Book a Demo
SOVA science4hire feat image 2 Ep44

HR Tech's Impact on the Future of Talent Assessment Tools with John Sumser and Jeanne Achille

Today on Science 4-Hire, I’m joined by HR technology visionaries John Sumser, principal analyst at HRExaminer, and Jeanne Achille, CEO of The Devon Group and chair of the HR Technology Conference & Exposition. These experts have a decades-long history as leaders in the HR tech biz, and when it comes to predicting the next trends, I trust their judgment. So I brought them here to talk about an important topic: How are changes in the HR tech landscape shaping what’s next for talent assessment tools?

Understanding the future of talent assessment requires a broader look at HR tech in general.

“Where I think HR needs to continue to grow, and I think we’re really at a nascent stage,” Achille says, “is as it respects personalization of each employee and the depth of employee engagement that we really need to foster in order to build strong workforces.”

The problem for many organizations is that their data flow across HR functions is still blocked by silos. “Because the data doesn’t flow across those stacks,” Sumser says, “you end up with pocketed experts who don’t understand the entire big picture of the world that they’re operating with and don’t make decisions with all of the data that they could be making the decisions with.”

Want to find out what nuanced experiences and integrated data pools mean for the future of talent assessment tools? You don’t want to miss my conversation with John Sumser and Jeanne Achille.

Context Matters

Many providers of predictive hiring tools claim to have categorized hundreds of skills. But in many cases, what they’ve really catalogued are labels that differ across different contexts. “The way you develop software at Intel is different than the way that you develop software at Apple,” Sumser says.

And in an increasingly tight labor market, those distinctions matter.

“It’s not just skills, but it’s skills plus context that creates a lot of difference,” Sumser says. “And you want to be able to predict who’s going to do well in which context.”

Talent assessment tools that take context into account will be needed to deliver situationally specific results in order to have any real value. What’s needed is “a deeper lens into what constitutes a job, who’s capable of doing the job, what do you need to do to train them?” Sumser continues.

Predicting success also includes factors you can’t reliably predict during the hiring process, like whether a new hire will move into a new role down the line and end up with a manager or team they don’t get along with. “I think context is this elusive variable that we have a very difficult time managing,” Achille says. This is nothing new, it is extremely difficult to account for all the factors that drive human behavior. But the future is all about solving difficult challenges – so the gauntlet has been laid down on this one.

Process Changes Are on the Rise

Assessments gained popularity in the 20th century, when data was expensive, Sumser says. While that’s no longer the case, most assessment vendors still sell solutions that are separate from other recruiting processes rather than integrated products that bring together recruiting and assessment.

“Talent acquisition has always sat off to the side,” which segments the customers, Achille says. Because different aspects of the recruiting process have traditionally been siloed, talent assessment vendors have delivered tools to fit those niches.

Vendors today should realize they have the power to change the paradigm and go beyond traditional processes. “And I think now, as we’ve seen many times where the vendors drive change, if we are looking at recruiting and assessments being knit together,” Achille says, “we will potentially see the change within the customer.”

What does this mean for the future of assessments? We have a prime directive to figure out how assessment data can live beyond the hiring process and become a more important ingredient in all people practices. The current movement to more platform based tools that can integrate broadly across all organizational systems of record is definitely a start.

Hire for Potential Instead of Experience

The tight labor market puts more pressure on employers to fill skill gaps internally. With many companies drawing largely from younger talent pools with less experience, talent assessment tools have to evolve to help HR make decisions based on potential.

“I think we’re going to see the emergence of tools that help you understand how somebody who you don’t think meets your requirements can fit into your organization,” Sumser says. “And so I think it’ll be initially perceived as having to lower your standards, but what it really is is increasing the employer’s responsibility for doing training.”

As the labor market forecast isn’t expected to improve, it’ll become standard practice to build more — and better — training programs into the hiring process. “I think you’re going to see a necessary merger of learning and development and recruiting, and that may take the shape of internal talent markets and different outreach at the front end of recruiting,” Sumser continues.

One thing we know for sure is that, with so many variables in play, we need to prepare for the unexpected. The risk of a global recession only adds to this urgency, Achille says.

HR and talent acquisition teams need continued tech developments, Achille adds. “I think you’re going to see more and more reliance on data and on technology to put a degree of certainty into these unpredictable times that we’re going to be living in.”

People in This Episode

Catch John Sumser on Twitter and by email, and Jeanne Achille on Twitter and by email.

Read the Transcript

Announcer:

Welcome to Science 4-Hire with your host, Dr. Charles Handler. Hiring is hard. Pre-hire talent assessments can help you ease the pain. Whether you don’t know where to start or you just want to stay on top of the trends, Science 4-Hire provides 30 minutes of enlightenment on best practices and news from the front lines of the employment testing universe. So get ready to learn as Dr. Charles Handler and his all-star guests blend old-school knowledge with new-wave technology to educate and inform you about all things talent assessment.

Dr. Charles Handler:

Hello everybody. Welcome to the latest edition of Science 4-Hire. Today, we have a double treat for you all. We have two longstanding bastions of HR technology, analysis and commentary in action. That would be Mr. John Sumser and Jeanne Achille. So welcome, y’all, and I have been on your podcast, full disclosure, and I’m very happy to have you as guests today on ours. And I always, always open up by saying who knows my guests better than they do? So it would be a disservice for me to introduce them. I’ll let them introduce themselves.

John Sumser:

Well, hi Charles, and thanks for having Jeanne and I on today. My name’s John Sumser and I am the principal analyst at HRExaminer.com. I track the comings and goings in HR tech, and I have a particular focus these days on the use of AI in HR tech and particular ethics associated with the use of AI in HR tech. Delighted to be here.

Jeanne Achille:

Yeah, thank you, Charles. It’s always fun when John and I get to be guests on someone else’s podcast. So thank you for having us. And I’m Jeanne Achille. I’m CEO of The Devon Group. And The Devon Group specializes in communication strategies for HR tech solutions providers. As such, I’m pretty steeped in the category and touching really all different parts of the functionality. I’m also chair of the Women in HR Tech Summit, which takes place at the annual HR Tech Conference in the U.S. and Singapore. And I’m chair of the virtual HR Tech Conference that will be held next February.

Dr. Charles Handler:

Awesome. Well, thank you. So thematically today, what are we going to talk about? We’re obviously going to talk about HR technology, and I’ve had a lot of guests who’ve done that. It’s such a great and burgeoning industry that there’s plenty of people talking about it. I would like to really orient our conversation. I just want our listeners to know — I would say old-school in some way, but that’s not fully appropriate because your viewpoints aren’t old school. But there’s a lot of tenure here in this industry on the other side of my microphone. And I really want our listeners to be able to benefit from the long-term perspective of HR technology because it’s been growing fast, but it has been around a while. And I think there’s a lot of people in this space who were very young probably when these two folks started to get going.

I think there’s a really good — especially Jeanne being so involved with the HR tech conference for so long, you’ve really seen that at an inside level and probably have a macro perspective that’s super-valuable. And John has always shared really amazing insights. I always tell a story, I told it on your podcast, I’m going to tell it again.

I think it might have been 20-some years ago. I was meeting with John at his place and he said, “Hey,” he started showing us this thing. He said, “This is going to be a new thing called a weblog. They’re going to call it a blog.” And there were maybe two blogs that people had ever even done probably at that point. And so I always think about that when we see that blog is pretty much a universal global word at this point. So there’s some good crystal balls, if you will, going on here.

So let’s talk. What’s most important that our listeners need to kind of know about the HR tech space right now, especially with the lens looking back in time? Where are we now on the spectrum of our existence? Let’s get existential here.

Jeanne Achille:

Yeah, so you know what? I’m smiling because I’ve been in this industry since, dare I say, the 1980s. And my first gig was with a company called Inside, which was responsible for HRMS on the mainframe. And we had such revolutionary things like timesharing your HRMS on a mainframe. I think at this point, we’ve really checked the box in terms of transactional HR, but it’s taken decades. I mean, here we are, 2022, it’s taken decades. I’m not sure we’re doing everything perfectly, even from a transactional standpoint. Where I think HR needs to continue to grow, and I think we’re really at a nascent stage, is as it respects personalization of each employee and the depth of employee engagement that we really need to foster in order to build strong workforces.

John Sumser:

So I can echo a bunch of what Jeanne said, but I’d probably put a different spin on it and say from about the time that Jeanne got involved until just about now, we’ve been very busy taking paper forms and turning them into things that you can use on a screen. And that fundamental thing of how’d we do the process before and how do we do it now? Nothing really changed about those processes except we moved the location of them onto a screen. And so we have automated the paper process, and we are on the edges of being able to do something bigger and way more interesting than that. And the thing that we’re on the edges of being able to do is integrate all of the data from all of the HR tech silos.

The problem with that transactional view that Jeanne was talking about is that recruiting is recruiting, and payroll is payroll, and performance management is performance management, and assessment is assessment. And you’ve got all these little stacks of stuff and the data doesn’t flow across those stacks. So because the data doesn’t flow across those stacks, you end up with pocketed experts who don’t understand the entire big picture of the world that they’re operating with and don’t make decisions with all of the data that they could be making the decisions with. And so you see, I just made a list yesterday of all of the companies that use the phrase talent intelligence in their —

Jeanne Achille:

Oh good heavens.

Dr. Charles Handler:

That’s so funny because I wrote that down coming out of HR Tech. It’s like this isn’t even differentiated anymore.

John Sumser:

62 companies have it in their tagline. 62 companies. Now what’s interesting is the definitions aren’t really the same. It’s one of those loose words that’s going to pick up a solid definition one of these days. But at the heart of it is the idea that people, opportunity and development are sort of the heart of what HR is about from a talent management and talent acquisition perspective. And you need to make those pieces play together with really rich looks at people, which means you need to fold in the HRIS data and you need to fold in the payroll data into this singular view of what an employee is or isn’t. And that gives you the ability to look at what the real requirements are because we’re going into a time that we’ve never seen in the history of the species, and that is sustained permanent labor shortages at the entry level of the labor force. We’ve done it to ourselves in the United States by squishing off —

Jeanne Achille:

It’s the birth rate, it’s the birth rate and immigration.

John Sumser:

Yep.

Jeanne Achille:

I mean the facts are the facts, the numbers aren’t going to lie.

Dr. Charles Handler:

Well, when the robots take half the jobs, then it’ll all balance itself out.

Jeanne Achille:

I want to go back to John’s point about talent intelligence, though. When was the last time an HR professional asked you for a talent intel? Is it a solution that does talent intelligence? Is that really the way the economic buyer is thinking about this?

Dr. Charles Handler:

That’s a great point because, I wrote here in my notes, it’s awesome that we have the ability to keep all this data in one place, and we’re starting to really own that. If you start thinking about the buying decisions that companies use to get all these pieces in place, unless you’re starting with a clean slate, there’s all these legacy systems with different times they were put in and different people in charge of changing them. So it’s almost like even though the technology is there, how do you get the company to be able to coordinate and organize everything so you can actually take advantage of it?

I think that’s probably the harder problem. We could have technology that can do things that are absolutely dependent and amazing, which we do for a lot of the part, but how do you really extract the full value out of it? Is that what you’re saying?

John Sumser:

The times are what makes the difference. The times are what makes the difference. Of course, there’s legacy problems and guess what? Kodak, the music industry, all of these places that have gone to hell, had legacy systems in place that were impossible to change. And this move that we’re seeing towards unified data systems will be a differentiator between who used to be a company in the 20th century and who’s a vibrant company in the 21st century.

And the problem, which is hyper-competition for labor that doesn’t meet existing specs for that labor, is going to require a deeper lens into what constitutes a job, who’s capable of doing the job, what do you need to do to train them? All of these things, because we’re going to have to draw on parts of the working-age population who are currently not working to fill the holes or we’re not going to see economic growth.

Jeanne Achille:

And so are we talking about a skills approach to jobs here, and how does that evolve? I mean, Charles has raised a very good point about how messy things are. No one starts with a blank sheet of paper. So are we looking at some sort of transition from traditional job taxonomies over to a skills marketplace?

John Sumser:

I think that’s one approach that’s out there. And there are — I talked to the people at Eightfold, and the people at Eightfold claim to have 600 million skills. And what they’re really saying when they say they have 600 million skills is that the way you develop software at Intel is different than the way that you develop software at Apple. And the way that you serve a hamburger at McDonald’s is different than the way that you serve a hamburger at Burger King. And so what you want to think about with skills is it’s not just skills, but it’s skills plus context that creates a lot of difference. And you want to be able to predict who’s going to do well in which context.

So if you’ve got time management skills from your Uber job, they may not actually work as time management skills in your software development job. And so you need skills plus context to be able to understand what’s transferable and what’s not transferable. And that means a rich look at all of the people in the company. And there’s some of the same problems we’ve already had, always had, of getting at that data, but we won’t have as much choice because there isn’t an infinite supply of people to hire once you fire the ones you don’t like.

Dr. Charles Handler:

Yeah, and I think it’s a consistent thing, but it’s being magnified even more now, is that we have to be able to define very clearly the things about people that are important for the job. You could call them skills, but they’re not always skills. I think that’s another label that sometimes isn’t locked down. In my profession, it’s pretty clear because we objectively define different — I don’t want to get into it here, but knowledges, skills, abilities, competencies, they all have kind of their own little role.

But more generally when you talk about this stuff, it’s hard skills and soft skills. That’s pretty much how people chunk this stuff out. And sometimes there’s overlap there, but it’s really about understanding people. Again, I’m a psychologist, that’s what I bring to the table, but you can’t understand how someone fits in unless you understand who they are at a very objectively definable level. And then what are you trying to get them to accomplish? You got to break that down, too.

So to me, it’s the technology. The more that technology can do that without a lot of input from people, and do it accurately, and facilitate the connections of what fits, that’s a good thing. And we’re getting there, I believe, and that kind of opens the door to the AI conversation.

Jeanne Achille:

Listen, Charles, you better than any of us on this call understand the science behind defining skills. When I hear that a vendor is touting hundreds of thousands of skills, my question is, who defined the skills? Are they self-reported skills? It’s like I can put something on LinkedIn about myself that isn’t true. I can say that I can do basket weaving. So what kind of guide rails do we have for going to a skills model? What kind of guide rails do we have to know what is real, truly a skill that is a tested, a production-tested skill?

Dr. Charles Handler:

Remember the whole movement — I’d say this was 10 years ago — I really thought oh, digital credentialing, digital merit badges. The ability for you to either — there were companies that crowdsource those, there were companies that would just freaking give them to you if you wanted them, there were companies that make you earn them. But I kept thinking, well, if we have these tokens that verify that you have these skills that could be attached to you, the digital you somehow, well that would certainly be a good way to help people verify and sort things out. But I haven’t seen that really go anywhere, that whole credentialing, merit-badging thing. What happened to that?

John Sumser:

I’m not sure that the idea of credentialing isn’t some kind of a boondoggle because there is an enormous amount of data that’s pretty close out there. And that enormous amount of data is the database of all jobs that have been posted in the last 20 years and all resumes that have been published in the last 20 years. And so the great thing about machine learning and text extraction and natural language processing is that you can attack something of that scale, and it’s a massive, massive pile of data. My view is that that approach, which is the sort of high processing approach, gets you 600 million skills. And a skill is maybe as simple as understanding what it takes to be in job A to job B and sort of subtracting to get the difference there.

There’s a side of AI that says you can’t really solve this sort of problem with machine learning, that you need structured taxonomies as the baseline. And there are a few emerging players who have structured taxonomies at the heart of a big machine learning endeavor. And my guess is that they will work. And the way I think that works is you stick this thing that’s got a skill label on it, and you let it attract all of the data in the database, and you get a cloud that wraps around that thing, and you draw a line around the cloud. And that’s the skill. And then you have to, in the intake process, validate that somehow. But it’s easier if you have great definition and good interview questions. You can get through the credentialing question, I think.

Dr. Charles Handler:

Well yeah, and I think that I was thinking about it as credentialing being structured data that just goes ahead and says, “This is what’s real,” versus the unstructured amalgamation of all the different stuff that tells you the same thing. It’s a more direct path to do it with the structured piece. But again, that approach didn’t seem to really grab on. But what you hit on one thing, though, John, that I think is really important to talk about, and I’m interested in hearing from you all. And it’s almost a no-brainer, that just a pure AI-based approach is going to be too empirical, too mathematic, mathematical, really. You got to have some human touch in there. You got to have somebody working with it and kind of layering together the various — interpreting things and managing things. So I know that’s kind of vague, but it’s critical that people are helping work with these things.

And what do you all think about that? How is that reflected in some of the products and things that we’re seeing in the market today? Do you think it’s still over-indexed on AI? I mean if you, John, had said — probably even now, but go back two years ago to the trade show at Eightwell before COVID, let’s say three years ago, whatever the last really big one was. This year was pretty big, but a little bit of a downturn over a couple years.

But how many people just were leading with AI? AI this, AI that, we have AI. It was kind of like if you didn’t say that, you felt like you’re missing out and nobody would love you anymore. So now we’ve kind of tempered that a little bit. And I think you started to see, I felt like this year at HR Tech, a lot more of the “hey, we’re doing stuff as humans for humans, but we’re using a lot of cool technology to do it.”

So maybe speak to that, the spectrum of how much people are really feeling compelled to plug AI into stuff versus people being a little more humanistic about it.

Jeanne Achille:

That’s a very interesting question you’ve posed. I think there’s — we have to address the practical nature of doing business and how technology fits into the stream of business. And John, it sounds like when we talk about AI, are we talking about large enterprise or are you seeing more instances of it moving downmarket — let’s say into the midmarket companies?

John Sumser:

I haven’t talked to a HR tech vendor in years now who doesn’t have some AI angle in their product line. AI is so commodified at this point in time that everybody — we’re all used to seeing AI all over the place in our normal interactions with the web, and that stuff has permeated all of HR tech. So you get suggestions for your search, you get all that sort of stuff. The more interesting stuff — I think this move to talent intelligence is really just a rebranding of artificial intelligence, with the idea that artificial intelligence has this promise of a general intelligence, and talent intelligence is niche specific. And so you get more right-sized expectations about what’s possible with talent intelligence than you did with AI.

But I think we’re talking about the same fundamental thing, which is the ability to predict with some degree of certainty what will happen in a variety of circumstances. And that ability to predict is key to improve. Recruiting, for instance, has a 50% failure rate. Historically, nobody’s been able to fix it, and if you can get a tool that is better at predicting success by 20 points, the cost saving to an enterprise are enormous. And so —

Jeanne Achille:

That would be an amazing number. I can’t think of any vendor that can promise 20%. I honestly think if you get 2 or 3% —

Dr. Charles Handler:

It is true. It is because when we look at this, and this is a much more zoomed-in esoteric thing, when we look at this as what we call a validity coefficient, what’s the lift you get from a test over randomness or over what you’re already doing? Not randomness, but at what you’re already doing. That lift man, if you’re able to get, if we’re able to get even 10%, we’re heroes, but at scale, that 10% is millions and millions of dollars.

And ultimately, what are we dealing with here? We’re dealing with people. Predicting what people are going to do is really, really hard, especially when you put them in an organizational context and you give them a boss they don’t like and you couldn’t account for that in the hiring process and all of a sudden they’re leaving. Or now, in this day and age, you can crochet pot holders on Etsy and make a good living. You don’t even need to put up with all this BS.

Jeanne Achille:

Well, Charles, how did you know that that’s what John and I are planning to do next in our careers?

Dr. Charles Handler:

That’s right. That’s right.

Jeanne Achille:

You’re raising a really good point, and I’m hearing the word “context,” and I heard John mention the word “context,” and I think we would really be remiss if we didn’t dig into that. Because a lot of times you join an organization and you love the hiring manager and three months later that person decides to take an expat assignment in Hong Kong, and you wind up with a manager you hate or you wind up on a team that you hate. I think context is this elusive variable that we have a very difficult time managing and predicting.

Dr. Charles Handler:

Yeah, and it’s where the rubber meets the road essentially. I mean it’s the reality that you put these people into.

John Sumser:

Yeah. I think one of the things that we’re going to puzzle out over the next decade or so is, are big companies useful in their current form? And when you talk about the context problem in the way that you described it, Jeanne, you’re talking about a problem that’s endemic to companies with more than, say, 2,500 people. But when you have smaller enterprises, context is something that you can actually get your arms around, and something like 85 or 86% of all people work in companies that are under 1,000 people. And you can actually — there are only a dozen variables that make a difference in context between organization A and organization B when you’ve got things of that scale. Industry, location, capital infrastructure, product maturity: Those things tell you everything that you need to know about what it’s like inside of the cult.

Jeanne Achille:

So are we making an argument for the fact that the smaller employers value the individual more and therefore, Charles, do assessments become even more important in terms of bringing in the right talent?

Dr. Charles Handler:

You’d think that. I think that assessment tends to be a big company kind of thing. You don’t — now from a technology standpoint, it is moving downmarket more and more, but you look at things like who’s dominating. Job boards don’t really do much these days. You look more at more comprehensive things like an Indeed, who does have an assessment arm, or ZipRecruiter, who doesn’t have any assessment. So as you start to push things downward like that, there’s these big people that are supplying a lot to small companies that might be able to inject it in there. But I just don’t think it’s in the mindset of smaller companies.

But to John’s point, too, for whatever it’s worth, when you can come in and meet the whole company and have almost a more authentic experience, maybe it’s easier to see a point of connection. I don’t know, maybe I’m wrong. But I can say, serving up, there is tons of value in any hire you ever make in saying, here’s some objective information about this person and what they’re bringing to the table. Use it for what it’s worth, does it raise any red flags for you or not?

So the market is making assessments more available at that transactional level, but they are more watered-down. As you scale it, you can’t build a precise model of who’s going to be the ideal employee of the corner lemonade stand when there’s three people or something. It’s not as easy to do. So I think that those are the kind of realities of it. And as a vendor, and you’re trying to get your tests pushed down to that level, it requires so much advertising, so much investment in scale. I haven’t seen it work because it takes so much to get people to have that awareness broadly enough at scale that they’re going to start buying your test for $10.99 or $15.99.

Those are all available now, it’s just — there’s a big trend, I’ll tell you, from the assessment side of things. As I did my latest kind of review of hundreds of vendors, vendors are starting to make more transactional tools. Those that have been doing it a long time are packaging up their data and building pretty darn decent, easy-to-digest and easy-to-purchase kind of point tests or having a platform where you can pay and run someone through there and maybe organize your candidates a little bit. Those are available, but there’s a lot of them and there’s a lot of people. So helping people figure out that they are available and which one to use, that’s a bigger problem. It’s not a technology problem.

Jeanne Achille:

Well, before we move off the topic of assessment, as a small employer, I just want to say that every time I ignored the assessment test results and we got to the end of the employee relationship, I went back to the assessment and said, “Oh, facepalm, it was there from the start.”

John Sumser:

That’s interesting. That’s an interesting data point. I was just thinking that there’s a lot of pressure for things to change right now. There’s a whole lot of pressure for things to change. And the assessment business is a survey business, and survey businesses are paper-based businesses and 20th century in their orientation. They were built at a time when data was expensive. And the difference that assessment companies made over the years was based on, “I’ve got the data and you don’t.”

In a world of more fluid stuff that isn’t so transactional and isn’t so quick to segment into value based on who’s got the data and who doesn’t have the data, you might expect to see more interesting products hit the marketplace that look like job boards. And if you start working with a job board, and the people that you get to interview from the job board work in your company, then you don’t have to go through that thing of calling an assessment and having an assessment process. You just get sort of finished goods out the end of the process. And that’s part of what I mean about collapsing the silos is — why is recruiting and assessment separate? Why in the world are those two things separate? Why doesn’t the recruiting process provide you with an assessed thing and forget —I don’t know about you Jeanne, but when I look at assessment reports, I can never understand them.

Jeanne Achille:

Well, you have that bachelor’s degree in psychology, John. That’s what — I’m kidding. But I think you’re making an interesting point, but also take a look at organizationally how HR has been structured. Talent or acquisition has always sat off to the side. And then, of course, you have a team of I/O psychologists looking at things like assessments. So the customer themselves were segmented. And I think now, as we’ve seen many times where the vendors drive change, if we are looking at recruiting and assessments being knit together, we will potentially see the change within the customer, as well.

Dr. Charles Handler:

Yeah, I mean they already are in some sense. It’s just that you’re delivering in the most part a manualized survey to somebody to complete as part of the recruiting process, and then the data follows them. I think what John’s saying is you apply for a job, and we know a lot about you already in this structured organized way relative to the job. And for me, I’m not against that, but I will tell you that I don’t trust something with no psychology baked in. I don’t trust a purely empirical approach on telling me who fits. I need to have some structured measurement. Now that measurement can be completely stealth and not on the surface. I’m OK with that, but just pattern-matching or making assumptions that aren’t refereed somehow in terms of how the system’s put together. And I’m not saying that’s what you’re suggesting, John, but inherent in the idea that there’s no friction in the assessment is that the person doesn’t have to sit there and answer a bunch of questions, you know, kind of thing.

So we’re not there yet on that. And I do think that there’s pieces of that that are working well. And then of course, I don’t even know if I feel like with whatever the 10 minutes we have left opening up the whole bias, can of bias worms here, with AI. But I mean there are concerns there. One of the things I think, and then I’ll shut up here and listen to what you all have to say, but one of the things I think people lose a little bit of sight of when you start talking about bias and things is that humans are also very biased in our decision making. We have to be.

So it’s not like we’re this pure vessel of unbiased-ness, and AI and computer science, whatever, is chock full of it. It’s a matter that it exists for a reason, and for different reasons, I think, for those different modalities. But it’s something that, we have to kind of join those forces together to combat versus just saying computers — AI, excuse me — are just inherently always going to be biased. Well, people are, too, I hate to break that to you. So we need to structure things.

Jeanne Achille:

Yeah, I think that’s why we look to our, whatever we’ve identified as our North Star for validation. I’m curious if you see, and this is a question for both you, Charles, and John. Do you think that for example, startups, let’s say a high-tech startup, is less likely to look for validation and more to rely on their gut, whereas larger, maybe a Fortune 50, is a bit paralyzed because they’re so wedded to looking at data for validation?

John Sumser:

Well, I got handed an impenetrable 80-page Meta study the other day that looked at all of the validity estimates for all of the assessment surveys done over a cajillion years. And it was looking to find the most effective selection procedure. And what it finds is that structured employment interviews are the most predictive tool that you could use and that the things that we’ve traditionally thought of as the science are less predictive than a well honed, structured interview.

Jeanne Achille:

I’m curious though, when you look at structured interviews.So if I’m a hiring manager doing structured interview, I agree that that gives me certain guardrails. But if I’m a hiring manager, I’m not doing interviews every day or every week even. So is there an element of training that also goes into that to make sure that the person doing the interview has been trained to do the interview?

Dr. Charles Handler:

You betcha. I don’t know if they said that in that article, but yeah, of course. In my opinion, it couldn’t be any other way.

Jeanne Achille:

Yeah, I’d be really curious to see if that’s a variable in the study.

John Sumser:

Over the years, I’ve seen good systems that provide structure in the interview process as part of the HR tech that you run the interview process on, so that what you do is ensure you start at the beginning and go, “Here’s what we want to find out.” And then over the course of the interview, you make sure that you find that out.

Dr. Charles Handler:

As we kind of play it out here, last commentary, let’s see, what’s going to be different between now and let’s just say five years from now in the world of HR tech? Five years from now, what’s the fundamental difference going to be between where we are now and where we are then, if it’s detectable at all?

John Sumser:

Yeah, I think we’re going to see the emergence of tools that help you understand how somebody who you don’t think meets your requirements can fit into your organization — because we’re going to be drawing from pools of people who have been traditionally considered unhirable, unemployable. And so I think it’ll be initially perceived as having to lower your standards, but what it really is is increasing the employer’s responsibility for doing training, which we did up until the mid-’70s for new employees. And so we’ll bring some of that stuff back.

It means you’ll see tools for distributed apprenticeship start to really hit as we bring these alternative kinds of people into organizations and learn how to utilize them. So I think you’re going to see a necessary merger of learning and development and recruiting, and that may take the shape of internal talent markets and different outreach at the front end of recruiting. And I think that’ll happen fast because there are 4 million open jobs that they can’t find anybody to fill, that used to be filled by immigrants, that we need to do something with.

Jeanne Achille:

Yeah, I think John is absolutely right about training. I’ve been really excited to see the resurgence of training programs and focus on both formal and informal learning in the workplace. I think that that’s something that will continue to build in momentum. I do think though that if we look out over a five-year period of time, we always operate with a certain element of uncertainty, but I think we are in for some pretty interesting times ahead in terms of a potential global recession. And so I think you’re going to see more and more reliance on data and on technology to put degree of certainty into these unpredictable times that we’re going to be living in.

Dr. Charles Handler:

So my quick one is, I think we’re going to just see even more focus on tools that allow people to take charge, employees, job seekers, whatever, to take charge of their journey via data and access to the ability to find where they fit best and where they want to fit best. So it is a two-way conversation, and the technology’s really going to continue to support that through a lot of the things we talked about today. So thanks so much for an enlightening conversation. And very last thing, let our listeners know how they can follow and find you all out there in the world — John and then Jeanne.

John Sumser:

You can find me on Twitter, @johnsumser, or send me a piece of email john@hrexaminer.com.

Jeanne Achille:

And I’m also on Twitter, @jeanneachille, and DMs are open. I’m on LinkedIn, and you are always welcome to email me at Jeanne, J-E-A-N-N-E at devonpr.com. Thanks so much, Charles. This was fun.

Dr. Charles Handler:

Science 4-Hire is brought to you by Sova Assessment Recruiting software powered by science. Sova’s unified Talent Assessment Platform combines leading-edge science with smart, flexible technology to make hiring smarter and easier. Visit SovaAssessment.com to find out how your organization can provide an exceptional candidate experience while building a diverse, high-performing and future-ready workforce.

November 25, 2022
Sova Assessment
Changing assessment for good
Join our community

We’ll keep you up-to-date with the latest developments, events and insights.

© 2022 Sova Assessment Ltd. All Rights Reserved.