Talent SummonerCandidate Comparison Worksheet for Panels
Use a candidate comparison worksheet to capture independent panel ratings, resolve evidence disagreements and record a human-owned next step.
A candidate comparison worksheet for panels is a calibration record, not a verdict. It gives each reviewer the same criteria, captures their independent reasoning and makes disagreement specific enough to resolve. A good worksheet shows the evidence behind a provisional rating, what remains unknown and what the panel should do next.
Use it for one role and one agreed candidate set. It works alongside an evidence matrix: the matrix organises evidence by criterion and candidate, while this worksheet captures how different reviewers interpreted that evidence and how the group resolved the difference. It is especially useful when a panel is troubleshooting inconsistent scores or strong reactions to a particular profile.
Prepare the worksheet before names are discussed
Start with the approved role brief and record its version, date, decision owner and candidates in scope. List only the criteria that can change the next step. Classify each as a must-have, nice-to-have or genuine operating boundary. Define what evidence would count for each criterion, including acceptable equivalent experience, before reviewers see the names.
Give every criterion an identifier such as C1 or C2. A criterion should describe job-related work or a real constraint, not a proxy for a preferred background. "Owned production API migrations" is more reviewable than "looks senior". "Can work the agreed schedule" is a separate constraint from technical capability. Do not add a criterion because one candidate makes it feel attractive, and do not quietly change its meaning during calibration.
Have each reviewer complete a private first pass
Ask reviewers to fill their own copy before the panel conversation. They should record the candidate identifier, criterion, source passage or link, narrow interpretation and status: supported, adjacent, unknown or contradicted. If your process already uses a rating scale, add a provisional rating only after the evidence note. The scale belongs to your process; it is not an objective measurement of candidate quality.
Require one sentence explaining each rating. "4 because strong" is not a useful record. "The CV describes owning a migration, but gives no production scope" preserves both the signal and its limit. Unknown means the supplied material does not answer the question. It does not mean the candidate failed the requirement. Reviewers should also mark which missing fact could change their provisional rating.
Keep the first pass focused. Review the same criteria for every candidate, use the same evidence standard and avoid looking at another reviewer's labels until the independent record is complete. This makes a disagreement visible instead of allowing the first confident voice to set the panel's frame.
Compare reasons, not just totals
The facilitator can now compare one criterion at a time. Place the reviewers' statuses, evidence notes and provisional ratings side by side, then classify the disagreement before trying to resolve it:
| Disagreement type | What to inspect | Useful panel question |
|---|---|---|
| Evidence reading | The same passage was interpreted differently. | What is the narrowest claim this source supports? |
| Criterion definition | Reviewers applied different meanings to the requirement. | What evidence would satisfy this criterion for everyone? |
| Weight or trade-off | Approved priorities were applied differently. | Was this priority agreed before names were reviewed? |
| Missing information | One reviewer treated silence as a negative. | What fact is missing, who can verify it and when? |
Do not average conflicting ratings until the panel understands why they differ. A mean can hide one reviewer's concern about a missing must-have. If a criterion is ambiguous, pause the comparison, revise the brief with the decision owner and create a new worksheet version. If the evidence is incomplete, write one comparable question or work sample for every relevant candidate rather than giving only the preferred person extra scrutiny.
Turn each disagreement into a check
For every material difference, write the smallest check that could change the action. Name the owner, source or method, due point and stop condition. For example, if a reviewer reads "led a platform migration" as ownership and another reads it as participation, ask the same person-specific question about responsibility, operating context and scope. Record the answer against the same criterion; do not replace the original source note.
Separate a disagreement about evidence from a disagreement about the role. If several candidates appear adjacent because the requirement is too narrow, the right action may be revise rather than forcing a ranking. If the panel needs another reviewer, mark hold with the reason and owner. A worksheet should make uncertainty actionable, not manufacture consensus.
Record the panel decision and preserve the trail
After checks are complete, record one current disposition for each candidate: progress, verify, hold, revise or close. Add the job-related reason, accountable owner and next review point. Closing someone because of rank alone is not an explanation. A close reason should state which approved requirement or boundary is confirmed not met by reliable evidence or a completed verification, while preserving any unresolved uncertainty.
Save the original individual worksheets, comparison view, brief version, evidence updates and final dispositions together. Note any changed criterion or weighting. This preserves why a decision changed: new evidence, a role change or a corrected interpretation. Follow your organisation's privacy and retention practices for candidate material.
Copy-ready panel worksheet
Copy this block into a working document and repeat it for each candidate and criterion:
Role and brief version: [role] / [version]
Review date: [date] Decision owner: [name]
Candidate: [identifier] Reviewer: [name]
Criterion [C1]: [must-have / nice-to-have / boundary]
Evidence standard: [what would count]
Source passage or reference: [quote or link]
Status: [supported / adjacent / unknown / contradicted]
Narrow interpretation: [what the source establishes]
Provisional rating, if already approved: [rating + reason]
Disagreement type: [evidence / definition / trade-off / missing information]
Verification question or check: [question or work sample]
Owner and due point: [name / date]
Disposition: [progress / verify / hold / revise / close]
Job-related reason: [record the basis]Where Talent Summoner fits
Talent Summoner is our product. Its current Candidate Ranking tool, checked on 2 September 2026, lets a user paste or upload a job description and upload up to 50 CVs in PDF, DOCX, MD or TXT format. It returns a ranked shortlist with plain-English reasoning and provides a shareable link or PDF report without requiring an account. Bring that report into the panel's first pass, then complete this worksheet from the source material; the ranking is not a panel consensus or hiring decision.
When the supplied pool is too small, candidate sourcing, also checked on 2 September 2026, starts from a role brief and searches LinkedIn, GitHub and other public professional sources across 200M+ profiles. It returns a ranked shortlist with must-haves, nice-to-haves and plain-English reasoning. Treat those public profiles as discovery material to verify, not proof of current availability, interest or working-pattern fit.
Should panel members see each other's ratings first?
Complete an independent first pass before discussion when practical. Comparing reasons after that makes evidence, definitions and trade-offs easier to distinguish from conformity to the first opinion.
Does a worksheet need a total candidate score?
No. If your approved process uses ratings, preserve each criterion and its reason before any total. A total can hide an unknown must-have or an unresolved disagreement and should not be treated as a prediction.
What should the panel do with missing evidence?
Label it unknown, state the missing fact, assign an owner and use a focused, comparable verification question. Missing information is not proof that the candidate lacks the capability.
Can Talent Summoner make the panel decision?
No. Its Candidate Ranking tool supplies a ranked report and reasoning for submitted CVs. The panel still inspects evidence, resolves disagreements, chooses dispositions and owns the hiring decision.
For your next panel review, freeze the brief, circulate the independent worksheet and require one evidence note and one next check for every material disagreement. End with a named owner and disposition for each candidate.
