One candidate answers every question with total assurance and leaves the panel nodding. Another hesitates, starts over, and gets marked down, even though their examples were stronger. By the debrief, the notes say “very confident” and “seemed unsure,” and no one in the room can point to the answer behind either.
This guide is for recruiters and hiring managers who interview candidates and sit in the debrief. Here’s a step-by-step way to judge confidence versus competence in an interview: decide whether confidence is part of the job, write down what a strong answer contains, probe a polished answer until it gets specific, give a nervous candidate the same chance, score each answer from what the candidate said, and bring the debrief back to the answers.
In an interview, confidence shows in how an answer sounds, and competence shows in what it contains: the situation, what the candidate did, and what changed. So my rule is simple: score what the answer contains, and when an answer stays general, however it sounds, ask for the detail.
Decide whether confidence is part of the job.
Before you can stop confidence from swaying a score, decide whether it belongs in the score at all. For some roles it does, and if you don’t name it, each interviewer will weigh it by gut feel anyway.
Start from the job. Someone who presents to customers, negotiates, or leads a team may need to speak with assurance under pressure. Someone who writes code, reconciles accounts, or answers support tickets by email mostly doesn’t.
If the role needs confidence, score confidence on its own. Write it as something a candidate does, such as explaining a recommendation to a skeptical stakeholder, and give it its own question and its own line on the scorecard. That keeps it out of the rating for every other skill.
If the role doesn’t need confidence, agree with your interviewers before the first interview that how fluently a candidate answers earns no points, and that every rating has to point to something the candidate said.
Write down what a strong answer contains before the interview.
A confident answer is hard to judge when the panel never agreed what a good one contains. Writing that down before the interview gives every interviewer the same thing to listen for.
For each question, list the evidence a strong answer includes. For a question about handling a difficult stakeholder, the list might be: a real situation and who the stakeholder was, what the candidate did themselves, what they said or wrote, and what changed as a result. Make each item something a candidate could say out loud, so no interviewer has to guess at a trait.
One way to build the list is to work out what your best interviewer already listens for, so the rest of the team can listen for it too. Viet Nguyen, head of global recruiting operations at Vercel, talks through that exercise with host Nolan Church on the 10x Recruiting podcast:
Share each question’s list with every interviewer who asks it. In the interview, it works as a checklist: an answer reaches those points or it falls short, however it sounds.
Your evidence list can also become the levels of a scoring rubric. A separate guide on how to write interview rubrics shows how to replace adjectives with what a candidate said.
Probe a polished answer until it gets specific.
A polished answer can sound complete and still say little. Follow-up questions show you whether there’s evidence underneath.
Listen for an answer that stays general: “we” where you need “I,” a description of how the process should work in place of what they did, or a result with nothing behind it. When you hear one, ask for the detail your list calls for, and keep asking until you get it or it’s clear the candidate can’t give it. Here are five general answers and what to ask after each:
| What you hear | What to ask next | What a competent answer adds |
|---|---|---|
| “We rolled out a new process and it went really well.” | What was your part in it? What did you do first? | The candidate’s own actions, in order |
| “I’d start by understanding the stakeholders.” | When did you last do that, and who were they? | A real example in place of a plan |
| “I always put the customer first.” | Tell me about a time that cost you something. | A specific situation with a tradeoff |
| “It made a huge difference to the team.” | How did you know? What changed? | How they judged the result |
| “I’m very detail-oriented.” | Walk me through a mistake you caught. | Something they did that shows the trait |
Use the same follow-ups with every candidate whose answer stays vague. If a nervous candidate gets probed and a confident one gets waved through, they’re no longer taking the same interview.
Give a nervous candidate the same chance to show evidence.
A polished answer is one way interviewers mix up confidence and competence. The other is the candidate whose nerves hide what they know. You can’t score evidence a candidate never got the chance to give, so set up the interview to draw it out.
- Explain the format at the start: how many questions, roughly how long, and that you’ll ask follow-ups. A candidate who knows follow-ups are routine won’t read one as a bad sign.
- Open with a question they can answer easily, such as what they do in their current role, before the questions you score.
- Let a pause run. A candidate who thinks before answering is working on the answer, and filling the silence for them cuts it short.
- Rephrase a question that lands badly, and ask for an example when the first answer is abstract.
- Come back to a question they rushed. Near the end, tell them you’d like to go back to that question, and give them a second go.
None of this lowers the bar. Use these steps with every candidate, so everyone gets the same chance to give the evidence you’re scoring. If the evidence still isn’t there after that chance, the score should say so.
Score each answer from what the candidate said.
Confidence can also sway a score after the interview, when an interviewer turns an impression into a rating. Rating each answer on its content, before you talk to anyone else, keeps the impression out. Four habits do that:
- Write your notes in the candidate’s words where you can. “Rebuilt the onboarding checklist after two new hires left in their first month” is evidence. “Very sharp” is a conclusion no one else can check.
- Check each note against the evidence list for that question. Mark the points the candidate covered and the ones they missed, and rate from that.
- Replace every word about delivery, such as “confident,” “polished,” or “nervous,” with the answer behind it. If there’s no answer to put there, delete the word.
- Submit your rating before you read anyone else’s or hear the hiring manager’s view.
Record what you couldn’t check, too. If an answer sounded strong but you ran out of time to probe it, say so on the scorecard, so the debrief knows that answer is still open.
Scoring from content counts most for behavioral questions, where a well-told story can hide how little happened in it. For those, there’s a guide to asking behavioral interview questions and scoring the answers.
Challenge words about delivery in the debrief.
The debrief is where impressions turn into a decision, and an interviewer who sums up a candidate with conviction can carry the room. A few questions bring the discussion back to the answers.
When someone describes a candidate as confident, polished, or articulate, ask which answer they mean, and then ask:
- What did the candidate say in that answer?
- Does that answer cover the evidence list for the question?
- Would we rate the same answer the same way if the candidate had given it haltingly?
Ask the same questions about a candidate someone calls nervous or unsure. If their answers covered the list, their nerves shouldn’t count against them, unless confidence is a criterion you set for the role at the start.
A separate guide to running an effective interview debrief covers the rest of the meeting, including how to prevent groupthink and what to do when the panel disagrees.
Where Metaview fits.
Every step in this guide depends on knowing what the candidate said. An interviewer taking notes while listening misses some of it, and a confident answer is easy to remember as better than it was.
Metaview is the Agentic Recruiting Platform. The Metaview Notetaker joins the interview as a visible participant and records the conversation. It transcribes what was said and turns it into structured notes. Where consent has been given, the recording captures every spoken word, so you can replay the exact answer behind a rating.
The Notetaker also drafts a scorecard from the conversation. The interviewer reviews it, rates the candidate, and submits it, so the rating stays with the person who heard the answers.
Reviewing that draft is also the moment to add anything the interviewer heard but couldn’t check. Shahriar Tajbakhsh, Metaview’s co-founder, made that point in a guide to writing an interview scorecard:
Interviewers shouldn’t be afraid to include hypotheses they didn’t get a chance to verify or falsify.”
In the debrief, the Metaview Assistant answers questions about your interviews in plain language and recalls exact details. When someone calls a candidate confident and names the answer they mean, you can ask the Assistant what the candidate said in that answer, and read it before the panel decides.
Metaview doesn’t decide who moves forward. Your interviewers rate the candidate, and your panel makes the call.
Check your last debriefs for confidence versus competence.
You can see how much confidence is driving your scores with the scorecards you already have. One role’s last few debriefs are enough.
- Pull the scorecards from those debriefs.
- Highlight every word about delivery, such as confident, polished, articulate, nervous, hesitant, or unsure.
- For each one, find the answer it came from, and mark the words with no answer behind them.
- Note whether each unsupported word sat beside a high rating or a low one. If “confident” keeps sitting beside high ratings and “nervous” beside low ones, delivery is driving the score.
- Pick the question with the most unsupported words, and write its evidence list before the next interview.
The unsupported words show you where your interviewers have been scoring confidence in place of competence. Start with that one question’s evidence list, and the next panel will have a way to tell the two apart.
Score competence from what candidates said.
See how the Metaview Notetaker keeps the exact answers on record, so your interviewers and your debrief work from the same words.
Frequently asked.
Isn’t confidence part of competence?
Sometimes. If the job needs someone who speaks with assurance under pressure, define that behavior and score it as its own criterion. For every other skill, confidence tells you how an answer sounded, and competence shows in what the answer contained.
Should you share interview questions in advance?
It can help a nervous candidate show what they know. If you do it, share the same questions with every candidate for the role, and use follow-up questions to test how deep each answer goes.
What if the hiring manager trusts their gut?
Ask which answer the feeling came from. A gut read often points at something real the candidate said, and naming it turns the feeling into evidence the panel can weigh. If no answer sits behind it, it shouldn’t move the decision.
Does this work for a recruiter phone screen?
Yes, in a shorter form. For each screening question, write down two or three points a strong answer covers, ask one follow-up whenever an answer is vague, and note what the candidate said in their own words.
How do you handle a candidate with rehearsed answers?
Ask a follow-up the rehearsal can’t cover, such as what they’d do differently now or how someone else on the project would describe their part. A candidate who did the work can answer a second question about it, and one reciting a script tends to repeat the script.