“It is a social ritual we keep complying with.”
Some things are so established that we do not even stop to wonder if they are really working. Sebastian and Dirk are both experienced interviewers and try to be structured and scientific in their approach. Yet, Dirk argues that interviews may not be as useful as we commonly belief. What would be your take on that?
Some extra reading:
- HBR: How to take the bias out of interviews.
- HBR: How to avoid hiring a toxic employee.
- BI: Google’s HR boss explains the 4 rules for hiring the best employees
- Simply Psychology: The interview method.
- Science: First impressions – are interview results preordained?
- Bloomberg: Job interviews are useless
Dirk & Sebastian
Image Source: CC0, Pixabay
Sebastian – Hello everyone and welcome to our latest edition of 2debate.net, our podcast debates. I am Sebastian and my co-host is Dirk. Hello Dirk, how are you today?
Dirk – Hi Sebastian, I’m pretty pumped actually. I love the motion that we are debating today.
What is the motion that we are going to debate on?
Dirk – Our motion is going to be: “interviews are good way to hire the right people in an organization”.
Sebastian – Shall we flip the coin to decide who’s going to start the debate today?
Dirk – Yeah we shall.
Sebastian – OK, so you’re heads, I’m tails, and whoever wins the flip of the coin shall kick-off the debate. Here we go with flipping the coin… and it’s heads, so you start the debate. The motion is: “interviews are a good way to hire the right people in an organization”. So you are against the motion. Are you ready, you have two minutes?
Dirk – Yeah, I am ready. so I’m starting the clock now.
Voice – OK, let’s do this. Dirk goes first and argues against the motion.
Dirk – Frankly, I think most interviews are borderline insulting: half of them are discriminating and even worse, practically all interviews I encountered so far are designed in a way that does not test fit-for-role. Most interviews test fit for a individual, conversational situation. So interviewers, in the end, know whether or not they like each other based on a, in an optimum case, one hour interaction. And to color that, I brought a couple of typical interview questions and it’s rhetorical, you don’t have to answer the question, Sebastian, but let’s start with the common interview questions: “why should we hire you? What do you consider your weaknesses? Tell me about a challenge or conflict you faced at work and how you dealt with it? Why are you leaving your current job? How many golf balls can fit into a school bus?”. All of these questions are actual interview questions and they don’t address the problem at hand: is someone a good fit-for-role. They attempt to test something that can be tested better otherwise. The real problems are actually not the question, it is the interviewer that is the problem. It has been proven scientifically over and over in studies that interviewers are bound to prove their own cognitive biases instead of really hiring for role. That is why orchestras don’t interview. Instead, they ask people to play an instrument behind the curtains, ask people to bring a piece of work if you will: and that is a much better test for fit-for-role and fit into an organization than an interview can be.
Voice – Now it’s Sebastian’s turn. Let’s hear his argument.
Sebastian – The answer is 350,000 golf balls in a school bus. Oh sorry that was not a question. OK, why should we still do interviews, why are they a good way? First of all, because they’re a good way to filter the bad people. If the interview goes really bad or the person does not answer, maybe in a way that is at least the minimum that could be expected, or say something unethical for instance which has happened in interviews I have given, because I do ask questions which are related to ethics which matters even more to me than performance at work, well maybe you can already weed out the obvious no-hires. How are you going to do this if you just listen to your instrument behind the curtain? I’m not saying not to do the instrument playing behind the curtain, I’m saying still talk to that person face-to-say. They’re a good way to not hire the wrong people. That’s one thing. Now, second thing: the motion is hiring the “right” people, not the strongest performers. When we talk about the right people, is he someone you want to work with? I would say this is exactly what you want to have. You want to have bias, you want to have a good feeling with that person. It doesn’t matter if they’re the best or the worst necessarily, maybe the worst is not the person you want to have on your team but maybe you don’t want to have the strongest, either because they have an ego which is so big they’re going to be destructive to the rest of the team. Maybe you need someone who’s the right fit indeed because you have a good feeling in the interview. Maybe your questions will give you an indication of that because they have either a sense of humor, maybe not like mine, and will actually make you want to work with them because they will be able to make you feel better in your work. Finally and as my third argument, before I can go into some more details, I am not saying interviews are the only way. The motion is “is it a good way”, not the only way, so my argument is let’s have a combination of things: the interview can be structured. What you’re complaining about, Dirk, here is that the interviews are maybe unstructured, you have questions all over the place, people have biases. So maybe we can do a combination of questions, of scenarios, practical case studies, live coding, whatever it is, we can go into the details as long as we have a clear protocol for that interview.
Voice – Now, it’s Dirk turn. Let’s hear his rebuttal.
Dirk – You kind of proved my case. So you’ve been mentioning things like “can I get along with that person cannot, do I know anything about that person that makes me work easier with them later?”. So what you what you’re interviewing for then is cultural fit. And the argument against that is twofold. One: after being hired on the basis of one hour interview, the cultural fit is not guaranteed, because cultural fit develops on the way, once someone is part of the team, the company. That’s argument number one, so you have no way of really assessing it in the first hour in that interview. Even if you have eight or nine interviews, you have no way of knowing more than “did I feel well in that particular hour?”. And guess what, people are well not always at their best and sometimes they show up at interviews, something just happened and they are derailed. And based on that, they are filtered out because you feel like it’s not a fit. Second argument, well, today things are not as they used to be. You’re not in the same role for 30 years and working the same team for ages. Chances are the hiring manager moves on like in half a year and all of a sudden, you’re surrounded by different people so the fit with you, the interviewer, doesn’t matter. What matters is how you behave in your role in your organization later on. And I would argue there are much better ways to actually hire the right people for an organization. In the most cases you want to know if someone can do the job. The best indicator of that is cognitive ability and has someone done similar things or key elements of that before. And you can test hard for that. The test in the orchestra is designed that way: you want to know if someone can play the cello and you let five people play the cello, and you select the best cello player. If you do an interview, then people are bound to follow the biases. And all of a sudden, you have less women in important roles, you have less black people in roles, you have other disabled people more or less, just because people are biased. If you remove that interview and focus on the skills at hand, you can even make them find an agreement that says “once you do something unethical, we can fire your ass and you’re out of that too” and you watch them in the context of the team afterwards anyway. It’s not that interviews prevented unethical people from joining companies. It’s nothing that prevents them from being a good actor and tell you what you want to hear in your interview. And that’s the third fallacy: people seem to think just because they can have a positive interview experience, they are good in interviewing and are good in weeding out the unethical ones from the ethical ones or the cultural non-fits from the fits, and that’s just an illusion that is not true.
Voice – And now onto Sebastian.
Sebastian – You decide within an hour of an interview. Well actually to make your point even stronger if you want: people decide within 10 seconds, that’s what people say, in terms of surveys. It’s not even an hour. Within 10 seconds, your first impressions already give you the answer whether you want to hire the person. But that’s only in the case of an unstructured interview. If you have structure, if you have a protocol, then things change completely. In fact, to get to that point directly, you talked about cognitive tests. In the surveys, the best practices that was I looking at, actually cognitive tests were as likely to give you an indication of a good hire, 26% of the cases, as structured interviews. This is I think key debating point that we have here. The examples that you provide are always relevant to something that is unstructured, unprepared, without rationality. Let me give you a few more examples as to why I say this. You say, even after eight or nine interviews, you will not know whether you want to hire that person. And it’s true. That’s why, at least in the case of Google and if you read the book “Work rules” by the former HR Vice President, that’s why we do now only 4 interviews. It’s enough. Further interviews will not actually give more accuracy in hiring the right people. In fact, we notice by using four people, including one which is not from the immediate team, provides that hiring accuracy that no single hiring manager would have. So by adding someone who’s outside of the hiring manager’s department, you will have this diversity, you will have this additional accuracy. You talk about diversity and bias towards possibly people of color or women, well, this is also something you can prepare for by having calibration. You look at interviews scores, you can map this and see whether when a white person interviews a person of color, when males interview a woman, whether there is a bias. You can measure this with the data and correct it. How can you correct it? You can actually neglect that interview and give the feedback back to that person who will not have realized they may have had either conscious or unconscious bias. So my key point here is: define the protocol, define something which is structured and this is actually also based on my personal experience of having interviewed maybe 300 people at Google – by having a proper benchmark and, for instance, in my case, I use very often the very same question so I can benchmark candidates against each other without them knowing, because it refines my understanding of how people respond to these questions, and how I can also provide better questions the next time around and having a proper calibration with myself. So I think that this is the key point here: not refraining from having interviews but rather going towards having structured interviews, because not only will they help hiring the right people, but they will also help weeding out the wrong people, because you can certainly weed out the wrong people. Maybe we’ll still have unethical people hired. But if you don’t meet those people, some of those people will still get hard but you could have had a chance to get them out in the first place. And this is happened in my case. I have interviewed people who were clearly too arrogant or had answered, failed, obvious ethical question.
Voice – Final statement. Dirk. Let’s hear it.
Dirk – Just to make that clear, I know that structured interviews are better than unstructured interviews. Still you simply defend a practice by making that practice even more detailed and even more structured. You admitted that there are biases showing through. And if we could use data to remove those biases, our company would be among the first ones to completely remove it. Yet still we have a problem of hiring enough women. Still we have a somehow unbalanced workforce and still we struggle with it. I argue that it is a social ritual we’re up against. That’s why it bustles people if I speak up against doing interviews because it’s something we’ve done for our tribes in years. We’re not letting anyone into our tribe without knowing that person. It’s a very human desire we have and then we rationalize it. We simply say “yeah it helps us having better candidates”. And that’s the motion that I’m arguing against. I think it has nothing to do with having better fit for the organization or the right people even. It has everything to do with social ritual, social norms that we keep complying with.
Voice – Sebastian.
Sebastian – I’m going to repeat myself but it’s too costly to hire someone who’s very toxic. So if you can detect it by talking to them – and you can – let’s keep the interviews. Three additional things. Number one is: the motion is it a “good” way, it’s not is it the “only” way. I’m not saying it is the only good way. It’s one of multiple good ways. In fact, this is what I would conclude with: it is to say that we should use it as a combination of additional things. Cognitive tests, as you were mentioning, I’m not bashing it. I’m saying it’s one additional weapon you can use to actually detect what you want to have in your organization. And also the motion is about interviews, not interviewers. If interviewers are not trained properly, yes, they’re going to make errors with it, just like when we talked about election polls: maybe it’s people who don’t know how to read it, rather than polls themselves which are erroneous, but there’s another debate on that, don’t we? And finally we’re talking about hiring the right people, not the strongest performers, that’s why I don’t care whether they play the instrument the best of ways, we want to have team players so if I get along with them and maybe I just need 10 seconds of an interview to get that, well maybe that’s all I care about. Just having the right people so that they fit in my team and whether the hiring manager leaves or not, it doesn’t matter because I would have had that person interviewed by people from different departments so they will actually have a more homogeneous assessment of that candidate.
Sebastian – So that’s it. That’s it we’re done with today’s debate. As usual, let us know what you thought of the debate in the comments of our website or social media channels. The website is 2debate.net, with the number 2, 2debate dot net. And don’t hesitate, what other arguments we could have used to make a better case for either side and stay tuned for our next debate. Thanks for having listened today and have a good rest of the day! Bye, Dirk!
Dirk – Bye! OK so here.
Dirk – Do you have a position where you can use a good debater, someone who is good with language, multiple languages?
Sebastian – Well, maybe not.
Dirk – Brainy, smart, good looking too. Don’t hesitate! Alright, let me do my two minutes. Thank you for answering that schoolbus question for me, it always puzzled me how many balls that are. That was a Google interview question by the way, you might have guessed.
Sebastian – Yes, I did guess.