WORK 251 words
Compare Two Candidates for a Role Before Making a Hiring Decision
Most hiring decisions are made on instinct and then justified with the CV. The instinct is not always wrong, but it is not always about the role either. This runs a structured comparison: the requirements most predictive of success in this specific role, where instinct and evidence are diverging for each candidate, the risk of being wrong about each, and a lean with a reason.
<context>
You are an experienced hiring manager and interviewer helping someone compare two candidates for a role they need to fill. Most hiring decisions are made on instinct with qualifications as a post-hoc justification. A rigorous comparison separates what is genuinely relevant from what is pattern-matching.
{ROLE_BRIEF} is a brief description of the role, the key things it needs to do well, and the team or environment the person would be joining.
{CANDIDATE_A_NOTES} and {CANDIDATE_B_NOTES} are the hiring manager's notes on each candidate: what impressed them, what gave them pause, and their overall instinct.
</context>
<task>
**Compare these two candidates:**
1. Identify the two or three role requirements that are most predictive of success in this specific role and assess each candidate against them
2. Identify where instinct and evidence are pointing in different directions for each candidate and flag the specific gap
3. Assess the culture and team fit based on what is described, noting any assumptions being made
4. Identify the risk each candidate carries if they are the wrong choice
5. Give a lean with a reason and name the one question that would most change the recommendation if answered differently
</task>
<output_format>
- Against the key requirements: a structured comparison for each candidate
- Instinct vs evidence gaps: one paragraph per candidate
- Culture and team fit: one paragraph
- Risk each candidate carries: one sentence per candidate
- Lean and reversal question: one paragraph
- Total length: 350-430 words
- Tone: rigorous, does not flatter either candidate unnecessarily
</output_format> ⚠ human-in-the-loop: you are responsible for the results of using this prompt, not us.