Best for
- Teams that interview by gut and hire by halo
- Designing the loop for a role you have never hired before
- Making interview results comparable across candidates and interviewers
What you give it
- The role: what this person must achieve in year one, and the team context
- Your current process, if one exists, and what has gone wrong before
What you get back
- A competency map: the five or six things that predict success in this role, defined observably
- An interview loop where each stage samples real work and owns named competencies
- Scoring rubrics with behavioural anchors — and a debrief format that resists the loudest voice
How it works
- Derives competencies from the role's actual year-one work, not from a generic title template.
- Designs each stage to SAMPLE work rather than discuss it: exercises shaped like the job, sized respectfully for candidates.
- Writes rubrics with behavioural anchors: what a 2 looks like, what a 4 looks like — in observable terms.
- Assigns competencies to stages so everything important is measured somewhere and nothing twice by accident.
- Structures the debrief against bias: independent scores committed before discussion, evidence quoted before opinions.
Example
You: We are hiring our first product manager and nobody here has interviewed one before.
Result: A competency map built from the role's year-one reality (prioritisation under constraint, customer evidence habits, written clarity, working with engineers). A three-stage loop: a realistic prioritisation exercise with your actual (sanitised) backlog shape, a structured past-behaviour interview with probes, a written brief sample — each stage with a rubric anchored in observable behaviours, a 45-minute interviewer guide, and a debrief protocol where scores are committed before discussion.
Limits — please read
- No interview predicts perfectly; it designs for validity and says where the uncertainty stays.
- Legal specifics of hiring vary by place; it flags the areas (protected characteristics, exercise compensation norms) for your local check.
- Rubrics resist bias; they do not abolish it — the debrief discipline matters as much as the design.