# Example: Prepare a Candidate Comparison Brief

Fictional company, role and candidates.

## What the user pasted

> Final two for Data Analyst, Brightmoor Retail. Decision meeting Thursday, Lena (head of analytics) decides. Criteria: SQL and data modelling (must), turning analysis into business recommendations (must), stakeholder communication (must), dashboard design (nice to have). Scale 1–4, 3 = meets the bar.
>
> **Candidate C-14:** SQL test (scored by Omar): 4, clean joins and window functions. Case study with Lena: 3, good recommendation on store stock levels but slow to get there. Stakeholder interview with Jo (ops): 2, "too technical, lost me on the second answer". Showed a Tableau portfolio, nobody scored it.
>
> **Candidate C-09:** SQL test (Omar): 3, correct but long-winded. Case study: done with Omar, not Lena, as Lena was off sick. Omar gave 4, "really clear story, knew what the CFO would ask". Stakeholder interview with Jo: 4. Dashboard: not covered.
>
> Omar says C-09 "just feels like one of us". Jo changed C-14 from 3 to 2 after the group chat.

## What the skill produced

# Candidate comparison: Data Analyst

**Candidates:** C-14, C-09, in the order supplied, not ranked · **Stages:** SQL test, case study, stakeholder interview · **Decision meeting:** Thursday · **Decision owner:** Lena

## Side by side (panel scores)

Bar: 3+ on 1–4 = meets, as supplied

Labels: **Met / Below the bar (panel)** (bar supplied; "(one interviewer)" if one scorer) · **Not assessed** · **No score recorded** · **Split** (scores on different sides of the bar, two or more points apart, or changed after discussion; independent first) · **Not comparable** (see note)

| Criterion | Must-have? | C-14 | C-09 |
|---|---|---|---|
| SQL and data modelling | Yes | 4 · Met the bar (panel) (one interviewer) | 3 · Met the bar (panel) (one interviewer) |
| Analysis into recommendations | Yes | 3 · Met the bar (panel) (one interviewer) · Not comparable | 4 · Met the bar (panel) (one interviewer) · Not comparable |
| Stakeholder communication | Yes | Split (3 independent → 2 after discussion) | 4 · Met the bar (panel) (one interviewer) |
| Dashboard design | No | No score recorded (portfolio shown, not scored) | Not assessed |

Independent (pre-discussion) scores. One column per candidate.

## Evidence by criterion

### SQL and data modelling

- **C-14:** Clean joins and window functions (Omar, SQL test)
- **C-09:** Correct but long-winded queries (Omar, SQL test)
- **Comparable?** Yes: same test, same scorer.

### Analysis into recommendations

- **C-14:** Sound stock-level recommendation, slow to reach it (Lena, case study)
- **C-09:** "Really clear story, knew what the CFO would ask" (Omar, case study)
- **Comparable?** No: different interviewers ran the case study.

### Stakeholder communication

- **C-14:** Independent score 3; Jo's note: "too technical, lost me on the second answer" (Jo, stakeholder interview). Changed to 2 after group discussion, no reason recorded.
- **C-09:** Scored 4; no notes recorded (Jo, stakeholder interview)
- **Comparable?** Partly: same interviewer, but one score changed after discussion and the other has no notes.

### Dashboard design

- **C-14:** Tableau portfolio shown, not scored.
- **C-09:** Not covered.
- **Comparable?** No evidence scored for either.

## Comparability notes

- The case study was assessed by Lena for C-14 and by Omar for C-09, so the 3 and the 4 are not on the same footing.
- C-14's stakeholder score was changed from 3 to 2 after the group chat. This brief shows the independent 3 and flags the change.

## Trade-offs to discuss

- C-14 scored higher on SQL (4 vs 3); C-09 scored higher on stakeholder communication (4 vs C-14's independent 3). Both met the bar (panel) on SQL. Risk for the role: stakeholder communication is a must-have. C-14's independent score meets the bar, but Jo's note raises a concern the meeting should test against the anchors.
- C-09's case-study strength comes from one interviewer who did not see C-14's case.

## Process check

- One comment was left out of the evidence as it gives no job-related example.
- Jo changed C-14's score after discussion; the independent score is used here and the reason for the change needs recording.
- Jo's 4 for C-09 has no notes behind it.

## Questions for the decision meeting

1. What led Jo to change C-14's score after discussion, and does the note support 3 against the scale anchors? (stakeholder communication)
2. Should Lena review C-09's case-study output, or re-run a short case, so both are seen by the same person? (analysis into recommendations)
3. Can Jo add notes behind C-09's 4? (stakeholder communication)
4. Does dashboard design matter enough to score C-14's portfolio before deciding? (dashboard design)

## Decision meeting agenda

1. Confirm the three must-haves and that 3 = meets the bar
2. Review the score changed after discussion and the missing notes
3. Agree how to handle the case study done by different interviewers
4. Discuss the SQL vs stakeholder trade-off
5. Decision owner records the outcome and reasons: Lena

## Open questions

- Is dashboard design needed in the first six months? [TBC]
- Comments were left out because they referred to protected characteristics (health).
