Transcript
Interview with Bas van de Haterd
July 2021
Kevin
Hello, everyone, today we’re going to be talking with the Bas van de Haterd in the Netherlands. Bas is a good friend of mine and is also an expert in assessment, which is really our topic for today. Bas, tell us a little bit about yourself and how you became interested in assessment?
Bas
Well, to be honest, I’ve been in the recruiting industry, I guess for about 2 years or so. Always a bit on the sidelines, I’ve always been consulting with companies. I did a lot on employer branding and candidate experience. And I’ve noticed that a lot of companies, especially at that time, when unemployment was a bit higher than it is right now, we were getting in so many applicants, and [recruiters were] all complaining about not getting enough quality. And I was like, how do you know, especially for entry level jobs at that point? They were like, yeah, this isn’t good enough, yet can’t be that this person has the talent. And then I met a company in Britain, who were doing game based assessments. And that was my first introduction with assessments because I already knew assessments, but I was like, questionnaires can be cheated, and I was as a professional speaker, not at your level but like you, Kevin, I was talking about all these new developments everywhere. And then what really was a trigger for me was when I was speaking in The Hague, I remember it so well, it was healthcare at the head of recruitment from a local health care organization who came up to me, he said, You’re talking about his game based assessment right now? Yes. You’re saying they can’t cheat on them. I said, Well, I won’t say can’t, but it’s a lot harder because you don’t know what’s being measured. She said, Yeah, because we do assessments for every hire. And we hire a lot of what we call site entries. So people who don’t register as a nurse yet, but we’re going to train them to be a registered nurse. But they can start with what we call a helping hand, in order to be a level two nurse. And we give everybody in a traditional assessment a questionnaire. She says, and we also give them a personal development plan on day one, based on their assessment. She says 1/3 of all my hires come up to me, after that probation period, which is two months in the Netherlands, and say, Yeah, can I do the assessment again, because I kind of cheated a bit towards what I thought you wanted to know. And now, I’m actually not getting the training, which I would have loved because I’m actually not that good at either this subject or that subject. And I would love to get a training on that. And not on this, because I’m actually a lot better, but you know, you don’t be perfect. So I cheated a bit on the assessment. She said, 1/3, I’m embarrassed to tell that honest. Yeah, but that’s after the probation period, so they can be fired anymore. That’s thats teh Dutch system, you know. And I wasn’t actually amazed by she actually said to me, so when one firm is willing to admit it, I’m pretty sure that more. So I would love to see a less cheatable form of assessments. And that’s why when I knew this was a really interesting area, and then a Dutch company started called Harver. And you probably heard of them. And their very first client was a contact center, they they started out in contact centers and being awesome at that. And they managed to reduce their attrition from over 100% a year to less than 20% a year. And I actually had a talk with their CEO and because there was a large customer and I first was like, it takes you 45 minutes to apply to a freaking contact center job. Who’s going to do that? And they actually saw that almost nobody dropped out. And then she let me read this was before GDPR. So privacy wasn’t that high of an issue back then. And she let me read one of the emails from one of people she rejected and an email literally read. Thank you for rejecting me this way. This is the most You mean rejection I ever had it, she said, it’s probably the only rejection he ever had were no human touched the entire process because he was so well below our acceptable rates in certain traits, like your hand coordination. Can you quickly type while somebody is talking to you that the system rejected him, but would a report saying, Listen, you’re rejected. These are the limits we have on all of these traits, we’ve measured you. I’m sure you just unfit to be a call center edge. And this guy also wrote in his meal, I now understand why I was so bad at his job. And I understand that I need to find a different sector to work in.
Kevin
Great. Let me ask you this. Before we go on. Where do you what do you think about the use of the typical interview as an assessment tool.
Bas
And I’m not against it. But let’s just put it like this, we’re not going to get rid of it for a long time. And I don’t think we should get rid of it. But the problem is, as soon as we start the interview, usually we’re biased, because of things we read on the resume. And so if we start the interview with actual data on the person, it can be a really, really good way of continuing. But if we start the interview being biased about the person, because of a hobby this person had, which either annoys you, because he was a member of the same fraternity you were a member of or because you have to save all my mother matter or whatever. It’s not as effective as it could be. And the one thing, I also if you’re doing an interview, it needs to be a structured interview. And that’s where a lot of people are still having a really difficult time, as you well know. So
Kevin
So if you were to use an assessment tool in general, what type of assessment tool is, is there a best time to start with in terms of a sort of a general overview of what somebody’s capabilities are?
Bas
Now, it depends so much on what you’re trying to measure. And so, for example, right now, I’m advising a client or government who wants to recruit trainees, and basically these trainees need not have any specific cognitive skills, because they actually wanted as open as they can possibly make it. It’s, it’s about it’s a local government, and it’s about solving problems in the city, and who best to solve a problem in somebody who’s actually from the neighborhood. So it’s about how do we get people to work out more in this deprived neighborhood? Or how do we stop street crime in a certain neighborhood. But they want these people to be learning agile, and team players. And well, those things can be perfectly measured, for example, with a situational judgment test. Yet the Dutch air traffic control needs people to be really stress resilient. And that’s basically the main thing you need is as an air traffic controller. For that you need a cognitive test. And usually again, based test where you actually experience a lot of stress works best, because the question that nobody will say they are really bad under stress, especially not when you are applying to the air traffic control, where you notice stress resilience is probably the one thing you are selected on. And due to give you that example, as when the air traffic control in the Netherlands started using game based assessments, they use a tool called brains first. And they managed to reduce the number of people they needed to train in order to have noticed that they always had the same size class. So they always started with them. And they used to end up with two or three at the end of the training. And now they end up with eight. Because in that training, they’re really assessed a lot on stress resilience. And now they can of course, are many, many other trades. So is there a best not bad definition, it all depends on what you’re looking for.
Kevin
I’ve heard you mentioned games a couple of times now. And I’ve, I’ve heard some read some reports that a lot of candidates are kind of pushing back against game based assessments they don’t really like to take play these games. Especially I’ve heard that’s true of new college hires, or new grads that are applying for jobs. So the other name, insight onto that?
Bas
I’ve heard it as well. And I can tell you what I see. But this is purely anecdotal about people also, openly talking about how terrible they are. On average, they’re really the, basically the white privileged people from top tier schools who we who are now failing at a test. And I, and so I, I have yet to see these comments come from the underprivileged, to be honest. So I think there’s a part in privilege, I do understand, and this is a big problem, we have to solve as an industry that you don’t want to take the same test over and over and over again. And we should be able to use different to use our same tests and our report for different companies get we don’t, because everybody’s building their own algorithm. I can actually tell you, I recently reacted to a really nice interim job for assessing assessments for a company. And they asked 12 character traits. And I asked them, which assessments Do you accept? in me proving that I’m really good at these character traits? They said, Now we just want to resume.
Kevin
I mean, the resume will never die, unfortunately. But you know, the other question I’ve got is a lot of companies now are using video interviewing. And in some cases, they’re attaching assessments to those, to those video interviews, sometimes, passively, sometimes actively. What are your thoughts about about any of those tools?
Bas
Well, the one bit the ones based on micro expressions, and facial expressions, and facial recognition, right now I stay far away from I’ve tested a lot of these tools. And there was one which I’m actually which which came out pretty good. And I was the only one which wasn’t automated, which is interesting. There’s still a lot of manual labor attached to the artist. So but I would stay away from that, every one I’ve seen is either extremely biased, biased, or has other difficulties. And one of the things with facial recognition is it just simply works less on people of color, because have the same reason you’re probably wearing a black shirt right now, it has less reflection on it, which makes you look thinner. But in a face, it has less contrast, which makes it impossible for camera to really recognize the expressions. That’s why they feel a darker skinned people. The linguistic part is a little better. So if they start analyzing your word use, etc, etc. It is actually 70 years of scientific research into that, but it’s still not on the quality I would have for any type of high impact hiring, for volume hiring. And, well, I’m always saying it’s better than a resume. But then again, we’re not putting the bar very high. Now.
It’s another issue that comes up all the time. And this is not directly related to assessment, but it clearly is a is an outcome of assessment. Because we talk a lot about quality of hire. And, you know, this is an area that I’ve had a lot of issues with, because I’m not sure how we define quality of hire. And the traditional measures, I don’t think are all that accurate, like, like for tender or Quick Time to promotion. You know, these are getting to nothing to do with it, in my opinion. So how if somebody asked you, we want to improve our quality of hire. How do you respond to that? What’s your Do you have a definition of quality of hire? Well, I absolutely agree with you that quality of hire is one of the most difficult things to measure. I mean, there are a few jobs where they’re actually pretty simple to measure with salespeople, they’re pretty simple to measure. And what you actually see is that there’s a lot of assessment tools out there, which have been optimized for salespeople, which is awesome. And they can actually prove their results really easily. For those others, I always do ask, you know, what’s your definition of quality prior, usually, we end up with, like you said, a bad a bad proxy. Right now I’m working with one company, we’ve defined quality of hire of the number of people who start on a one year contract who then get a full, a permanent contract, which is, but that’s typically Dutch that you start off with a one year contract, and that it might be or might not be prolonged. Then your is one of the things. I’ve had some customers successfully positions where customer satisfaction was measured in there. But usually, it ends up with things like tenure or stuff like that. What I tend to do when we’re talking about quality of hire, is eventually we ended up not coming to a great definition, although in accountancy, and with lawyers, and they tend to go for billable hours, which isn’t, by definition, the best one, but it’s, again, it’s the least worst we have. But so what I’m then talking about is what are the traits you want people to possess? Yeah, but that’s really, that’s the best thing. Yeah. Yeah. And the beauty of when you start measuring it quality on trade level, so when you actually define traits, you start measuring the quality on trade level. And you take that as the input, instead of the resume for your, for your first selection, your pre selection, we’ve always seen diversity increase, which is just amazing. And, for example, there’s a KPMG, the Dutch branch of it, implemented an assessment for their early hires, so they did early career starter trainees. And what happened was when they started saying, Okay, these are the leadership skills we expect people to possess. And they simply increased the number of women they were recruiting by, I think, a quarter, they went up from under 30% of women to think just over 40% of the total number of hires, because the bias that a woman wasn’t showing the leadership skills they wanted, disappeared when they actually started measuring the leadership skills they want, which I think is really amazing.
Kevin
I just heard an interesting quote the other day that men are hired for potential women are hired for accomplishments. Now, I thought that was quite interesting research that it will give a man an opportunity because we say, Oh, he’s got great potential, but we rarely do that for a woman, we have to look at her. What is she actually done? Right?
Bas
And there’s actually a lot of research as well, Kevin, which shows that a man is allowed to have flaws that a woman isn’t. So I was actually I worked at a university for a while, which is interesting that this was working in the recruitment at the university, you find that it’s the most unscientific way of hiring I’ve ever seen in any corporate which is, I mean, I couldn’t even go bring up the subject of structured interviews was just nogo. They didn’t believe in it, which is really interesting for for professors. But what we saw there was that the main reason women were rejected, especially from the higher positions This was more than promotion part was a lack of leadership skills. The main complaint About male professors was a lack of leadership skills. And we had zero complaints about women’s lack of leadership skills for the HR department.
Kevin
And this is kind of the dilemma, I guess, of assessments versus personal opinions, biases, attitudes, it’s so much a bigger part of hiring. You know, when I look at a hiring decision, it’s maybe 10, or 15%, qualifications, and 80 plus percent, personality, attitude, dress style, background, and it’s very unfortunate, but that’s the reality.
Bas
Well, but personality and attitude matter, it The problem is we need personality and the attitude to fit what you expect from a person. So we know for example, that this was an interview from pu and pa, pa W. In in US fat people are more considered to be lazy and wild thing, people I’m considered to be a creative and have tenacity. And that’s, I mean, the only bias in the US which is increasing right now is beauty bias.
Kevin
It’s so interesting. And you know, I think you may have heard about the symphony orchestra was only only men being hired, right? And then they had every candidate play the instruments behind the screen. And suddenly the number of women doubled.
Bas
And then they had everybody come walking on a socks, because and then it actually became 5050. Yeah,
Kevin
Because they couldn’t hear the different shoes or whatever. Yeah, yeah, yeah. So much is so much bias.
Bas
But let me share another piece of bias with you which, which was amazing research from if I’m not mistaking MIT two or three years ago, they had two similar resumes of a man and a woman. And when they added, clearly less resume from a male candidate, the men would have been selected 100% of the time, when they edit a resume of a female candidate, which was clearly less, the female would be selected 60% of the time, because hiring is a risk averse business. And if you have two males and one female, the woman is the risky choice, because apparently, unconsciously, subconsciously, the men is the norm. And all of a sudden, when you change it to the woman is going to be the norm, the woman gets a serious chance. And we also know that this was Danish research from University of Copenhagen, when a woman said the exact same thing. So they basically put the same quotes and they just said somebody else said it. She was considered to be 20% less competent. Amazing, isn’t it? Absolutely. Yeah. And GitHub, did amazing research on code. Female coders, right? I mean, you think ones and zeros are as unbiased as it gets. They found out that a women whose names identified as women, so basically, Charlene 21, or something, were rated 20% less than the women voters who had a non female screening.
Kevin
I was just reading an article a little bit off our subject, but it is an interesting side note, that in the 60s and 70s 80% of all coders and computer people were women. And it changed in the late 70s, early 80s. and suddenly, it’s a profession that almost only for men. Fascinating research. Because in the early days of computers, it was considered women’s work. It was clerical, that, you know, coding was like a clerical job. And that women were much more suited to that like typing than men were right.
Bas
And then suddenly, do you know that computer actually used to be a job title? Yes, yeah, I know that. Somebody would do data entry was a computer.
Kevin
And they were all women.
Very interesting. So anyways, but just to change again, a 90 degree turn here. There’s a lot of talk today that a lot of companies are using AI chatbots at the front end of the recruitment process as a screening or even an assessment tool. What are your thoughts about that? And do you think that’s a legitimate and valid assessment process?
Bas
And it depends on how you do it. But in the basic, I can say, yeah, it could work. And if you look at the actual answers given, I know that one of the very first I know that this was a German company, and this woman said, Listen, I asked, she started off with four interns. She said, I asked all of my interns the same four questions, always, because I love structured interviews. And I’ll build a chatbot asking every applicant those same four questions. And I do a better pre selection, and I can have an interview, I can really go more in depth on the person behind those four questions. There are some who are now using linguistics behind it. That could work as well. I know there’s one company in the Philippines, who’s actually doing it on voice. So basically phone and they have a must admit really great AI, which almost real time understands what you just said, and is able to adjust the next question based on that. So they showed me a demo where the guys asked, like, what’s your main programming language? And he said, I’m not a big IT guy. But I think he said something like Ruby on Rails, and then Alright, so are you then familiar? Or do you prefer this or that framework, which are two Ruby on Rails frameworks, which was really, really cool. And what they mainly did, because their software, this was an Australian company with a lot of friends in the Philippines, that was it. They said, the Philippines have a lot of contact centers. Again, we go back to the contact centers, where with American clients, so we’re actually testing the actual level of English. And they’re going to get graded on how well they speak English. Which is, of course, really good, because you’re not looking at the last name, and oh, this person had some, maybe American family somewhere. And unfortunately, his name is still American, you’re now checking on how well does this person actually speak English? So those kinds of chat bots, I think, are really good. If you, you can make a structured interview in there. And the problem is with automated grading, some let’s just say there are a few which are able to do it well. But for most I’m not impressed.
Kevin
Now, that’s I just wanted to get sort of your ideas on that, because they’re becoming I think, more and more popular now. And clearly a way to reduce the volume of unqualified candidates that might be, you know, applying for trying to apply for jobs. It’s a great way to reduce the number of them.
Bas
Yeah, and what I personally always hope is that more qualified candidates are now getting a chance.
Kevin
I hope so as well. But that all depends on the quality of the supplier. Sorry. Absolutely. And how well those chat bots are constructed and validated and so forth.
Do you think there any kinds of assessments that are better for categories of jobs? For example, should a call center person be assessed on one set of criteria consistently? And should engineers perhaps on another set of criteria? Are there different types of tests that fit a category?
Bas
Definitely, definitely. I think actually, call center jobs is one of the best validated job Mark jobs there are. And it’s a combination of psychometrics and cognitive traits, because they’ve actually been able to really validate those because of the long history because they were one of the first to roll out at scale. And the great thing about call center jobs, you have so many applicants and so many hires and so much attrition that it’s really easy to measure. So what we see in call center jobs is a combination between a few specific cognitive traits, like your hand coordination, and Stuff like that. And some character traits. We know for retail jobs, it’s all about character. You know, it’s all about your attitude about how would you react to certain things. I’m personally a fan that it people are screamed at genuine coding tests. So how well can you go? How fast can you go? How good is your code looking how much processing speed is in there. And for all the other ones, usually what we do si is cognitive test. And I’m not talking about IQ test, here, I’m talking about specific cognitive traits tend to be the main qualifier of a candidate. So I’ve seen for example, it does depend on how the job is managed. So for example, I’ve seen corporate recruiters where you have a lot of hiring managers, one of the traits, apparently, which are really important to them is what they call fast switching. So are you able to switch between tasks, because you’ll have a hiring manager on your desk asking where to dam applicants are about twice or three or four times a day. So if you’re more of a straightforward, I need to schedule my own type of guy, you probably shouldn’t be in that type of recruiting position, you might be more of a social. And they’re in for example, if you’re looking more at the, the hardware engineering, and we know that analytical skills are really important for those, so and most of them, I would still say, if you look at game based assessments, different types of games, and again, there are different suppliers, some good, some less good, some really, really, really terrible are usually one of the most validated and most fitting for what you’re measuring. But again, in some cases, you just want character. And then usually, things like situational judgment tests are more valid.
Kevin
Where do you put personality testing in the spectrum of assessment testing?
Bas
Well, personality is what I show you what I call a character traits. So I would prefer usually personality testing based on situational judgment test rather than the old school questionnaire. Because you can manipulate a questionnaire how much they say it isn’t. And one of the problems I have with personality questionnaire based is they’re not unbiased, despite the fact that everybody says they are, when you’re doing international hiring, for the same job, they are not. And to give you a simple example, you know, as Dutch, we are never about the extreme. So on a five point scale, I will always say I’m either a two or four. The American will always say, easy to a one or a five. And I’ve seen this with actually one of my friends, he’s married to an American woman, she will always say it’s amazing, or it’s crap. And he will always say like, it’s good, or it’s unacceptable. But so, and I’ve actually seen this with one of my clients, they put a questionnaire based assessment in there. And it turned out that everybody would have bicultural background in the Netherlands, basically filled the character test. Now you’re not telling me that all of these people have the wrong personality? Come on. That’s, that’s statistically impossible. So that’s, that’s basically why I believe more in situational judgment tests for personality traits. And as well, there are a few games who can also measure personalities. Another question I’ve got for you is around another type of testing that sort of become popular certainly was popular before the pandemic, which was a culture fit test. What are your thoughts about that? Is it is a valid, I’m afraid my own opinion, is that it’s really about you know, do you look like me? Yeah. Well, culture, culture fit is probably the worst type of testing you can have. Absolutely. It’s do you look like me? Now, value fit is something different, but then we’re talking about actually saying how do we want somebody to respond in a certain situation? What are our values? And I can tell you, I had this discussion with a few boards. And we never were able to come to a conclusion what the values of the company were. I mean, everybody’s like, yeah, we need to be, we need big integrity. And then I say, Okay, so we have this situation. What do you do? Do you call your client out on it? You leave your clients? Or are you letting it be? An all board members have a different opinion about it? I’m like, Yeah. How can we hire fulfill you fit? If you can’t agree what your values are, everybody says integrity, but we now need to make it measurable. And so values, value fit, yes, cultural fit. Not that much.
Kevin
There is a professor at MIT, Ed Schein, who’s written many books on culture, organizational culture. He says there’s two kinds of values, the espoused values, those values that we talk about, and we print on the website, and everybody carries on their ID card. And then there’s the real values, the values that the company actually practices. And in my experience, most tests go for the espoused values, not for the real values of the company actually lives by. And that means that all the tests to your point are useless.
Bas
Yeah, ideal that doesn’t exist. Exactly. And you’re absolutely right, the main argument of having cultural fit is to kick out people who don’t look like us, it’s the worst argument you can ever have no value fit again. But then you need to make the values measurable, and have them be honestly expressed. And one of the things which I’ve noticed, I’ve actually done done a little research on this a couple of years ago, values tend to differ between different departments of organizations. And that’s something nobody wants to admit. So I once was at a company, and it was so clear that the marketing department has completely different values than the IT department and the sales department and the finance department, for example, on customer satisfaction, were try to put that into one test, or no was the main question I always had is, what are the values we want to have from From now on, we’re going to have finance and it also, because they were lowest on the customer satisfaction? importance, as you might consider, this was this was like, as traditional as it gets. I said, are we gonna hold them accountable to what we say our corporate values?
Kevin
I totally agree. So let me let me ask the final and multimillion dollar question. Are we going to be able to automate assessment? make it so that no humans need to get in touch with the person? Can I just have someone come to my career site, take some tests, and then I offer them a job, human unseen.
Bas
And this is actually important Amazon is already doing for the warehouse staff. So can it be done? For certain jobs? Yes. For other jobs, maybe in the future? Will it be done? I questioned it. And do we want it? Maybe not? Because the most important thing, I think right now with assessments is not to have the assessment decide for us. It’s to interpret results that come from it. And the one goal, where we’re not even researching on cracking, is team assessment. who fits best in this team? That’s the one thing which now at one of my clients, I’m where we’ve got a game based assessment as the first point of entry. And then they go to the hiring manager, say like, Okay, we’ve got 20 applicants. 10 of them seem to be pretty good fits. What do you need most in your team? Do you need a more analytical person? Do you need more team player because I’ve got these different profiles here will all fit but all in a different way. So tell me what do you think? think would be the best add to your team. And so a lot of them are like, yeah, yeah, we have a lot of team players. So don’t you want maybe somebody who’s a little less as a team player, but who score so much higher than the analytical, the analytical skills? Isn’t that a great ad to your team? And this, I don’t think we’ll ever crack the code of automating so. But let’s focus on what we humans are good at. Doing what it isn’t good at seeing greater picture seeing the strategic total picture of recruiting and the team.
Kevin
Couldn’t agree with you more about thanks so much for your time today. Really appreciate this sort of free flowing chat. It’s been great. A lot of fun. And thank you.
Bas
Thank you for having me.
Recent Comments