The Podcast for Employers Who Are Hiring At Scale
Are you involved in the hiring of dozens or even hundreds of employees a year? If so, you'll know that the typical sourcing tools, tactics, and strategies just don't scale. This podcast features news, tips, case studies, and interviews with the world's leading experts about the good, the bad, and the ugly when it comes to high-volume hiring.
AI and high volume hiring: the good, the bad, and the ugly
Keith Hulen is the Co-Founder and CEO of SmartRank, which stack-ranks and filters job applicants without using a resume. It helps employers achieve their DE&I initiatives by removing unconscious bias from the screening process.
Welcome to the High Volume Hiring Podcast. I'm Steven Rothberg, the founder of Job Search Site College recruiter. We believe that every student in recent grad deserves a great career. This podcast features news tips, case studies, and interviews with the world's leading experts about the good, the bad, and the ugly when it comes to high volume hiring. Thanks for joining us. Today's guest is Keith Huen, co-founder and c e o of Smart Rank, which quickly and easily stack ranks and filters job applicants without the use of a resume according to smart Rank. Eliminating the use of the resume from the screening process increases d e i initiatives by effectively removing unconscious bias. Keith, welcome to the
Show. Thanks, Steven. Great to be here.
So for the listeners who might not know that much about you personally give them a little bit of background other than smart Rank, who are you? What do you do?
Yeah, sure. So I've, I've been in business for little over two decades, and most of that time I was a hiring manager. So that's actually where kind of the impetus to create a company to solve some of these challenges came from, because that was certainly my least favorite part of being a leader was the hiring process. And so I knew that there was obviously dysfunction there and there had to be some low hanging fruit there to to solve for.
All right. And I love the fact that you've sat on the other side of the desk. You, you've much better appreciate. It's, it's, it's not just a tech vendor who thinks that they know what the world's problems are, but you actually know what the problems are. And we're gonna dive into some of the solutions that you've come up with. So let's talk about some of what you've learned about high volume hiring. What is it that you think that that talent acquisition leaders and professionals need to be aware of when it comes to high volume hiring?
Well, I think what's interesting is that it's still the same set of tools that people are using that don't have high volume. So what do I mean by that? I mean, they use job descriptions, they use resumes, and they are probably using some traditional applicant tracking system where they get to go through one of these applicants at a time. And I find that really interesting that, you know, the lack of automation that a lot of organizations are using, even when they do have high volume, I mean, I've spoken to people that have 2, 3, 4, 500 in some cases, over a thousand, and they still have to go through each applicant one at a time, open up a resume and compare the keywords between the resume and the job description. So I, I'd say that's probably the the biggest thing that was a, a wake, a wake up call for me is how, is there not more automation in this particular area, especially as it relates to high volume?
Yeah, boy, I mean, I can just think of so many analogies. It's, you know, in the conversation we had before this, you know, found out you're in Denver, you found out I'm in Minneapolis, call it a 13 hour drive depends what the weather in Nebraska's like, but it's, it's, it's typically about 13 hours. You know, if I wanna drive from my place to yours I'm gonna wanna be in a nice car for those 13 hours, not on a bicycle bicycle, get me there, but it's gonna be a real pain and it's gonna take me a real long time. And I'm not a professional cyclist. So no, I'd love, I love that the, the, the sort of the, the same tool, different problem is issue issue.
I, I just used an analogy the other day on LinkedIn. I said, you know, why would you drive 50 miles from your home and go grocery shopping at a gas station? I mean, sure, you can get a couple things there. You can get milk, maybe you might be able to get even some bread depending on the gas station and certainly some chips. But why do that? I mean, if you, why would you go grocery sh shopping at a gas station and why would you do it 50 miles from your home? That is essentially what I kind of see a lot of the folks doing. And, and, and there's a lot of problems that come along with that, right? Because, you know, the other issues you have is if you have 400 people that are applying, are you going through 400? I mean, most of the answers to, to those questions when I ask is no.
They're gonna go through the first 30, 40, 50, 60 maybe and they're gonna find the best people that they think are the most qualified and then pass those along to the hiring managers because they don't have time to go through 400 or 500. But, but this is the problem. What if your best applicant happened to apply 498th out of 500, you're just never gonna get to them. And even if you do, likelihood of them still being in the, in the hiring process is few and far between. They've probably already accepted another role. So there's, you're missing, most of the companies are missing the absolute top talent, and that's a shame.
Yeah. Some, some of the job boards out there will tell the candidate, this is how many people have already applied to this role. And, and, and I think it's because of that. The reality is, if you're 499th to apply, the likelihood of you getting hired is much lower than if you're third. And candidates have learned that, but there's really not a good reason for that. If, if you're using the right technology and those 499 applicants have all come in within a few days of each other, not over the course of eight months, then you should be treating the 499th applicant the same as as the third. That is, if you wanna hire the best people, not just the first people. What are, what are some other common misconceptions that, that TA departments have when it, when it comes to high volume hiring?
Well, I think, you know, that it needs to be done the same way that it was 20 years ago. I mean, there's very little there's very little improvement and innovation that I've seen in working with a lot of these companies, right? Like I just mentioned, they're all using those three core tools, job descriptions and resumes and applicant tracking systems. So it's a misconception that it still needs to be done the exact same way as it was done 20 years ago when at S'S first came out. Here's another big misconception. You don't need recruiters to do the screening. Now that sounds a little crazy. You don't, because here's what you have. You have a hiring manager that knows exactly what they want or they should know exactly what they want. If they don't, you're in trouble, right? Because and you need to find somebody else that does.
But the hiring manager should be your core resource that knows exactly what they're looking for. And then you have an applicant that you know, doesn't really know exactly what these qualifications are. Most recruiters don't know exactly what these qualifications are, right? You have an intake meeting that a recruiter does with a hiring manager, and they're gonna ask 'em questions like, what are your nice to have? What do your need to have? What are your knockouts and is there anything else I need to know about this role? And then they have a job description that they ask the hiring manager to review, which, by the way, hiring managers don't like that. They, they, that that's a, that's busy work for them and that, and that's not something that is fun because they also get no value from it. And then after that, you know, they're gonna have these applicants applied to these vague job descriptions because they say things like experience with knowledge of familiarity with that doesn't mean anything.
Nobody has defined what those terms mean. So it doesn't mean anything. And by the way, when you look at the resume and you see proficiency in X, Y, and Z and you see experience with that also doesn't mean anything cuz there's no context, there's no specificity to that. So, so you have a job description that you're comparing to a resume and you're looking at the keywords. Why do you need to pay somebody to do that? Right? Like, that doesn't make any sense. So that, that's a process that absolutely should be automated. And what you really need to do is remove the recruiter, get the, get the hiring managers, essentially screening a hundred percent of the applicants with exactly what they're looking for. And now you remove the ambiguity, you remove this time. The number one, if you look at data I know Aptitude Research put out a a, a research document about this not too long ago.
But you see it all over the place. The number one time-consuming task that a recruiter typically has is reviewing and screening applicants. Well, what if they don't, don't need to do that? What if you're giving them back? And that could r range anywhere from 30 to 50% of their time is being spent doing that. So what if that time could be spent doing other things like passive recruiting and going out and finding people instead of just looking through the people that have already applied? What if they are doing process improvement or data analysis or, or spending time with those candidates that they really want to join the company or they're hiring managers for that matter. There's a hundred things that they could be doing better than looking at resumes and comparing keywords to a job description.
Or even just bringing on that new hire in two weeks rather than in two months, because now they've got the time and the day. And that's also gonna improve their, their conversion rates, their application to interview, interview to hire, et cetera. Cuz they're not gonna lose those, those candidates. They're not gonna leak outta the process because they found a better job someplace else. You know, a couple things that you said a second ago, Keith wanna dig into. One is a couple of times you've, you've mentioned resumes and what I'm hearing is, and and, and certainly correct me if I'm wrong, what I'm hearing is that you feel like they're not necessary perhaps not necessary with with the right tools. Perhaps smart rank is part of that. Are you advocating for the elimination of resumes? Are you advocating for resumes to be used by recruiters and hiring managers alongside as assessments? What's your view on on that?
Yeah, so I definitely want to clarify here. I think there's a place for a resume. I don't think it's in the screening process. So yes, I think it's fine to have a resume. I think it's, you know, certainly if you're gonna be interviewing somebody, you would probably want to have a resume. And so there is a place for that resume to come in, but it's not in the screening process. Why? Because it doesn't give you any specific information you're looking for. If I see somebody, you know, says I, I was 110% of my quota. Okay, great. What types of clients were you calling after? What size of an organization were you going after? What level of the organization are you calling on? What were you selling? There's so much context that's missing from there. The way I like to say it is, resumes give you everything that the applicant wants you to know about them as opposed to what you really want is everything that you want to know about them, right? It should be the other way around, but that's not how it ends up working. And so you're trying to pull information out of these things, just ask them, just ask a hundred percent of the applicants out there and, and that solves that screening process.
We'll be back right after this break.
Back to the high volume hiring podcast.
A screening to say it another way is something that absolutely should be automated. It should be automated. There's no reason that you need somebody to open up a resume and a job description one after the next and, and just sit there painstakingly go through keywords trying to find something that matches that doesn't make any sense. It's not efficient and frankly, it's not even close to being effective either because almost every recruiter is not a subject matter expert for the roles that they're hiring for. That's just the reality. It's an impossible ask to ask a recruiter, I need you to be a subject matter expert even though you've never done this role ever in your career. Right? But I'm gonna make sure that I, I'm gonna ask you to be a subject matter expert for this. That's not fair to them. And if they're being honest with you, they're, the recruiters will tell you that carrying that burden is very stressful because they don't know. But they don't want to tell the hiring manager, I have no idea what you're looking for, but if they're being honest, they would tell you. They tell me that behind closed doors, yes. I don't know exactly what they're looking for. I have some words, I have some keywords I'm looking for, but I don't know exactly what they're looking for cuz I've never been a full stack software developer. How on earth would I know exactly what that is? Yeah.
At best they're gonna know a subset of what the hiring manager knows, and the likelihood is that, that the hiring manager isn't actually hiring for the same job as he or she has. They're hiring for a job for somebody who reports to them. And that person should know more about the job than the hiring manager knows. And then the hiring manager can only pass some of that onto the recruiter, and then the recruiter can only put some of that into the job description. And so you end up with a, it reminds me of the game that we all used to play with with kids called like telephone. You sit around in a circle and one kid whispers something into one the, the next kid's ear and then so on and so on. And so by the time it gets backs, the first person, it has nothing to do with it.
The you know, it does what you're saying about that. There's no reason for using a resume for screening. I think there was in 1952 before there was really technology and maybe even in the eighties, but it's been about 30 years now since the nineties since there have been automation tools and it keeps getting better and better and the technology keeps getting better and better. We now a lot of organizations like, like Smart Rank are using machine learning, even artificial intelligence. Are, are you, is smart rank using that or are you a proponent of that? Where, where do you guys stand on the use of AI in screening and and and ranking?
Well, that's a loaded question, isn't it Steven? That <laugh> right?
It is absolutely, absolutely. On, I'm on this side of the microphone. I get to, I get to ask the
Questions, <laugh> yes. Okay, so, so let's go into ai. But really quickly, fun fact, you were talking and, and I agree with you in the fifties, certainly in the eighties, absolutely. There was a, there was a need for that. You know, I've, I'm 43 my entire career. The resume has been the, the entire document that you use. Yeah, fun fact. The resume is actually 540 years old, so I don't know if you knew this, but Leonardo da Vinci you can look this up is actually credited with creating the first resume. You can actually find it online. You can actually see his resume.
Oh, I do, I do remember he was, he was being considered oh, for, he some like nobleman wanted to hire him or something, and he had to kind
Of dress the Duke of Milan. Okay.
And he had to like put together like a, I've done this work before kind of documents. That's right. Oh, okay.
Yes. But you know, I say that and it's a fun fact. It's also a little sad because we're still using this document virtually in the same way. Yeah. Right? That's the key phrase there in the same way that we were 540 years ago. There's almost nothing that we use in the exact same way that we did 540 years ago. So, but yes, over the last, you know, few decades. Okay, so let's talk about ai. I I think that there is a place for ai. I think that there are ways that it can be used but I think that companies need to be very careful about how they use it. And they need to not just listen to whatever somebody is telling them about how AI is gonna fix their problems. It is not magic and it's not just a silver bullet. Silver bullets don't exist.
You can't just plug something in and then put your hands up and say, oh, AI's gonna take care of it for me. Nothing that is that easy is going to, you know, give you the result or the impact that you're looking for. Anything that is worthwhile is going to have some effort and some level of work to it, right? So if it sounds too good to be true, might be right. So a couple things to consider when it comes to ai, number one. So again, there's pros and cons here, right? But you know, it can do things faster, okay? So if you are somebody that is, let's say you are not very, you're not a very innovative leader and you're somebody that likes to live in the status quo, okay, that's fine. Nothing wrong with that, but you want to improve the speeding up process of reviewing resumes. Sure. AI can do that and it can do it faster than a human. So if you just want to compare a resume to a job description and look at add to keywords, yes, it can do it faster, but it cannot do it any more effectively than a human can. So what do I mean by that? AI cannot tell you what that person meant when they wrote proficiency in Excel,
AI cannot do that. Yeah. And I sure hope that it never can because then we're in trouble and terminators are gonna come and get us, right? <Laugh>. it, it cannot give you the context and the specificity. And it also isn't gonna give you, you know, the objectivity and all those things that you really need to get. The only way you're going to know is if you say, Hey applicant, what does proficiency mean to you? Right? What, explain what that means. Oh, you know, I can put in numbers and cells and I can, I can even su do the whole summy thing. Oh, that's your level of proficiency. Ooh, okay, well that's not what we're looking for here. Do you know what a pivot table is? Nope. Never heard of it. Okay, well this job requires it. You do, you know, that's what's being missed in this whole process.
And so AI isn't going to do that. The other big thing that's very important for the listeners to understand is that AI can be notoriously biased. And there are number of, of, of cases out there. Amazon is one of them. Amazon tried to speed up because obviously everybody wants to work for Amazon. So they are high volume hiring at its best or worst, however you wanna look at it. And, and they tried to put in, and they have billions of dollars. So they have the technology, the resources, the money. And so they, they tried to do this and what they ended up finding, and they, they got rid of the program because they found that it was specifically biased against women. Now why? Because AI uses machine learning. And in order for some, for the machine to learn, you have to give it some inputs and some data.
So if you have 85% of your development team, I'm just gonna give an example. I'm not saying this was Amazon, I'm just saying any example, and it's all white guys over the age of 40. And then you plug those resumes into your database, that's what it's going to learn. And it's going to look for, you know, trends and it's gonna look for, you know, things patterns, right? And those patterns are things that it's gonna pop out on the other side. And you're gonna find that in some cases it can be biased. And by the way, I want to throw in, because this is also very important for the listeners to learn the states. And even at a federal level, it's, they are, they are putting that onus back onto the companies. If you are going to use technology from a hiring perspective that that automates some of these things like that, you are responsible. At the end of the day. You can't say, oh, well we use this tool and it does the AI and we don't know how the AI works. That's actually not acceptable anymore. California, New York Illinois has put in some some different statues. And even at a federal level with the algorithmic accountability Act, there is a number of things that are coming out and
That's as it should be, right? It's been mm-hmm <affirmative> decades since employers, at least in the US have been able to shield themselves from liability by hiring a staffing company to say, we don't wanna hire women, we don't wanna hire people who are black, we don't wanna hire people who are disabled, et cetera. Now, if there's a legitimate job reason for it, you're, you know, you're a Broadway show and you're casting a role for the, for the, you know, the, the lead actor and that actor is playing is gonna play a role where it's like a white male, you know, is it's, it's a, you know, a dad or, or whatever. Well, fair enough. You know, then, then go ahead and discriminate cuz it, the, and there actually is a job requirement. You know, if you've got a disability of some type that prevents you from repetitively listing, lifting boxes that weigh 50 pounds, you know, u p s, FedEx, et cetera, they can legally discriminate against you on the basis of that disability cuz they need you to do that.
But you can't make up that requirement if you're, if you're hiring software engineers, you can't say, oh, you need to be able to repeatedly lift boxes of 50 pounds or more, cuz that's not part of your job. But it, it was puzzling to me for a number of years where it seemed like employers understood that they can't use a staffing company to discriminate, but they can use software and, and so I'm glad that the law is catching up with the technology. It's, it's not there yet. So just to kind of like close the, the loop on on this last question before we need to le leave off. So your company smart rank, what's, how are, are you using machine learning? Are you using ai? If you're, if you're using them how are you finding that to be helpful?
We don't use machine learning on the front end. So that's where you see there's bunches of these companies that are scraping resumes and looking for the keywords, right? There's a hundred, there's hundreds of those companies. We don't do that at all. Our solution's, very practical. It just asks a hundred percent of the applicants exactly what the hiring manager is looking for. It, it automates that process at, at, at its core, right? So what that's doing is there is no AI happening, there's no magic happening on the front end. It's very practically just a asking every single applicant the very specific questions that you want to know. And then they, they have a set of answers there. What we use machine learning for is more on the backend. So once we've gathered that data, now we've already, you know, now we can start putting together, you know, things to say, Hey, by the way, based upon, you know, hiring manager, the people that you've liked and, and said you want to interview here is a you, you're 10 times more likely to probably select this person or, you know, based upon what you have done in the past.
Now again, you know, our, our solution, you know, is, is stack ranking. So at the end of the day, you're looking at the most qualified people and those are the ones that are getting sent to, to the hiring managers and that's specific to what they're looking for. But you know, it's not looking at, that's the key thing is like, it's one thing to look at all the data that you've collected on the backend and then be able to say, you know, things like you know, X percent are, are more likely to be fired if they select this answer. They're more likely to be hired if they select this, they're more likely to get passed on if they selected this answer. But that's not data that's making the decisions. That's just kind of like, hey, good information for you to know. And then there's a lot of areas that, that we can go down the road with that. But yeah, we're not using it on the front end. And I think that's the key thing for the listeners to understand is you, you, you gotta have a, a practical way of just getting these answers to exactly what the hiring manager's looking for if you really want to be able to make some change without getting yourself in trouble with AI
<Laugh>. Oh, very, very cool. So Keith, for listeners who might wanna learn more about you or smart rank how, how should they reach you?
Yeah, thanks for asking. So you could just go to our website obviously and www.smart rank.ai after I just sat there and talked about how <laugh> ai, but yeah, but that because again, we use it on the back end. But yes, so www.smart rank.ai or you can email us at info smart rank.ai.
Brilliant. Thank you very much for your time. It's been great talking to you.
Likewise, Steven. Thanks for the opportunity.
Thanks for joining us today on the High Volume Hiring podcast. I'm your host, Stephen Rothberg of job search site college recruiter. Each year we help more than 12 million candidates find great new jobs. Our customers are primarily Fortune 1000 companies, government agencies, and other employers who hire at scale and advertise their jobs with us. You can reach me at [email protected] The high volume hiring podcast is a co-production of Evergreen Podcasts and College Recruiter. Please subscribe for free on your favorite app, review it five stars are always nice, and recommend it to a couple of people you know who want to learn more about how best to hire at scale. Cheers.
Hear More From Us!
Subscribe Today and get the newest Evergreen content delivered straight to your inbox!