The people who need the hire most are often the worst ones to judge it alone.

Not because they are bad interviewers. Because they want the seat filled, they want to like the person, and they will work next to them on Monday. That is a lot of pressure on one decision.

OpenAI seems to agree. This month it started testing an interview run by someone who doesn’t work on the hiring team at all.

1. What’s happening

Fortune’s Emily Forlini reported on 5 October that OpenAI is piloting a mission interview. It’s an extra step where candidates talk about how they would add to the company’s mission.

The interesting part is who runs it. The interviewer comes from a different team than the one the candidate applies to. OpenAI is also building a training programme so staff can get certified to run these interviews. The pilot starts with marketing and HR.

OpenAI isn’t the first. Fortune notes that Amazon has long used “bar raisers”: interviewers from outside the hiring team who check candidates against Amazon’s Leadership Principles. Meta runs its own version too.

The goal, per the article, is one consistent measure across teams as the company grows. That’s the part worth copying. Not the mission talk.

2. Why the team can’t always judge its own hire

There’s a well known study on this. Sociologist Lauren Rivera interviewed 120 people who hire at top banks, law firms and consulting firms, and sat in on a hiring committee (American Sociological Review, 2012).

Interviewers looked for people like themselves.

They favoured candidates who shared their hobbies, experiences and way of presenting. Often that mattered as much as, or more than, skill. Many judged “fit” by how comfortable and excited they felt in the room.

Now add the hiring team’s own pressure. The manager has a gap in the team and a deadline. The engineers are tired of covering for the empty seat. Everyone wants the next candidate to be the one.

So the bar slides. Not on purpose. It just slides.

Someone from outside the team doesn’t have that problem.

They don’t need this seat filled. They won’t sit next to this person on Monday. They can ask the plain question: is this person good at the job we said we need?

3. What I see from the outside

This is my job every week.

At AssessDev I interview candidates for other companies’ teams, as a third party with no stake in who gets the seat.

A few habits make this work. I use the same questions and the same scoring guide for every candidate in a role.

I take notes the whole time (old school, right?). Every candidate gets written feedback within 24 hours. And I ask candidates not to use AI in my interviews, and tell them why, so I see their own line of thinking and approach to solving problems.

None of this needs an outsider. Any team can do it. But it’s much easier when you’re not the one who’ll be short-handed if you say no.

4. The honest part

An outsider can’t judge everything. They don’t know your codebase, your team’s quirks or what the last hire struggled with. The hiring team still knows things nobody else does.

And a “mission” or “values” round can go wrong fast. If there’s no scoring guide, it turns into a vibe check. That’s exactly the “people like me” trap Rivera found. An outside interviewer with no rules just brings their own version of fit.

So the outsider isn’t a replacement for the team. They’re a second, steady pair of eyes.

5. How to add one

1. Pick someone with no stake in the seat. Someone from another team, or an outside interviewer. Not the hiring manager’s closest peer.

2. Give them a clear job. One or two skills to test, not “get a feel for them”. Write down what Fail, OK, and Great look like.

3. Use the same scorecard as everyone else. Same skills, same scale. Otherwise you can’t compare notes.

4. Score before the debrief. Everyone writes their score and notes before they talk. That way the loudest voice in the room doesn’t set the result.

5. Train them. OpenAI is certifying its interviewers. You don’t need a programme, but one practice session and a short brief help a lot.

6. Tell the candidate. Say who they’ll meet and why. “This person isn’t on the team. They’re here to keep the process fair.” Candidates like knowing that.

6. Your checklist

  • At least one interviewer has no stake in filling this seat
  • Each interviewer knows which skills they’re testing
  • Every round uses the same scorecard and scale
  • Scores are written down before the debrief
  • “Fit” or “values” rounds have a scoring guide, not just a feeling
  • The outside interviewer had at least one practice run
  • Candidates know who they’ll meet and why

TL;DR: The people who need the hire most feel the most pressure to say yes. One interviewer with no stake in the seat keeps the bar where you set it.

Who in your hiring process has nothing to gain from saying yes?

Your Outside Interviewer

AssessDev interviewers can be the outsider on your panel. They are senior engineers and CTOs who use your scorecard and have no stake in who gets the seat.

Tech Assessments

Structured technical interviews run by senior engineers and CTOs, with written feedback and a report within 24 hours.

Tech Interviewer Training

Eight hours of live training that helps your own interviewers run structured, fair interviews.

Curious on how bringing an outsider will look like?

Read our A Buyer's Guide to Interview-as-a-Service article.

Sources

This post first appeared in Shortlisted by Kat. Photo by Nikolay Vasiliev on Unsplash.