Hiring as an EM: how to run a hiring process and decide

In this article

1. Before you start
2. Why interviewers ask this
3. Hiring as an EM from first principles
4. Worked example
5. Trade-offs and edge cases
6. How to answer it in an interview
7. Common mistakes
8. Follow-up questions
9. Sources

Before you start

This page sits in the engineering manager interview questions guide. It assumes no hiring experience.

Why interviewers ask this

Hiring is one of the few decisions an EM makes that lasts for years, so questions about it test judgment more than process knowledge. Candidates report being asked at several companies how they plan hiring for a team, what their hiring bar is, how they build a diverse candidate pool and what they did after a bad hire.

A weak answer describes a feeling: "I look for smart people who fit the culture." A strong answer describes a system: the role is defined first, the criteria are written down, each candidate gets the same questions, and the decision follows the evidence. A Staff-level answer also says how the manager checks later whether the decisions were right, and what the system costs in time.

This round usually sits inside the people management round, next to questions on growing engineers and handling a low performer.

Hiring as an EM from first principles

Rubric

A rubric is a written description, for one interview question or attribute, of what an outstanding, solid, borderline and poor answer looks like. It lets several interviewers score the same answer the same way.

Hiring bar

A hiring bar is the minimum standard a candidate must meet on the written criteria to receive an offer. It is stated before interviewing starts and applies equally to every candidate for the role.

In plain words

Think of judging a diving contest. If every judge invents their own idea of a good dive, the scores mean nothing. If the judges agree on what each score means and score separately, the numbers compare. The analogy breaks in one place. A diver repeats a known dive, but a candidate answers open questions, so the rubric describes answers, not movements.

In one precise sentence, hiring as an EM is a repeatable decision process: define the attributes, score every candidate on them with a rubric, and compare the scores to a bar set in advance.

Gut feel alone is a weak basis for a hiring decision. A well-known risk is confirmation bias: an interviewer forms an early impression and then hunts for evidence that supports it. A written rubric tests that impression against evidence.

Step 1: Define the role and its attributes

The hiring process starts by defining the role and its attributes. Write down the job level, the problems the person will own in their first year and what good looks like. Then choose the attributes you will assess.

Google's guide uses three categories: role-related knowledge, problem solving and leadership. For a manager role, you might define leadership as empowering a team, setting direction and developing direct reports; that is our suggestion, not the guide's wording. A behavioral question ("tell me about a time...") asks what the candidate did before; a hypothetical question ("imagine if...") asks what they would do.

Input: the headcount request and the team's needs. Output: a one-page role description with three to five attributes. You now know what you are hiring for. Without this step, a vague "cultural fit" tends to fill the gap; defined attributes are a better basis for the decision.

Step 2: Write the rubric before the first interview

The rubric comes before any interview. For each attribute, pick one or two questions and write what an outstanding, solid, borderline and poor answer sounds like. Google's guide says scoring with standardized rubrics gives all reviewers a shared understanding of those four levels.

Why before? Once you have met a charming candidate, you will unconsciously bend the criteria to fit. Writing the standard first removes that room. The same page reports that pre-made questions, guides and rubrics save on average 40 minutes per interview.

Input: the attributes from step 1. Output: a rubric per question, plus the hiring bar, for example "an overall score of at least a set number and no attribute below a set floor." You now have a standard that does not move when a candidate walks in. Rubrics take effort to build, and it helps to refresh the questions over time so that candidates who compare notes gain no edge.

Step 3: Run the loop with the same questions

The interview loop runs the same questions for every candidate. Give each interviewer one or two attributes. Attributes overlap across interviewers, so each attribute is scored by more than one person, but each interviewer asks different questions. Brief the interviewers on the rubric, and ask them to write down what the candidate said, not only a score.

Google lists recording comprehensive feedback of candidate answers as one of the four parts of structured interviewing, so that evaluators can review responses easily. The fourth part is interviewer training and calibration, which means practicing the rubric on sample answers until reviewers agree.

Input: the rubric and the candidate. Output: independent written scores and notes from each interviewer, submitted before anyone talks. Independent scoring stops the most senior or most confident voice from setting everyone else's view. You now have comparable evidence for every candidate.

Step 4: Decide on the evidence and close

The decision compares each candidate's scores to the hiring bar set in step 2. In the debrief, each interviewer shares their written score first. Disagreements are settled by returning to the notes and the rubric: what did the candidate actually say?

A scorecard is the table of scores per attribute and per candidate that the debrief works from. The worked example below builds one. After the decision, a strong EM closes the candidate with the specifics of the role, and later compares how the hire performs against the interview scores.

Google reports that structured-interview candidates gave higher satisfaction scores in feedback, and that structured interviews were more predictive of job performance than unstructured ones. That result is Google's, not a universal law.

1 · Define2 · Rubric3 · Run4 · DecideReplay

Step-by-step description

1. Define the role and its attributes.
2. Write the rubric and the bar.
3. Ask every candidate the same questions.
4. Compare scores to the bar.

• The hiring process starts by defining the role and its attributes.
• The rubric comes before any interview.
• The interview loop runs the same questions for every candidate.
• The decision compares each candidate's scores to the hiring bar.

Worked example

Two candidates, A and B, for an illustrative role that leads people, are scored on three attributes by two interviewers each, on a four-point scale (poor 1, borderline 2, solid 3, outstanding 4). The weights are illustrative: role-related knowledge 0.3, problem solving 0.3, leadership 0.4. The bar, also illustrative, is a weighted score of at least 3.0 and no attribute average below 2.5.¹

Candidate A averages 3.5 on role-related knowledge, 3.0 on problem solving and 2.0 on leadership, a weighted score of 2.75.¹ Candidate B averages 3.0, 2.5 and 3.5, a weighted score of 3.05.¹ The plain unweighted means are 2.83 for A and 3.0 for B.¹ A is below the bar on two counts: the weighted score is under 3.0 and leadership is under 2.5. B meets the bar. A gut-feel debrief might have favored A, who is strongest on technical knowledge; the scorecard shows why the role, a people-leading job, needs B.

The time arithmetic is also computed. With the 40 minutes per interview that Google reports saved by pre-made questions and rubrics, an illustrative loop of five interviewers and six candidates is 30 interviews and saves 1,200 minutes, or 20 hours.¹

Text description of the diagram

• Candidate A averages 3.5 on role-related knowledge and 2.0 on leadership.
• Candidate B averages 3.5 on leadership.

Trade-offs and edge cases

Structure costs time up front. Rubrics take effort to build, questions need refreshing, and some interviewers who trust their own judgment resist the change. The payoff is comparable decisions you can defend.

Evidence for structure is strong but dated. Schmidt and Hunter (1998) found that general mental ability combined with a structured interview reached a mean validity of .63 for job performance, the same as general mental ability combined with a work sample test. Sackett et al. (2022) revisited the estimates and report that mean validity values fall by .10 to .20 points, while structured interviews rank first among selection procedures. Sackett et al. revise the validity of the structured interview alone from .51 to .42, still the highest of the procedures they re-estimated. Cite the revised figure, not the 1998 one, as the current estimate.

Text description of the diagram

• Structured interviewing uses the same questions for every candidate.
• Structured interviewing scores answers on an identical scale and sets criteria in advance.

Edge cases: a very small team may have one interviewer, so use a written rubric and ask a peer from another team to score one attribute. A role with no existing standard needs the criteria written with the people who will work beside the hire. When a candidate does something the rubric did not foresee, record it and discuss it, but do not change the bar for that candidate alone.

On candidate pools, a strong answer shows the same discipline: widen where you source, apply the same criteria to everyone, and look at where candidates drop out of your process. Avoid quoting statistics you cannot source. The bias question has its own page: how do you ensure biases don't impact your hiring. How you then grow the people you hire is covered in growing engineers.

How to answer it in an interview

The spoken structure: define the role, say the rubric comes first, say scoring is independent, then name the trade-off and the follow-up check.

Common mistakes

• "I trust my gut; I'm good at reading people." (wrong) Gut feel is what structure replaces. (right) "I use a rubric so my impression is tested against evidence."
• "I look for cultural fit." (wrong) It is vague and lets bias in. (right) "I assess defined attributes: knowledge, problem solving and leadership."
• "We write the criteria after the interviews, once we know the candidates." (wrong) The criteria bend to the person. (right) "The rubric and the bar are written before the first interview."
• "Experienced interviewers don't need structure." (wrong) Interviewer confidence is not accuracy: in Google's own data, structured interviews predicted job performance better than unstructured ones. (right) "Even experienced interviewers score against the rubric."
• "Our debrief starts with the most senior person's view." (wrong) It anchors everyone else. (right) "Written scores come in first; discussion follows."

Follow-up questions

How are you planning on hiring for your team?

Start from the team's gaps and the approved headcount, define the role and level, write the rubric and bar, then source, interview and decide in that order. Name the timeline and who owns each step, and say how you will review the plan if the market or the team's needs shift. Link it to team health: hire for what the team lacks, not for a copy of yourself.

What's your experience building and maintaining a diverse candidate pool?

Describe what you did, not an abstract belief: where you sourced beyond your own network, how you wrote the role description, and how you applied the same rubric to everyone. Say what you measured, such as where in the process candidates dropped out, and what you changed. If your experience is limited, say so and describe what you would do.

How do you build a team?

Build around the team's goals: list the capabilities the work needs, compare them to the current people, and hire or grow to close the gaps. Balance experience levels so that seniors can mentor, and watch how the new mix works together, as in two engineers who can barely stand each other. If the role is technical leadership rather than management, see the tech lead interview questions.

How have you changed a team's trajectory?

Use a situation, action, result structure: the state of the team, the specific changes you made to hiring, expectations or coaching, and the measurable outcome. Say what you would do differently. For a team you did not build, see inheriting a team. The framing in your management philosophy should match the story.

What's your hiring bar, and what qualities do you look for in candidates?

Your hiring bar is the minimum standard a candidate must meet on the written criteria, stated before interviewing starts: an overall weighted score of at least a set number and no attribute below a set floor, the same for every candidate. Say you hold it even under hiring pressure. Then name the attributes you defined for the role, such as role-related knowledge, problem solving and leadership, and how you assess each one with a rubric.

What would you say if asked about a time you made a bad hire?

Be honest about the signal you missed, when you noticed it and what you did: clear expectations in one-on-ones and hard feedback, a time-bound plan, and a decision to part ways if it did not work, as described in handling a low performer. Finish with the change you made to the hiring process.

Sources

1. A guide to structured interviewing for better hiring practices — Google re:Work, 1 Mar 2026
2. The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings — Psychological Bulletin 124(2) (Schmidt and Hunter), 1 Sep 1998
3. Revisiting Meta-Analytic Estimates of Validity in Personnel Selection: Addressing Systematic Overcorrection for Restriction of Range — Journal of Applied Psychology (Sackett, Zhang, Berry and Lievens), 1 Jan 2022