Your best engineer just got a calendar invite: interview a candidate on Thursday. She has never interviewed anyone before. She will get a resume, a job description if she is lucky, and no guidance at all.
So she does what most first-time interviewers do. She improvises. She asks questions she remembers from her own interviews, forms an impression in the first five minutes, and submits a rating that says more about her mood than the candidate.
This is not her fault. It is a training gap, and it is one of the cheapest problems in hiring to fix.
The Real Cost of an Untrained Interviewer
An untrained interviewer produces noise instead of evidence. Their ratings swing on charisma, shared background, or how the conversation happened to flow. Two untrained interviewers can meet the same candidate and reach opposite conclusions for reasons neither can explain.
The damage compounds in the debrief. When one panelist has structured notes and another has a feeling, the feeling often wins because it is delivered with confidence. Candidates notice too; a rambling, unfocused interview signals a disorganized company.
There is legal exposure as well. Interviewers who improvise are far more likely to wander into questions about family plans, age, or health, topics that create real risk in many jurisdictions.
Start With the Role, Not a Question Bank
Most interview training starts in the wrong place: a list of questions to ask. Questions are the last step, not the first.
Begin by writing down the three to five capabilities the role actually requires. Be concrete. "Communication skills" is too vague to evaluate; "can explain a technical decision to a non-technical stakeholder" is something an interviewer can test and score.
Once the capabilities are defined, assign each one to a specific interviewer. This is the single highest-impact change most teams can make. When everyone owns a lane, the panel stops asking the candidate the same three questions five times.
Teach a Structure, Not a Script
New interviewers do not need a script. They need a repeatable shape for the hour, and a simple one works.
First five minutes: settle the candidate. Explain who you are, what this session covers, and how much time is reserved for their questions. Nervous candidates underperform, and that costs you signal.
Middle forty minutes: evidence gathering. Ask about specific past situations, then follow up. "What did you do next?" and "What was your specific contribution?" do more work than any clever question. Push past rehearsed answers politely but persistently.
Final ten minutes: their questions. Answer honestly. This is also evaluation time; the questions candidates ask reveal how they think about the work.
Shadow, Then Reverse Shadow, Then Solo
Nobody should conduct their first interview alone. The progression that works is boring and reliable.
First, the trainee shadows two or three interviews and only observes. Afterward, they compare their notes and scores with the experienced interviewer and discuss every gap.
Next comes reverse shadowing: the trainee leads while the experienced interviewer observes and stays quiet. The debrief afterward covers what the trainee asked, what they missed, and how they handled follow-ups.
Only then do they interview solo. The whole progression costs three or four hours of overlap per interviewer. One bad hire costs months.
Make the Scorecard the Habit
A scorecard is not paperwork; it is the mechanism that turns an interview into evidence. Keep it short: the capability being tested, a rating scale with defined anchors, and a required field for the specific examples that justify the rating.
The rule that matters most: notes are written before the interviewer hears anyone else's opinion. Scores discussed in the hallway converge; scores written independently reveal real disagreement, and real disagreement is where hiring decisions get better.
Practice the Awkward Parts
Most interview failures are not about questions. They are about moments. The candidate who goes silent. The answer that will not end. The person who asks about salary in minute ten. The rambler who needs a polite interruption.
Run a one-hour practice session where interviewers role-play these situations with each other. It feels silly and it works. Include a short segment on what cannot be asked, with concrete examples, so the boundaries are clear before anyone is in the room.
Close the Loop
Training does not end after the first solo interview. Twice a year, pull the data. Which interviewers' ratings actually predicted who succeeded after being hired? Whose scores cluster in the safe middle of the scale? Who is consistently harsher or softer than the panel average?
Share the patterns with the team. Interviewing is a skill, and like any skill it improves with feedback and decays without it.
The Payoff
A trained interview panel makes better decisions with fewer interviews. Candidates walk away with a clear picture of the company even when they are rejected. Debriefs get shorter because evidence replaces impressions.
None of this requires new software or a big budget. It requires treating interviewing as a real skill rather than something smart people are assumed to know how to do. Hand your next first-time interviewer more than a resume and a calendar invite, and every hire after that gets a little more reliable.