A more scalable solution to this is pair interviews.
The first company I worked at for 4 years made use of this, in addition to meeting in person, at the end of the day, to make a decision.
Benefits of the pair interview: Two different opinions on the candidate for the same "experience". If either interviewer had poor tendencies, they'll be curtailed a bit because they know someone else is there to witness them.
If both interviewers are in agreement, that's a pretty strong data point.
When it came time for "review" at the end of the day, we started with an around the room "thumbs up/down". No hiding behind an anonymous email to the recruiter/manager. If it was unanimously up...offer. Unanimously down? Obvious no hire. If it was mixed, we'd go around to all of the thumbs down, and ask, "Are you on the fence, or are you 100% vehemently opposed to this candidate?" If they said, "NO WAY JOSE", we'd ask them to give their reasoning. Then, we'd ask if anyone had a positive enough experience to argue in favor of the candidate. If not? No hire. If yes, we'd then go around the room and ask all of the interviewers to give a full description of the interview. You'll note that this part gives the full data, but takes the most time, but could be avoided (an optimization) if it was going to be ultimately fruitless.
This was by far the best hiring decision making process I've been a part of. No process, is perfect, but I feel this was pretty fair.
Oh - lastly, if a candidate was particularly hard to decide on, and we went around the room, and some people were for them, and some were against, and it ended up in deadlock - we ended the meeting and it was up to the ranking members of the team to decide (i.e. managers.) This meant that after one round of debate, we wouldn't keep wasting the company's dime trying to force consensus, and instead moved the decision to the people who the company had already decided it wanted to make decisions. So, if democracy worked...great, but this was a business, and we had get shit done, so lacking clear consensus, the burden was reduced.
After typing this out, I really, really whole heartedly respect this process - compared to much of what I've seen in the Valley.
We do this at my current company. The tiebreaker stuff is slightly different, and every candidate that gets a vote of confidence actually ends up chatting briefly with our CEO as the very last bit, which is more of a cultural fit thing than anything else. I agree that it's good to have a potential counterpoint to odd personal interactions.
Interviews are clearly stressful situations for a majority of candidates, even if their technical chops are without question. This ends up manifesting in ways that can come off as personality defects / lack of fit. I can think of a few times where the person I went in with came out with a specific point in spacetime that was wrinkled for them - a comment they thought was off, something about their behavior - and more than half of the time I've noticed comments the candidate made, or compensating factors, that reasonably explained these wrinkles. A lot of the time, I'm trying to balance the things I care about asking with paying attention to my co-interviewer so I can provide objectiveness down the line.
The first company I worked at for 4 years made use of this, in addition to meeting in person, at the end of the day, to make a decision.
Benefits of the pair interview: Two different opinions on the candidate for the same "experience". If either interviewer had poor tendencies, they'll be curtailed a bit because they know someone else is there to witness them.
If both interviewers are in agreement, that's a pretty strong data point.
When it came time for "review" at the end of the day, we started with an around the room "thumbs up/down". No hiding behind an anonymous email to the recruiter/manager. If it was unanimously up...offer. Unanimously down? Obvious no hire. If it was mixed, we'd go around to all of the thumbs down, and ask, "Are you on the fence, or are you 100% vehemently opposed to this candidate?" If they said, "NO WAY JOSE", we'd ask them to give their reasoning. Then, we'd ask if anyone had a positive enough experience to argue in favor of the candidate. If not? No hire. If yes, we'd then go around the room and ask all of the interviewers to give a full description of the interview. You'll note that this part gives the full data, but takes the most time, but could be avoided (an optimization) if it was going to be ultimately fruitless.
This was by far the best hiring decision making process I've been a part of. No process, is perfect, but I feel this was pretty fair.
Oh - lastly, if a candidate was particularly hard to decide on, and we went around the room, and some people were for them, and some were against, and it ended up in deadlock - we ended the meeting and it was up to the ranking members of the team to decide (i.e. managers.) This meant that after one round of debate, we wouldn't keep wasting the company's dime trying to force consensus, and instead moved the decision to the people who the company had already decided it wanted to make decisions. So, if democracy worked...great, but this was a business, and we had get shit done, so lacking clear consensus, the burden was reduced.
After typing this out, I really, really whole heartedly respect this process - compared to much of what I've seen in the Valley.