HiringSolved: The Use Of AI In Recruiting In A Post-Pandemic World With Shon Burton

William:  00:33
Ladies and gentlemen, this is William Tincup and you’re listening to the RecruitingDaily podcast. Today, we have Shon on from HiringSolved. We’re going to be talking about the use of AI in recruiting, in a post-pandemic world. And we’re getting dangerously close to a post-pandemic world, so it’s great that Shon is actually going to give us some ideas and some thoughts around what’s going to happen, what’s around the corner. So Shon, would you do us favor, the audience a favor and introduce both yourself and HiringSolved?

Shon:  01:03
Sure. Thanks William. I’m Shon Burton, I’m one of the founders of HiringSolved. HiringSolved’s a talent intelligence company. We’ve been around for not quite 10 years, but almost. And what we tend to do is automation and artificial intelligence in the HR tech space within your own tech stack.

William:  01:24
And so folks that, I mean obviously, y’all are extremely well-known. People use you for a lot of different things, right? They use HiringSolved to source candidates, but that’s just one use. What are some of the other uses right now that folks are using you for?

Shon:  01:43
We get a lot of diversity. Like you said, William, we really started in the sourcing world. Easier to find people. That started to branch out more into what I call analytics, and I guess the market is starting to call talent intelligence. Where it’s like, oh, it’s great to find people, but how do we find diverse people? How do we find diverse people that are in a consulting firm, rolling off a project within four weeks, that have these skills? How do we find people that we haven’t engaged with in the last 30 days or 60 days.

Or for example, with somebody like Lowe’s, these little nuggets of intelligence that are in the organization but aren’t necessarily being written down or in Forrest. Which is, hey, maybe we always want to talk to somebody who’s worked at Home Depot first, in all of our applicants. So we bubble that sort of stuff to the top automatically. That’s sort of the intelligence piece is, there are all these wonderful patterns that are being created by the humans in the company all the time. What we do is go in and discover them, and start using them, and bring them to light.

William:  02:49
What I love about that is, you’re looking at both candidate formation, what would be help top of funnel, and you’re also looking at employee data or employee information so that potentially it can help it with an internal mobility and internal movement.

Shon:  03:08
100%. In fact, I’m not supposed to talk about it, but we’re going to be talking a little bit more about internal mobility later this year.

William:  03:18
Good, well, I can’t wait. So AI gets kind of a, I won’t say a bum rap, but I think people we’ve talked about AI maybe a little too soon, years ago when maybe AI wasn’t quite there. It was a lot of machine learning, maybe a little NLP. But what I really want to get your take on kind of, when we get past the pandemic, what do you see? Some of the uses and some of the ways that people are going to use that both to create great candidate experiences but also great hiring your manager experiences, recruiter experience, sourcer experiences, et cetera. Like, if you put your futurist hat on, what do you see in terms of the true use of AI?

Shon:  04:06
Well, I think like you pointed out, and I’m just going to go ahead and call it AI because that’s what the market’s doing. But AI is not actually sexy. It’s not C-3PO, it’s not what we’ve seen in the movies. But what it is, it is something, like I said, that can make some of our work easier. It can make some sort of low-level decisions for us.

And I always think about it just as simple as a spam filter. Call it ML, call it AI. What it’s really doing is it’s reading your email and, in something like Gmail, it’s telling you not only what you should see and what you should never see, but also it’s categorizing it for you. Oh, well only pay attention to this if you’re interested in looking at your social stuff. So that, that really, to me, that is the basis. That’s probably the best example, that we all use every day, of functional AI.

It’s doing something that used to be specifically the domain of humans. Which is, deciding whether something’s important or not and kind of categorizing it.

Shon:  05:07
When I think about that from the HR tech perspective, it’s really very much the same thing. When we think about that email example, there’s a lot of patterns in that data and humans are great pattern matching machines. What AI is, that is based on these artificial neural networks, it becomes a great pattern matching machine.

So where that can take us when we think about a post-pandemic world, we’ve never seen the sort of demands that we’re seeing today on the employment side. I’m talking to customers but I’m also talking to just people on the street every day, and the businesses are dying because they can’t hire people.

Shon:  05:44
When we think about where does that go in the future, I think that the candidate and hiring manager or hirer experience is going to change a lot within the next few years, because there’s just still a tremendous amount of what I call toil. Which is sort of, it’s pain and it’s pain in not knowing what I can do.

Did you know that the burger stand down the street in Salt Lake City, Utah literally is paying $18 an hour right now for anybody that’ll sit there and cook food? Did you know that? And what’s that like, and is that really worth it, and do I want to do that? And to the hiring manager, is that enough? Is that going to bring people in? What kind of people is that going to bring in? That sort of stuff.

Shon:  06:31
So when I think about a post-pandemic world where you have this friction point that’s been put on us by not having enough workers in the market anymore, and really trying to get business re reignited, I think it’s a fantastic opportunity to bring out new technology that starts to lower that toil, lower that strife. And it’s both on the candidate side and the company side, where you’re bringing in that power of analytics, you’re bringing in that power of sort of automation.

But again, it’s not C-3PO. It might take the shape of a little blinking light or it might take the shape of a text message. It’s going to be something fairly simple in the way that we think of as a spam filter, but it’s going to be very helpful at the end of the day. Connecting people with jobs faster and more reliably in jobs that they want to do, and letting people understand what their opportunities are and kind of play the market, so to speak, so you can really understand.

Shon:  07:28
And we’re talking to our daughter about this all the time. It’s like, hey yeah. It’s funny, we just did this. We were on a road trip and she’s like, hey, here’s three professions. She wanted to play this game. Like, what would you rather do? Would you be a veterinarian? Or something like, it was a dog walker, a tax accountant, or like am In-N-Out cook.

And we had this great conversation trying to help her understand like, yeah, what might be the life of those people? She loves dogs, so she wants to do the dog walker. But there’s not a lot of available information of what that really means. And how does it really impact your life, and how often can you really eat sushi if you’re the dog walker? Maybe all the time, I don’t know. But I said, you got to figure it out. And I think these tools are going to help us do that.

William:  08:14
What I love is, you’re imagining a world where it’s helping the candidate experience. It’s helping with recruiters in terms of getting some of their time back. So it’s displacing some task and some things that you called a foil, which I think is great. It’s friction. It’s, maybe we’ve done it for 20 years and that’s just kind of how we work. That it’s getting rid of some of that stuff so that they can spend more time with candidates and also deal with volume and deal with more candidates in maybe even a deeper way.

William:  08:50
Let me ask you kind of the candidate difference. Like, okay, pre-COVID, if you will, or pre-pandemic. We weren’t as fast as candidates. Like, you talk to almost any recruiting team, December of ’19, January of ’20. And by the time we got back to a candidate, they already had another offer, four other offers, a job, they’d already started somewhere.

And like, we just weren’t fast. I think there was a real problem with speed. Whether we go through the pandemic and on the other end of that, and even for some industries, it’s still true, that we’re still not as fast. So where do you see the candidates and how they’re different in a post-pandemic world? Like, what are their needs or desires or attributes, or just, how are they different in the way that they interact with hiring?

Shon:  09:47
I mean, I think what’s really interesting about this post-pandemic world is that it basically hit the reset button for everybody. It paused the rat wheel, hamster wheel, and it let us just step off of it for a minute. And now everyone in the world, I think, is looking back at it, going, hang on a second. Now, do I want to get back on that wheel, and start doing the same thing as I was and just running in place? Do I want to get back on that treadmill or do I want to do something different? Is there a new opportunity for me? And it’s really an interesting, just I mean, I’m sure it will be phenomenal in 10 or 20 years when the anthropologist sort of figure it all out.

Shon:  10:27
But I think from an employment perspective, what it’s doing, like you said, there’s still companies that are not anywhere near fast enough. I’m sure you see it, William. I mean, I’ve been all across the United States in the last couple months and lots of stuff is still at half capacity or less. You go to a hotel, they can’t clean your room. Right. I’m sort of privileged, so I get to see it from that angle. But whether that’s, you can’t get a place in the restaurant or you can’t get a bagger for your groceries, we’re seeing it all over the place.

And I think one of the things that’s happening, like I said, is just, people are starting to figure out, it’s a great time to reassess, what am I potentials. On the speed side, one thing we’ve noticed in a post-pandemic world, just on the professional side, the business side is staffing is crushing it.

Shon:  11:24
I think the reason staffing’s crushing it is because they are fast, they’re motivated to be fast, they get paid if they’re fast. So we’re still seeing exactly what you’re talking about and we see it tremendously in automation and AI. Big corporations, particularly, are slow to adapt. And even though they might be dying on the vine and running their businesses at half capacity because they don’t have the employees to fully staff them, they’re still concerned about compliance.

They’re concerned about, what about bias in AI? And they’re just having a hard time even defining it. So they kind of go to staffing firms. And I’m talking about across the board. It doesn’t matter if it’s a staffing firm to fill up the best bias so that they can sell cell phones again because there’s a line out the door, or a restaurant or whatever it is, or an SAP administrator. So we’re seeing that. We’re really focused on staffing firms because of that right now, because we’re seeing a lot of growth there and we’re seeing the large corporates still slowing down.

Shon:  12:27
On the other hand, like I said, literally, that was a true story. I was in Salt Lake City, Utah, where my mom lives, and driving down the street. And this place called Arctic Circles, this old burger stand that was there when I was a kid had, instead of like, burger and fries for $6.99, the sign said we’re paying $17-$19 an hour. Please come in. Apply within.

William:  12:52
Wow.

Shon:  12:52
So they’re moving fast. Those smaller businesses are actually moving really fast, and I think that’s also super interesting. And I’m seeing that everywhere too. Seeing the sushi restaurant nearby who’s still not at full capacity and I love. But they’re like, look at how they’re advertising to the candidate now. I mean they’re saying, wait staff can get 401Ks here. That’s outside on the window. Amazing, right? I’ve never seen anything like it.

William:  13:20
Well it’s also, you’re dealing with supply and demand or scarcity and surplus. And again, some industries that have been hardest hit through the pandemic are having to figure out really creative ways to refill that talent. And also, to keep that talent because you’re hearing the same things I am about people that will accept a job. T50% of those folks start the first day. And then you go through onboarding and 50% of those people will go forward. Like, it’s crushing, especially in high volume areas, that it’s not as easy as it was pre-pandemic. You hired somebody, they started, they got onboarded, and they worked a gig.

William:  14:08
Well I think, again, some of this is just, we’ve been through trauma. And I think you’ve really nailed it when you said, this is kind of a great reset. Like, we as candidates have figured this out. We can behave differently. It’s choice. We can go back to the things that we were doing prior, but we don’t have to. I want to get your take on kind of, if you and I were building a Maslow’s hierarchy of needs, in terms of AI and for recruiters.

And let’s deal with the corporate folks, because the staffing and RPOs, you’re right, they’re hard wired to get price, quality and speed. And so they’re going to innovate, if for no other reason than their clients are going to demand it. So I’m not worried about them. I am worried about the corporates, and those recruiters, hiring managers, sourcers, et cetera,. Getting over whatever issues they might or might not have with AI.

If you were building that triangle, where would you say, okay, start here and then get that done, and then build this or buy this and then get that done, and then kind of move your way through. Like, how would you move a reluctant global head of TA through this process, post-pandemic?

Shon:  15:38
That’s a great question. What we see, William, is that there’s actually just a ton of low-hanging fruit. I would say the biggest area of strife that we see is they really, really want automation because the workload is crushing. The applicant load is high. But there are some major roadblocks. Just super simple roadblocks, but roadblocks that we see all the time.

One of them is, you’d be, well, you probably wouldn’t be, but I am surprised. I’m very, very surprised when we talked to these great companies. And I mean, these are world-class technology companies, they’re world-class retail, whatever. And they’re saying, hey, we want to automate. And we look at their data and what we find is, there’s just nothing there. Like, they won’t have. I mean, there’s stuff we can do.

William:  16:36
Right, of course.

Shon:  16:38
But I’m talking about, well, we’ve got 800,000 candidates and 60% of them don’t have a resume. Or I mean, literally, so what we do now is, everybody we talk to goes through what we call a data quality assessment. Because as soon as we start talking to AI and automation, the cat now is out of the bag.

And I’ve been saying this for years, but what you don’t get is C-3PO. He doesn’t show up and fix all your problems, and just magically wave an AI wand and everything’s fixed. What it really is, is just, it’s like all other automation, it’s putting a conveyor belt on some piece of the machine that maybe moves things faster.

But if those things are not the right things that you want to move faster, then it’s going to make mistakes faster. So what we see is, for example, no contact information. Literally no way to contact. No social, no email, no phone. On globally, we see this in at least 10% of all records. In some companies, it’s 30%. So there’s these little things that they can do. Another big one is just asking, like everyone seems to have gone the other way and said, oh, let’s make the apply really, really simple. And that’s great for the candidate experience.

Shon:  17:52
But if we’re talking about, we’re hiring a nurse, we’re hiring a commercial trucking driver, we’re hiring somebody for Lowe’s. Asking those people. For the nurse, are you a registered nurse? In what states and what’s your license? Sort of looking like a structured data coming in. So it can’t be, what we’re seeing is these simple mistakes where they might ask that question and then it’s a free form text field where a person can literally write a story.

Well, I was certified and I had a license, but then my husband got sick and I had to take care of him, and he’s also a nurse. And then there’s this big long story. And when you go to automate that, that is just garbage. That takes tremendously more effort for the machine to figure out. And by the way, your AI vendor might not tell you this, but there is always an error rate and it’s pretty high. I mean, you’re talking about.

William:  18:42
Right.

Shon:  18:44
So simple things like that.

William:  18:47
I love where you started, Shon, because you didn’t start with the sexy things, which is what most AI vendors would start with. They’re going to go with all the sizzle. And you started with something that’s relatively boring, but critical: get your data right.

Shon:  19:05
Get your data right. Yeah. Figure out that, you don’t have to ask them and they don’t have to fill out a 15 field form. I think it’s wonderful that we’re not doing that to people anymore. But ask them three questions. How about we classify jobs in classes and then we ask them three questions, and we bring that data in as structured. Do you have a commercial driver’s license, yes or no? It’s a radio button, man. You can’t tell me a story.

William:  19:29
Right. In fact, the story works against us.

Shon:  19:32
Right. The story works against us.

William:  19:35
I love that.

Shon:  19:36
[crosstalk 00:19:36] I know what you can create is this beautiful structured data, that then, there isn’t an error rate because you asked the person and they said yes or no. Of course, they could be lying on that, but, that’s another stage.

William:  19:47
Well, I mean, again, the data is not, there’s no errors in the data. It might not be accurate because somebody lied, but that’s an altogether different issue.

Shon:  19:57
Yeah, that’s further down the pipeline, right?

William:  19:59
I think some of the things that are really interesting after they’ve gotten. So again, we’re building that pyramid of the hierarchy of needs and how, the corporate recruiters in particular, how they take this on and start to think of these things. I love that they get their head around automation, which again, there’s so many different ways to think of that.

Conversational bots on the front end that can ask probative questions. They can knock out questions or communicating with candidate. There’s scheduling tools embedded in some of these things, that are smart. They can schedule across multiple schedules, so I think that’s great. Screening, followup. Just, there’s all kinds. When you say automation, that’s an umbrella for a lot of different ways to give time back to the recruiter so that they can do the things that they need to do.

Shon:  20:59
Yes. And focusing on, yeah, it’s simple stuff. It’s like, if we think about a chat bot or a text, a simple SMS system that asks those questions. I mean, we see the same thing in internal mobility. The biggest challenge in internal mobility is data. Well, this person’s been an employer for 16 years and we have a six year old resume.

And it used to be, when I started in this, the thinking was, well, we’ll just match that person’s LinkedIn resume or LinkedIn profile, in there against their will. And that will be their. Well, you can’t do that. So it turns out, that’s not actually that cool. So what if we sent a text message or two to everybody and said, well, do you have this experience? We’re considering you. Are you interested, yes or no?

William:  21:45
Right. Yeah, you’re asking, there’s consent there, which is baked in, which I think is really smart. But again, you’re also, subtly, you’re also saying, hey, we care. You don’t have to go somewhere else to get to another place. We have plenty of things, and you might not have even thought of and we might not have historically thought of you for. Are you up for it?

Which I think is just genius. And I can’t wait, well, we’ll have another conversation down the road on internal mobility. Let me ask. And again, some of this might be historic, but some of this is also in our future. Do we talk too much about how AI works, with practitioners?

Shon:  22:31
100%. Oh my gosh. Thank you for, I mean my God, a voice of sanity and the darkness. Yes. Oh my god.

William:  22:40
First of all, no offense to anyone in HRTA. I love you. But I don’t think they care.

Shon:  22:42
No.

William:  22:44
I think they care that it works. Like, Alexa, what’s the capital of Ohio? I think they care that the result gets them what they need, but the zeros and ones, and a lot of the technical things, I see a lot of firms really spend a lot of time, money, and energy talking about the how.

Shon:  23:09
So I’ll make a confession on this, because I was definitely part of this problem. My confession as an idiot vendor is, we always got asked, and still do, that question. And when I was on all those calls, how does it work? How does the AI work? Yeah, I used to, oh well, it’s got a genetic algorithm that models…

William:  23:30
Let’s geek out.

Shon:  23:32
Yeah. And it was interesting because they never said no, no, no stupid. That’s not what I’m asking. So I would just rattle it on and just keep going. And man, you know me, I can talk. So it took me probably eight years to realize, oh my gosh, when they say, how does it work? They’re not asking, they don’t want to know. Like you said, they don’t really care what the algorithm, they want some confidence that it works and they want some understanding of the use cases. And they’re just saying, what does it do and how can I be sure that it’s going to work for me?

William:  24:07
Yeah. I think the only thing, and you know this as well, is another tool. One of the things that they’re not saying that they should probably say more often is, I’m afraid that my people won’t use it. It’s the greatest thing. I love the demo and I can see it immediately. I can see what it would do to help us, and give our time back, and make us more efficient, and create a better experience and all of those things. But I’m worried about my folks actually utilizing it and using it.

Shon:  24:41
And it’s such a valid concern. Because it’s funny, we look at this from an HR tech lens, but the truth is that over half of enterprise software doesn’t ever get used. I mean, however you measure that, half the licenses, half the packages.

William:  24:54
Right.

Shon:  24:54
So it’s such a valid concern and that’s something that we’ve been. It’s funny, I just got back from Chicago from a strategy meeting and that was the single biggest thing that we came away with. Was, man, we’re, we’re asking these people to do yet another thing. And it’s really hard to provide that value curve. And I think you hit on something, William, that we see.

Which is, if you’re talking to an executive level, you can probably sell them on a lot of pretty cool charts and graphs, and stuff, and case studies, and you can make the sell. But when the recruiter is forced to use it, it’s like, none of my sales guys will use Salesforce. They hate it. And I hate it. It’s not going to get used. But we buy it every year and a lot of those licenses go unused. So that’s a great point.

They should be saying that, is, how do I know people use it? What’s your plan to get my people to use it? Because realistically, I can yell and scream. But at the end of the day, these recruiters are under the gun. They got to make a hire. And if they can’t instantly understand that this is going to help them do that. And I mean, instantly, the first time they log in, they’re not going to use it. Doesn’t matter what we say.

William:  26:08
No. I think we’ve gotten past the fear stage phase of AI. So one of the things I’m glad and I’m hopeful is that, there was a lot of literature around how AI was going to take my job. And you see less of that stuff. First of all, it was all fear mongering anyhow. But it was always going to augment, and it was always used in and thought of as an augmentation in helping you get some of your time back. But we’re past that idea that this is going to rob me of my job. It’s just going to help you do your job better.

William:  26:49
Shon, I could, I could obviously talk to you forever, but I know that you’ve got other things to do today. And I appreciate you, A, giving us a glimpse of what you think is going to happen post-pandemic, what you see happening post-pandemic. And for the audience, if you take nothing away from the call today or the show today, is thinking about structured data and thinking about your data in a different way so that you can do all the things.

Because it is garbage in, garbage out. You want AI to be able to help you make great business decisions, but you can’t do that if your data is, let’s just say, not great. We’ll keep that generic.

Shon:  27:35
Yeah. I want to leave you with this, William and the audience, is that, remember those scary videos of the robots, the Boston Dynamics robot? The dog? They’re all super scary, right?

William:  27:46
Right.

Shon:  27:46
And we see them doing stuff that, for 10 years, really looked like, oh my God, that thing, it can do back flips, it can roller skate. It’s definitely going to take my job. It’s AI, right? So this is the one piece. Is that, Boston Dynamics, as cool as it is and as much as I love those videos, it got bought by Hyundai for a billion dollars just recently. And they’ve been for sale for a long time. So as much as it looked really scary in those videos 10 years ago, it looked like it was just totally going to happen, slack got bought for $27 billion.

The difference is, Slack as a company that generates intelligence. It generates intelligence from every single employee by creating structured data, creating intelligence, non-structured data. Boston dynamics was trying to move the needle forward, but we’re yeah, like you said, it’s not taking our jobs. At best, it makes things a little better. I can search from my payroll calendar in Slack now. That’s awesome.

William:  28:46
Right.

Shon:  28:48
And a bot tells me, you’re going to get paid tomorrow.

William:  28:51
Thank God.

Shon:  28:52
Yeah.

William:  28:52
Well listen, everyone, thank you for listening to the RecruitingDaily podcast. Shon, thank you so much for coming on and breaking things down for us.

Shon:  29:02
My pleasure. Thanks, man.

William:  29:03
Alrighty. Take care.