidence-linked view from a preference, missing evidence, a proposed criterion change or a decision. If the meaning is unclear, ask before classifying. Take unresolved items into the next discussion with a named decision owner.
The rest of this article gives you the record, the agenda and a fully synthetic worked example. It also covers the ways this workflow tends to break and how to recover.
TL;DR
- Log every shortlist comment with attribution. Record who said it, which candidate it concerns, which agreed criterion it relates to and what evidence it cites. Do not average it, score it or summarise it into a consensus.
- Separate evidence-linked views from items needing clarification, more evidence, an agreed criterion change or a decision. Do not turn an unclear comment into a requirement by interpretation.
- Apply an agreed criterion change consistently to the relevant comparison group and reconsider affected earlier decisions. Check missing evidence wherever the same job-related question applies.
- The record supports human judgment. It does not remove bias and it does not produce a hiring outcome. Disagreement stays visible until the agreed decision-maker resolves it and gives a rationale.
The feedback record: copy this first
Use two small tables per shortlist round instead of one wide one. The first table records what was said. The second records what the comment needs next. The two are linked by an ID.
Table 1: Comment log
| ID | Who (role) | Candidate | What they said (close paraphrase) | Agreed criterion |
|---|---|---|---|---|
| F1 | C1 / C2 / none | |||
| F2 |
Table 2: Evidence and next action
| ID | Evidence the stakeholder cited | Evidence in our file | Type | Next action and owner |
|---|---|---|---|---|
| F1 | ||||
| F2 |
Field rules
- Who. Record the name and role, for example "CEO" or "Board member, audit chair". Never write "the client".
- What they said. Use a close paraphrase that the stakeholder would recognise. If you are unsure what they meant, write the question you need to ask rather than guessing.
- Agreed criterion. Use the criteria reference from the kickoff. If the comment does not map to any criterion, write "none". Do not stretch a criterion to make the comment fit. The mandate kickoff checklist is where those references should come from.
- Evidence in our file. Point to a specific observed example from interviews, referencing or research. Do not use an adjective. The candidate research notes structure makes this lookup quicker.
- Type. Choose the entry type below when its meaning is clear. Otherwise mark it pending clarification and record the question and owner.
- No fit score column. A single number hides who thinks what and why. If someone asks for a ranking, record it as a stakeholder view with that person's name on it.
Entry types and what each one needs
| Type | Test question | What happens next |
|---|---|---|
| Evidence-linked view | Does the comment interpret stated evidence against an agreed criterion? | Preserve the attributed view and source pointer. Compare conflicting interpretations before a decision. |
| Preference | Is the meaning clear, and is the comment not tied to an agreed criterion? If so, it is a preference. If the meaning is unclear, mark it pending clarification instead. | Record it. It does not move a candidate unless the client turns it into a stated criterion, which makes it a changed criterion. |
| Missing evidence | Is the criterion agreed but the evidence absent, thin or second-hand? | Plan a targeted follow-up and apply the same check to every candidate where it is relevant. |
| Changed criterion | Does the comment assume a requirement that differs from what was agreed? | The accountable stakeholder confirms or rejects the change explicitly. If confirmed, date it and re-review all candidates against it. |
| Decision | Has the agreed decision-maker stated an outcome against the criteria? | Record the decision, its rationale and any dissent, with names. If it depends on open items, mark it as proposed. |
"Her plant-expansion example meets C1" is an evidence-linked view, not a progression decision. A vague "not senior enough" remains pending until its author explains the job-related concern. Neither comment should silently become a new must-have.
This structure borrows an organising idea from federal assessment guidance. OPM's job analysis guidance connects job tasks with the competencies they require. Applying that link to search-firm feedback is this article's adaptation. It is not an OPM requirement for executive search.
The reconciliation agenda
Build the agenda from Table 2. Leave out anything that is already resolved. Send it to the stakeholders ahead of the next shortlist discussion so nobody is surprised.
| # | Item | Purpose | Who decides | Output |
|---|---|---|---|---|
| 1 | Proposed criterion changes | Confirm or reject each one before discussing candidates | Agreed accountable stakeholder | Dated criteria list |
| 2 | Missing evidence | Agree which follow-ups to run and confirm they apply to all relevant candidates | Consultant proposes, client confirms | Follow-up plan with owners |
| 3 | Conflicting evidence-linked views | Put both views on the table with the evidence each relies on | Agreed decision-maker | Resolution and rationale, or a named open question |
| 4 | Proposed decisions | Confirm progression decisions now that items 1 to 3 are settled | Agreed decision-maker | Recorded decision with any dissent noted |
| 5 | Preferences | Note them openly and ask whether any should become a criterion | Accountable stakeholder | Either "noted" or a new item 1 |
The order matters. If you discuss candidates before settling the criteria, a new requirement can quietly decide the outcome for one person while the others are never assessed against it.
Worked example (synthetic)
Everything in this example is fictional. No real client, candidate or engagement is described, and nothing here implies a result.
Context. A fictional company, Northfold Instruments, has retained a search for a Chief Operating Officer. The kickoff agreed these criteria:
- C1: Has led multi-site manufacturing operations through a capacity expansion.
- C2: Has built and retained a senior operations team.
- C3: Has worked directly with a board on capital allocation.
The kickoff also agreed decision rights. The CEO decides progression after consulting the board member and the CHRO.
Stakeholders (fictional): Dana, CEO. Marcus, board member. Ines, CHRO.
Shortlist (fictional): Candidates A, B and C.
Comment log
| ID | Who | Candidate | What they said | Criterion |
|---|---|---|---|---|
| F1 | Dana, CEO | A | Not convinced A is a COO. A reads more like a strong plant leader. | C1? |
| F2 | Marcus, board | A | A is the strongest operator on the list, based on the expansion story. | C1 |
| F3 | Ines, CHRO | B | Has heard that two of B's direct reports left within a year. | C2 |
| F4 | Dana, CEO | C | Great energy. Would enjoy working with C. | none |
| F5 | Marcus, board | B | B's board exposure looks thin. | C3 |
| F6 | Ines, CHRO | B | Disagrees with Marcus. Thinks B's committee work covers C3. | C3 |
| F7 | Dana, CEO | All | Move A and C forward and drop B. | — |
Evidence and next action
| ID | Stakeholder evidence | Our file | Type | Next action |
|---|---|---|---|---|
| F1 | None stated | A ran three sites and led one expansion programme, but did not hold enterprise-wide P&L | Pending clarification | Ask what job-related evidence Dana needs. Is enterprise P&L being proposed as a requirement, or is there another concern? Do not infer a change. |
| F2 | Expansion example | Matches our interview notes | Evidence-linked view | Bring to agenda item 3 alongside F1 |
| F3 | Second-hand, from Ines's network | No verified retention evidence | Missing evidence | Agree a permissioned C2 follow-up for the relevant comparison group. Keep source attribution in restricted records, outside the wider shared log. |
| F4 | None | Not applicable | Preference | Note it and ask whether it reflects a criterion |
| F5 | General impression | B presented to an audit committee on a regular cycle | Pending clarification | Ask Marcus which C3 evidence he considers thin, then take it to agenda item 3 alongside F6 |
| F6 | Committee work | Same as F5 | Evidence-linked view | Agenda item 3 |
| F7 | Not stated | Depends on F1, F3, F5 and F6 | Decision (proposed) | Hold until items 1 to 3 are settled |
Review of the example
What the record prevents. A quick summary might have said "Client positive on A, lukewarm on B." That sentence erases Dana's doubt about A, the disagreement between Marcus and Ines about B, and the fact that the concern about B rests on second-hand information. Each of those is now a separate, attributed line.
The possible criterion change. F1 does not establish that Dana wants enterprise-wide P&L. The consultant must ask. If she proposes and confirms that requirement, apply it consistently and revisit affected earlier decisions; the candidate rediscovery workflow helps structure that review. If no change is intended, ask which evidence against the existing criteria concerns her. Do not automatically relabel the comment as a preference either.
Comparable evidence. F3 could become a check aimed only at B. Instead, agree how the same relevant C2 question will be examined across the comparison group, within the permissions for each person. A rumour is not a finding. Consistent questions help the review; they do not make it bias-free.
The proposed decision. Dana has the authority to decide, and the record does not override her. What it does is make visible that F7 currently rests on an unconfirmed criterion and an unverified concern. The consultant asks Dana to decide once those are settled, and records her rationale. If Marcus or Ines still disagree, their dissent is recorded by name.
The preference. F4 is not dismissed. It is shown openly as a preference, which is more honest than allowing it to influence the outcome without being named.
The resulting agenda (synthetic)
- Clarify F1. Ask Dana for the job-related concern before proposing a criterion change. If enterprise P&L is then agreed, date it and plan a consistent re-review.
- C2 retention evidence. Confirm the referencing plan covers all three candidates, and agree who conducts it.
- C3 for Candidate B. Marcus and Ines each walk through the evidence behind their view. Dana resolves it, or the group names the specific missing piece.
- Progression. Dana's decision, with rationale and any dissent noted.
- Preferences. Note F4 and ask whether any criterion is missing.
Workflow failures and how to recover
1. Someone writes a consensus that does not exist
Symptom: The internal note says "client aligned on A" after a call in which one stakeholder raised doubts.
Recovery: Reopen the comment log and restore the attributed views. Correct the note before it reaches the candidate debrief or the next client update. Unattributed summaries tend to get repeated until people treat them as fact.
2. Criteria drift without anyone recording it
Symptom: A candidate is turned down for lacking something that was never in the brief, and nobody checked the others for it.
Recovery: Treat it as a proposed changed criterion. Ask the accountable stakeholder to confirm it. If confirmed, date it and apply it to the full shortlist and to relevant earlier candidates. If it is rejected, revisit the turn-down.
3. Reactions arrive without explanations
Symptom: The client marks candidates up or down in a shared system but leaves no reasons.
Recovery: Ask for a comment tied to a criterion. Clockwork's long-list documentation describes reactions and comments as separate records. Our inference is that a reaction on its own may not explain a decision.
4. Sensitive second-hand information lands in a shared document
Symptom: A stakeholder's informal back-channel information about a candidate is pasted into notes that the wider client team can see.
Recovery: Contain further sharing and follow the firm's incident procedure. Preserve an appropriate restricted record of the source and what happened; the responsible owner determines correction and communications. Route any follow-up through the agreed permissions. The AESC Candidate Bill of Rights addresses member-firm confidentiality expectations for candidate information, sources and references.
5. You rebuild what your system already holds
Symptom: You build a separate feedback spreadsheet while candidate notes already sit in your search system.
Recovery: Check the existing views first. As one example, Clockwork's status reports overview describes candidate status views that show profile data and recent notes to firm and client users. Add only the fields you are missing, such as type and criterion reference.
6. A follow-up is asked of only one candidate
Symptom: Extra probing on a criterion happens only for the candidate who drew the concern.
Recovery: Ask the same job-related question of every candidate where the criterion applies. OPM's structured interview guidance describes predetermined job-related questions and common rating standards. The template in this article is not a validated instrument, but that consistency principle is still a useful check.
Keeping it separate from the client update
This record is a working document for reconciliation. It is not the update you send to the client. The client update template covers market response and progress reporting. Feed it the outcomes of reconciliation, such as confirmed criteria, decisions and follow-ups, rather than the raw log.
For AESC member engagements, the AESC Client Bill of Rights describes expectations of confidentiality and ongoing communication about candidates and obstacles. The reconciliation agenda gives you a clear way to raise those obstacles.
A practical run order
- Within a day of each feedback conversation, fill in Table 1 while the wording is fresh.
- Before drafting anything for the client, complete Table 2 and assign types.
- Send the agenda to the stakeholders, listing only the open items.
- During the discussion, take criteria first, then evidence, then decisions, then preferences.
- Afterwards, record each outcome and its rationale, keep dissent attributed, and update the dated criteria list.
- Carry forward any item still open as a named question with an owner. Do not mark it resolved.
Where Laisy fits
You can run this manually in tools your firm permits. Laisy offers personally supported early access for specialist executive-search firms, starting with an agreed task alongside current tools. Applications are reviewed individually; access and response dates are not promised.
A useful starting point is a single live assignment. Agree which permitted inputs to use and review every output before it is shared. Define in advance what evidence would show the work is useful, and decide when you will continue or stop. No full migration is needed. If you would like to explore that, you can apply to discuss an assignment.
Frequently asked questions
What if the decision-maker decides against the evidence?
Record the decision, the rationale they give and any attributed dissent. The record exists to make the basis for a decision visible, not to overrule the agreed decision-maker. If their rationale implies a criterion that was never agreed, raise that explicitly as a changed criterion.
Should I share the feedback record with the client?
Usually you should share the agenda and the reconciled outcomes, not the raw log. The log may contain candidate-sensitive information and paraphrases that each stakeholder has not checked. Decide what to share based on your engagement terms and your confidentiality obligations.
Why not just use a weighted scorecard?
A single score hides who holds which view and which evidence supports it. That makes disagreement harder to resolve, not easier. If a client wants ratings, record them per stakeholder and per criterion, alongside the observed example behind each one.
What about feedback from someone outside the agreed stakeholder group?
Log it with their name and role, and type it like any other comment. Then ask the accountable stakeholder whether that person's input should shape the decision. Do not let it change criteria or decisions without that confirmation.