💡 Core idea
Hiring is one of the highest-leverage things a manager does. Tools can widen the top of the funnel. They cannot take responsibility for who joins the team and why.
A hire is a years-long decision (Chapter 6’s one-way door). Treat it that way. The process should be designed before anyone is sourced. Starting interviews before you agree what “good” looks like is how you hire for comfort and call it instinct.
The same pattern shows up when companies fill manager roles. Gallup (Beck and Harter, 2015) reports that organizations miss the mark on high managerial talent in 82% of hiring decisions, and that managers commonly said they were chosen because of success in a previous non-managerial role or because of tenure. Those are weak predictors of the job this book is about. Design the scorecard for the work of managing, not for the work the person used to do.
Design the role before you source
Write the work to be done in the next twelve to eighteen months, not a fantasy wish list. Separate must-haves from teachables. Translate values into behaviors you can observe. That scorecard is the contract you will use in debrief, and later in performance conversations (Chapter 10). If you cannot write it, you are not ready to open the role.
Judgment belongs on the scorecard more than it used to. As production gets cheaper, the scarce skill is the ability to evaluate work — including machine-assisted work. Cong Wang describes taste as judgment compressed by time. Itamar Gilad’s competence/judgment gap (Chapter 6) is a hiring problem: if tools inflate output, interviews that only inspect output will be fooled. Sean Goedecke argues that you cannot prompt like an expert without the underlying domain mastery. Hire for that mastery, not for prompt fluency alone.
Tolan describes one attempt to rebuild technical interviews when models can write the code. The principle travels beyond engineering: test how a candidate frames a problem, checks a result, and explains a trade-off, not only whether they can produce a fluent artifact.
Structure beats vibe
Same core questions, scored against the scorecard, for comparable candidates. A short work sample, close to the real work and respectful of people’s time, will tell you more than an unstructured “coffee chat.”
Ted Gioia offers techniques for reading character. Use them cautiously. Do not turn them into a pseudo-scientific score, and do not let “character” become a polite word for “people like us.”
Debrief with evidence first, vote second, and no late private lobbying. “Culture fit” without behavioral evidence is how teams hire clones who cannot disagree.
Juniors are a design choice
Marc Brooker asks what happens to junior roles when senior work changes. If you stop hiring people who still need to learn the craft, you have not only made a headcount decision. You have cancelled an apprenticeship path (Chapter 13). If you hire juniors, you must be able to say what they will do, who will teach them, and how they will see enough real work to get good. That belongs in the role design, not in a later hope. In the language of Chapter 2, that path is mostly Scope it (mode 2): you scope, they produce, you review, until they can Release it and own the framing.
Charles Gorintin and Maxime Le Bras describe professional effectiveness as a triangle of technical ability, taste (discernment), and agency. AI collapses the execution side, but “L’IA ne supprime pas le besoin de compréhension technique. Elle déplace le curseur de l’exécution vers l’orchestration.” Agency and taste become the scarce differentiators; you still need enough technique to orchestrate well. On junior hiring, Le Bras calls the apparent logic — why hire a beginner when AI can do the work — “une incompréhension fondamentale.” “Les juniors d’aujourd’hui sont les gardiens du discernement de demain. Mais seulement si nous leur donnons les conditions pour le développer.” Use AI as a thinking partner, not as a substitute for effort. Chapter 13 returns to the learning path.
Tools
- Hiring scorecard. Must-haves, teachables, values in behavior, and how evidence will be collected — including judgment and verification, not only production.
- Structured interview kit. The same core questions, scored against the scorecard.
- Work-sample brief. Short, close to real work, respectful of time, hard to fake with an unreviewed draft.
- Debrief protocol. Evidence first, vote second, no late lobbying.
🤖 AI-era tip: Assist the process; do not secretly score the person
You may use an approved tool to draft a scorecard, generate structured questions, tighten a work-sample brief, and check a job post for exclusionary or vague language. Do not use a model as a hidden personality judge, and do not upload candidate materials to an unapproved tool. Legal and policy rules vary by place and will change; when in doubt, ask HR or counsel rather than inventing a practice from a blog post. A human owns the hire and should be able to explain it from evidence.
Takeaways
- Design the role and the scorecard before you source.
- Structure the process; prefer work samples; debrief with evidence.
- Hire for judgment, verification, and domain mastery, not for fluent artifacts alone.
- Junior hiring is a design problem: if the learning path is missing, do not pretend the role still exists.
- A human is accountable for the decision; a tool is not.
Evidence
- Beck, R., and Harter, J. (2015). Managers account for 70% of variance in employee engagement. Gallup. Organizations miss high managerial talent in 82% of hiring decisions; managers often reported being hired for prior individual success or tenure rather than for managing talent.
- Gorintin, C., and Le Bras, M. (2025). Le triangle de l’efficacité and Aura-t-on encore besoin de stagiaires ?. Practitioner essays: as execution gets cheap, agency and taste become scarce; collapsing junior tasks is not a reason to stop hiring beginners. See also Chapter 13.