Hiring: Defining the Role and Structured Interviewing

From a Group to a Team showed what turns a collection of people into a team that produces reliable outcomes. But before you can build that team, you have to choose who is on it — and every person you add changes what the team becomes. Hiring is the single highest-leverage decision most managers make. A great hire lifts everyone around them for years. A poor one drains energy, absorbs your time, and is painful to undo. Yet most hiring is done badly: a loose job description, a friendly chat, and a decision that rests on a feeling.

This page gives you a better way — one you can apply whether you’re hiring a nurse, a warehouse supervisor, a barista, or a software engineer. The method has three parts: define the role before you meet anyone, interview the same way for every candidate, and decide from written evidence instead of the loudest voice in the room. None of it requires charisma or a talent for “reading people.” It requires discipline, and discipline is learnable.

Start by defining the role — on paper, before any interview

The most common hiring mistake happens before the first interview: nobody has agreed what the person is actually for. The job title exists, a vague description exists, but no one has written down what good looks like in concrete terms. So each interviewer walks in with a private, unspoken picture of the ideal candidate — and they end up measuring people against different rulers.

Fix this first. Before you post the role or talk to anyone, write down two things:

• Outcomes — the concrete results this person must deliver in the first 6 to 12 months. Not “manage the ward” but “run the evening shift so patient handovers happen on time and complaints stay low.” Not “do marketing” but “launch the monthly campaign and grow the email list by a set amount.” Outcomes are what you’d point to a year from now and say: that’s why we hired them.
• Competencies — the handful of skills and behaviors required to produce those outcomes. Keep it short: four to six. For a shift supervisor that might be calm under pressure, clear communication, planning, and coaching junior staff. For a cafe lead: speed under a rush, consistency, warmth with customers, handling cash honestly.

ROLE DEFINITION — Evening Shift Supervisor (Ward B)

Outcomes (what success looks like in 6–12 months)
  1. Handovers completed on time every shift
  2. Staffing gaps covered without last-minute panic
  3. Junior nurses report feeling supported

Competencies (the few things that produce those outcomes)
  • Stays calm and decides under pressure
  • Communicates clearly across shifts
  • Plans and prioritizes a busy floor
  • Coaches and develops junior staff

This one page does an enormous amount of work. It aligns everyone involved in the hire. It tells you what to actually test for in interviews. And it becomes the yardstick you score against later — so the decision measures the same thing for every candidate. If you can’t write this page, you are not ready to interview; you’re ready to think more.

The common trap — hiring for a résumé instead of a role

It’s tempting to shortcut role definition by looking for the most impressive background — the biggest company on the CV, the most years of experience. But experience is a proxy, not the target. Someone can have ten years of experience and none of the competencies your role actually needs, while someone with a shorter résumé has exactly them. Define the outcomes and competencies first, then let the CV matter only insofar as it predicts those. Otherwise you’ll hire impressive people who can’t do the specific job in front of you.

Structured interviewing: the same questions for everyone

Once the role is defined, the temptation is to “just have a conversation” and see if you click. This is the most natural way to interview and one of the least reliable. An unstructured chat drifts wherever rapport takes it, covers different ground with each candidate, and mostly measures how comfortable you feel talking to someone — which is not the job.

Structured interviewing means the opposite: you ask every candidate the same core set of questions, tied to the competencies you defined, and you evaluate their answers against the same criteria. It sounds rigid. It is far fairer and far more predictive.

The reason it works comes straight from first principles. If you want to compare candidates, you have to hold the measurement constant — the same test, the same conditions. A conversation that wanders down a different path with each person gives you nothing to compare; you’re left with a vibe. The same questions across everyone give you parallel evidence you can lay side by side.

The most useful question type is behavioral: ask what someone actually did in a real past situation, because past behavior predicts future behavior far better than hypotheticals (“what would you do?”) that just test how good someone is at describing an ideal.

COMPETENCY: Stays calm and decides under pressure

Behavioral question (same for every candidate):
  "Tell me about a time a shift went badly wrong —
   short-staffed, an emergency, something breaking.
   What did you actually do, and what happened?"

Follow-ups to probe for real evidence:
  • What were you weighing in the moment?
  • What did you decide first, and why?
  • What would you do differently now?

Notice the follow-ups. A good structured interview isn’t a rigid script read robotically — it’s a fixed set of questions you dig into, listening for concrete detail. Vague, tidy answers (“I stayed calm and handled it”) are a signal to probe; real experience comes with specifics.

The evidence — structure beats gut feel

Decades of research on hiring point the same direction: structured interviews predict on-the-job performance substantially better than unstructured ones. Reviews of the evidence generally find that the free-flowing “get a feel for them” chat — the format most managers trust most — is among the weaker predictors, while asking consistent, job-relevant questions and scoring them consistently is among the stronger. The uncomfortable lesson is that the confidence you feel after a great conversation is largely disconnected from how well that person will do the job.

Score with a scorecard, not a gut feeling

Structured questions only help if you record what you heard in a structured way. That’s what a scorecard is: for each competency you defined, a place to write the evidence you observed and a rating for it. You fill it in right after the interview, before you talk to anyone else, while the detail is fresh.

SCORECARD — Candidate: A. Rivera   Interviewer: You

Competency            Rating (1–4)   Evidence I heard
──────────────────────────────────────────────────────────
Calm under pressure       3          Described a real ER
                                     surge; prioritized well,
                                     but glossed over the
                                     staffing decision
Clear communication       4          Concrete example of a
                                     handover fix she led
Plans/prioritizes         2          Answers stayed general;
                                     no specific system
Coaches junior staff      3          Mentored two new nurses;
                                     gave detail on one

Overall: lean YES on communication; concern on planning

Two rules make a scorecard work. First, rate the evidence, not the person. You’re scoring “how much did I hear that shows this competency?” — not “how much do I like them?” Second, write the evidence down, not just the number. The sentence “answers stayed general, no specific system” is worth more than a lonely “2,” because it can be discussed, challenged, and compared. A number with no evidence behind it is just a feeling wearing a costume.

The scorecard also protects you from a subtle failure: remembering the conversation rather than the content. A week later you’ll recall that someone was warm and easy to talk to — but not whether they gave any real evidence of planning skill. The scorecard freezes the evidence at the moment you have it.

Design out the biases — because they’re automatic

Here is the hard truth that makes everything above necessary: your brain forms hiring judgments fast, confidently, and often wrongly. These biases aren’t a character flaw you can will away; they’re default wiring. So the strategy is not “try to be unbiased” — that doesn’t work — but to build a process that makes the biases matter less. Name the main ones so you can spot them:

• First-impression / halo effect. You form a judgment in the first minute — from a handshake, an accent, a confident opening — and then unconsciously spend the rest of the interview gathering evidence to confirm it. One good trait (say, being articulate) creates a “halo” that inflates your rating of unrelated traits.
• Similarity bias — the “culture fit” trap. You warm to people who remind you of yourself: same background, same humor, same way of talking. It feels like good judgment (“we just clicked”) but it’s often just recognition. Left unchecked, it builds a team of near-copies.
• Recency. The last candidate you saw looms largest simply because they’re freshest in memory, regardless of merit.

The structured method already counters these: fixed questions stop the halo from steering the interview, a scorecard filled in immediately blunts recency, and rating evidence rather than rapport limits similarity bias. A few extra habits help — having more than one interviewer, agreeing the questions in advance, and deliberately probing past your first impression rather than confirming it.

The common trap — “culture fit” as hiring people like yourself

“Culture fit” is one of the most abused phrases in hiring. In practice it often means “I’d enjoy a drink with them” — which quietly filters for people of the same background and personality as the people already there. That’s how teams become homogeneous without anyone intending it, and homogeneous teams tend to share the same blind spots. Reframe it. What you actually want is values alignment — do they share how the team treats people and work? — and culture-add — what perspective or strength do they bring that the team is missing? “Would they fit in?” becomes “Do they share our values, and what do they add?”

Decide from the evidence — compare scorecards before you discuss

The final step is where good process most often collapses. Everyone has interviewed the candidate; now the group gathers to decide. If the most senior or most confident person speaks first — “I loved her, great energy” — the room quietly anchors to that view. Softer doubts go unspoken. The decision becomes a reflection of the loudest voice, not the strongest evidence.

The fix is simple and powerful: each interviewer submits their scorecard before anyone discusses. Ratings and written evidence go on the table first, independently. Only then does the conversation start — and it starts from the data, not from a personality. Where scorecards disagree, that disagreement is useful: it tells you exactly where to dig, rather than being smoothed over by whoever talks first.

DECISION MEETING — order matters

WRONG                              RIGHT
─────────────────────────────      ─────────────────────────────
1. Senior person: "I loved         1. Everyone submits scorecards
   her!"                              (ratings + evidence)
2. Room agrees                     2. Compare side by side
3. Doubts stay silent              3. Discuss where they DISAGREE
4. Decision = loudest voice        4. Decision = weight of evidence

This is the same principle behind writing outcomes down before interviewing and evidence down before discussing: get the independent judgment on paper before the social dynamics kick in. Strong voices anchor rooms; paper doesn’t.

Case study — Google’s structured hiring (from the 2010s)

Google, having interviewed enormous numbers of people, publicly concluded what the research already said: that unstructured interviews and clever brain-teaser puzzles predicted job performance poorly, and that the fix was consistency — the same job-relevant questions, clear rubrics for scoring answers, and multiple interviewers rating evidence independently. The company also moved decisions away from a single hiring manager’s gut toward a committee reviewing the written evidence together. The lesson generalizes far beyond tech: the reliability came not from smarter interviewers but from a structure that made ordinary interviewers’ judgments comparable and harder to bias.

In practice — the hire the room almost talked itself out of

A cafe owner was hiring a shift lead. In the debrief, her most experienced barista jumped in first: “The second guy was great, really confident.” She stopped him and asked everyone to hand in their notes first. On paper, the “confident” candidate had given almost no concrete examples of handling a rush, while a quieter candidate had described, in detail, how she’d reorganized a chaotic morning service at her last job. The confidence had created a halo; the evidence pointed the other way. They hired the quieter candidate, who turned out to be the best lead they’d had. The scorecards didn’t make the decision — they made the evidence loud enough to hear over the charisma.

Putting it together

The whole method is one idea applied four times: replace private impressions with shared, written evidence measured against a definition you agreed in advance.

1. DEFINE   Outcomes + competencies, written down, before anyone
2. ASK      The same job-relevant questions for every candidate
3. SCORE    Rate the evidence per competency, right after, in writing
4. DECIDE   Compare scorecards first — then discuss the disagreements

It’s more work up front than a friendly chat. But hiring is the decision you least want to get wrong, because it’s the hardest to undo and its effects compound over years. A little discipline here saves you enormous pain later — and, just as importantly, it’s fairer to the people you interview, who get judged on what they can do rather than on how much they happened to remind you of yourself.

How it connects

Hiring is the front door to everything else in this part. Get it right and onboarding has good raw material to work with; the clear outcomes and competencies you wrote here are exactly what a new hire needs on day one. Defining who does what feeds directly into role clarity and RACI. And reframing “culture fit” as values-plus-add is how you build a team with enough difference to have real psychological safety and a healthy culture of behaviors — rather than a room of people who all agree because they’re all the same.

Try this

Take a role you’re hiring for now, or the last role you filled, and write the one-page role definition from scratch: the concrete outcomes for the first year and the four-to-six competencies that produce them. Then draft one behavioral question for each competency — a “tell me about a time you actually…” question. If you’re hiring soon, build a simple scorecard from those competencies and ask every interviewer to fill it in before the debrief. Notice how much the conversation changes when it starts from evidence instead of impressions.

Reflect

1. For a recent hire — good or bad — had anyone actually written down the outcomes and competencies before the first interview? If not, what were people really measuring against?
2. Think of the last time you “just knew” someone was right in an interview. What concrete evidence backed that feeling — and how much was rapport?
3. When you picture your ideal candidate, how much of that picture is “someone like me”? What would a culture-add hire — someone who brings what the team lacks — look like instead?
4. In your last hiring decision, who spoke first and how much did the room anchor to them? Would independent scorecards have changed the discussion?
5. Which single step — defining the role, structuring the questions, scoring the evidence, or deciding from paper — is weakest in how you hire today, and what would fixing it cost you?

Show reflections

1. Most honest answers reveal there was no shared definition — just each person’s private picture. That’s the root cause of interviewers who “disagree about the candidate” when they’re really disagreeing about the job. Writing the definition first is the cheapest, highest-return fix available.
2. The aim is to separate signal from rapport. A strong answer can point to specific things the person did that map to the role; a weaker one realizes, on reflection, that the “knowing” was mostly comfort and confidence — exactly the halo effect the method is built to counter.
3. This is the similarity-bias check. The useful move is to describe your ideal candidate honestly, notice how much resembles you, then ask what perspective or strength the current team is missing — that’s the culture-add lens, and it’s how teams stay diverse enough to see their own blind spots.
4. Watch for anchoring: the first confident voice usually sets the tone and silences softer doubts. If you suspect it did, that’s the strongest argument for collecting scorecards independently before any discussion — so the evidence, not the seniority, leads.
5. There’s no wrong answer, only an honest one. Naming the weakest link tells you where to spend your discipline; often it’s the last step — deciding from paper — because it requires overriding social habits in a live room, which is harder than writing a question at your desk.