20 Recruiters Making 150,000 Hires with Shay Johnson

In this episode of the Chad & Cheese Podcast's AI Series, hosts Chad Sowash and Joel Cheesman sit down with Shay Johnson, VP of Strategic HR Partnerships at Compass Group. Managing over 7 million applicants and up to 200,000 annual hires, Compass Group is redefining high-volume recruiting at scale.   

Shay breaks down how the global hospitality giant evolved from manual, grassroots hiring to leveraging cutting-edge talent acquisition tech from native job-board integrations and automated re-engagement, to implementing a full end-to-end conversational ATS.   

Tune in to discover how on-demand AI interviewing agents, fraud detection, and candidate coaching are filtering out the noise, capturing rich talent data, and elevating recruiters from tedious admin tasks to high-value human experiences.   

PART 2 - CAN AI INTERVIEW CANDIDATES BETTER?

PART 3 - AI DIDN'T REPLACE RECRUITERS

PODCAST TRANSCRIPTION

Intro Voiceover: Hide your kids, lock the doors! You're listening to HR's most dangerous podcast. Chad Sowash and Joel Cheesman are here to punch the recruiting industry right where it hurts. Complete with breaking news, rash opinion, and loads of snark. Buckle up, boys and girls, it's time for the Chad & Cheese podcast!

Joel Cheesman: All right, let's do this. We are the Chad & Cheese podcast talking today with Shay Johnson of Compass Group. He's the VP of HR Partnerships. Shay, welcome.

Shay Johnson: Thanks, guys. Happy to be here.

Chad Sowash: Glad to have you.

Joel Cheesman: Again!

Chad Sowash: Second time, yeah, second time on the show.

Joel Cheesman: Yeah, three more and you get a velvet jacket with our faces on it.

Shay Johnson: Plush.

Joel Cheesman: Aren't you lucky?

Shay Johnson: Nobody wants that. Nobody wants that.

Joel Cheesman: Yeah, North Carolina, it'd be a little warm, the velvet, but hey, we could do that. So, for our listeners who didn't catch our last interview, what should they know about you and Compass Group?

Shay Johnson: Yeah, so, a little bit about me. VP of Strategic HR Partnerships. I spent the first 10 years of my career, all with Compass, by the way, I've been there since I got out of college, spent it in TA, mostly focused on our technology partnerships, implementations, and strategy. So, now I do that across all of HR for Compass Group. And for those of you that don't know, Compass Group is the largest contract food service and facilities management company in the world. So, about 300,000 employees in the US, about 600,000 globally.

Joel Cheesman: So, when I go to a Colts game and I get my extra long dog, it's usually a Compass employee who's serving it to me.

Shay Johnson: For sure, and sometimes we control the pricing, and sometimes we don't.

Chad Sowash: I think that's a monopoly, Chad. I think that's called a monopoly.

Chad Sowash: So, 10 years. Nothing's changed in that time in talent acquisition, so it's been easy. Literally, so much has changed just over the last six months to a year to 18 months. For you, especially with Compass Group, there's a lot going on. How many hires are you guys making a year? How many applications are you receiving in a year? Let's just start with that base.

Shay Johnson: Yeah, I mean, from a hiring volume perspective, it ranges, obviously, just given what kind of year it's going to be and what kind of business, but anywhere between 160 and 200,000, if you're going all in, frontline, salaried professional, everything like that. Seasonal doesn't count, like our temp labor and gig. But from an application standpoint, that's where it's really interesting because it used to be really hard to get, you just wanted anybody with a pulse to apply. And now we're in the overwhelming stage, right, of I think this past year we had over 7 million applicants. So yeah, reverse engineer the...

Joel Cheesman: All real humans, I'm sure. All unique, real human beings.

Shay Johnson: Exactly.

Chad Sowash: We'll get into that.

Joel Cheesman: Okay, sorry. I'm so excited.

Chad Sowash: So knowing that, what has changed in the last 10 years, especially here we're at the AI sessions, how are you guys using new technology to manage all of that?

Shay Johnson: Yeah. Yeah, I mean, well, 10 years, that's a long way to go back. I'll try to go back, maybe I'll just go back five. Yeah, when I go back then, we'll call it pre-COVID, you know, at that time I feel like it didn't even feel like we had to do that much recruiting. We had a pretty steady pace of hiring, I think things were in a mellow place. And so it was more about just running your jobs through an ATS, right? Got to get the job posted, not a lot of your applicants, our applicants at the time, came from online. Most of our hiring happened through hiring events, through word of mouth, through local partnerships.

Chad Sowash: Just five years ago?

Shay Johnson: Yes, 5 to 6 years ago. So, it's like we'd have a lot of online activity, but I feel like the majority of our hiring was really happening through non-digital sources. And a big part of that at the time were things like we didn't have a direct integration or any facilitation between major job boards like Indeed and our ATS. So, there was a lot of just fragmented kind of, what's the posting posture that we were doing or our online presence, and what's the actual hiring activity. I'd say that's the biggest change is just how much we have actually connected, solidified, streamlined, automated, and done all of that from a technology perspective, whereas we had some tech 5 years ago, but I feel like it was all just visibility and surface, and the real hiring was like grassroots, local, nose to the ground, hiring effort. And now a lot of it truly is done online and heavily automated.

Chad Sowash: But that's distribution, so you actually plugged into distribution lines. And after that happened, you started to see, I would assume, some bumps in applications and what not. So, take us further into the timeline.

Shay Johnson: Yeah, yeah. Shortly after that, so I think one of the biggest things that we did, and I'll just call out a vendor, a partner in this one, was our partnership with JobSync. So about 3 years ago, again, no direct integration with Indeed, so we had all of our jobs feeding there, and just so you guys know, we have about 15 to 16,000 jobs live at any given time. So, that's what's running to Indeed every day, but with the experience we used to have, a candidate would land on a site like Indeed, they would click the apply button, that would redirect them to our career site, and that's about where 80% of the drop-off would happen right there. So, when we introduced a partnership with JobSync, it's just like it sounds, we're trying to synchronize our job sites and our systems, we made it so that we provided that native Indeed Apply experience for every one of our jobs so the candidate doesn't ever have to leave Indeed, they just get to complete their application, but it's actually our application. They're filling out, you know, our disclaimers, our requirements, our screening questions. The candidate actually makes it into our ATS automatically. So not only are they going to get into the hiring manager's hands faster and process through to hire, but now they're in our ATS and our CRM. So all of our retargeting, our re-engagement, our actual hiring event invitations, everything, we get to capture that entire group of people that was flowing through the behemoth that is Indeed, whereas before we were losing all of that.

Chad Sowash: And you now get to use them and retarget them because you spent money to acquire them.

Shay Johnson: Yes.

Chad Sowash: It's beautiful.

Joel Cheesman: Genius. It's beautiful. Genius.

Joel Cheesman: Were people dropping off after coming from Indeed because the experience was so bad compared to Indeed? Was it because it was just a different system? Like what were your thoughts on why the drop-off was so high, and how has it improved since making the change?

Shay Johnson: Yeah, to me, I mean, I would think it's not necessarily that it's a bad experience, but if you've already gone through the trouble of creating a profile on Indeed and doing all the things they recommend, bolster your profile, add your resume, do all these things, you don't want to click apply only to land on a company's career site and have to start from scratch, right? So, to me, the whole point of being on a job network like Indeed or LinkedIn is the quick apply experience. You already have all this information, just hand it to the employer. So, it wasn't necessarily that it's a bad experience. You know, it could be a very quick apply experience on our career site, you just don't want to have to do it. And so, we just wanted to facilitate that automatic connection.

Joel Cheesman: And what has that meant to improving the numbers of applies, how much in savings? How has that translated into...?

Shay Johnson: Yeah, overnight when we turned that on, it was about a 150% increase in applicant volume. So that was real people. That wasn't bots, that wasn't AI at the time, that was just simply this is how many people are trying to apply to our jobs on Indeed organically, not even just through sponsored jobs. Like, all these people want to apply, but we're losing them because they're being redirected to another place. So just by saying, you don't have to leave Indeed, you don't have to leave LinkedIn, ZipRecruiter, I mean, we connected 6 or 7 major job boards, but you just actually get to complete your application right here as you clicked on it, and now, the moment you apply, you're going to get a follow-up from our actual system saying like, 'Hey Chad, thanks so much, we saw you applied on Indeed, now we're going to actually usher you through our custom process that we've, you know, spent again, tons of time and money and energy building, you know, we want to be able to leverage it with everyone.'

Chad Sowash: Talk about that process now, because most companies, they buy candidates over and over, and they never use that applicant tracking system or their CRM to re-engage. How did you guys, when did you start doing that number one, and number two, how are you doing it today? Because obviously, it's working.

Shay Johnson: Yeah. I mean, we started it about the same time, I mean, we were doing some...

Chad Sowash: Right around JobSync time?

Shay Johnson: Yeah, around that time. Yeah, I mean we were already trying to do retargeting and re-engagement, but with very, very little results because your applicant pool isn't large, a lot of them aren't engaged. So, once we had all that volume coming in from JobSync, and you start using automated re-engagement through a couple different partners where, you know, you're getting different signal, right? You're saying, 'Hey, this candidate has applied to multiple jobs, maybe they're sitting in the pipeline for a job that is getting ready to close because it already has candidates in the interview or in the hire stage.' So, we know we want to automatically re-engage them and say like, 'Hey Chad, I know you just applied to this job, but actually, you know, this other one is getting ready to open right now and you'd be an early applicant.' So, you're getting buy-in from a candidate that you're getting re-engaged immediately and led to the right job that they actually are probably going to get hired for. We started doing that around that time. We started just being able to leverage this talent pool to invite to our hiring events, which is where a lot of our activity still does happen today. So, all of this became connected with just now that you have access to all these people, you can use these tools to re-engage them, and you can re-engage them really quickly. You don't want to just let them become stale for 6 months to a year, and then say, 'Ah, I need some people, maybe these people that applied a year ago would be interested.' I mean, sometimes, but rarely.

Joel Cheesman: And is the re-engagement automatic? Is someone doing that? And the engagement, I assume, is different with like a talent community versus sort of a passive, 'Just keep me sort of in the loop on what's going on.' How do you guys approach that challenge?

Shay Johnson: Yeah, all of the above. So, you've got kind of the last one you mentioned, job alerts, right? Like you kind of sign up, you say what you're interested in or you apply to a certain job, just as jobs get posted, maybe on a weekly cadence, you're going to get job notifications from us. That's just kind of general automation, nothing super specific or super targeted. You've got manual, which is where our recruiters are actually going in, and whether it's for jobs that they want to push applicants to, or a hiring event, they're going to go through and create kind of their filters and their drip campaigns and initiate those. But then our targeted automatic re-engagement that I was kind of talking about before, which is using all those different signals to say, 'How do you target someone as quickly as possible after their initial engagement when they're most likely to be convertible?'

Chad Sowash: What are you using to target them, though? Because you obviously have to have some type of logic that's in there that says, 'This individual is not only applicable for this or meets the requirements for this job, but this one also over here.'

Shay Johnson: Yeah, so today we have another partnership with a company called Dalia, and that's who's been really critical in that, and that we've done a lot of this testing with, where we've said, 'Hey, you all already have the mechanisms to do this re-engagement,' and JobSync comes back into play because they have all of that signal to say, 'We've got all these candidates, we know from which job board they came from, which job pipelines they're sitting in.' They can look at all that different candidate volume, feed it to Dalia, and so Dalia can use it to determine how they want to target which candidates, when and where. No recruiter ever has to dictate that. It's all happening automatically behind the scenes.

Joel Cheesman: Curious about your hiring events, which seem a bit passé post-COVID. But you do them, I want to know specifically what the strategy is, but you use technology in a very unique way from most TA professionals. How do you marry your face-to-face events and technology, and getting people in the system and then targeting them through marketing, etc.?

Shay Johnson: Yeah. Well, this is actually one of the things, the very first things that came during the pandemic, was—so hiring events have always been a big part of what we do, especially for sports and entertainment, college hiring, like when we're doing the food and beverage on college campuses. A lot of your hiring, I'd say like 60 to 70% of your initial workforce that you need to hire for the start of like, let's say a basketball season, is going to happen through their hiring events. And it's not just ours, a lot of times they're hosted with a partner or client, you're hiring for the whole arena or the whole ballpark, so all those different roles that are not just ours, maybe you got parking and ticketing and things like that. So, hiring events have always been important in terms of just that volume, but where we started using technology a few years ago was we needed virtual events during the pandemic because we weren't doing in-person events. And then after we started using a virtual event platform, which in our case was Paradox, they actually came to us at that time and said, 'Hey, we've just spun up an events platform, do you want to try it?' And we were like, 'This is a game changer,' because what it does is it brought that conversational experience that we had been using for job search and apply to events. So, whenever we're doing these outbound marketing campaigns that I'm telling you about, instead of encouraging someone to apply to a job, you could do that. They could be paired together, but essentially you're saying, 'Hey, we have a hiring event next Thursday for the ballpark, click here to learn more,' and they're greeted by a landing page with Olivia, and it's all about getting them registered for the event, pre-sorted into certain interview lobbies, which hiring managers they're going to connect with, they could choose their times, they can get sorted, so it all just depends on the nature of the event. So, you're already giving this kind of conversational intro into an event, that also allows you to do things like, 'Hey, let's follow up with this person the day before the event and tell them about parking and where they need to go and who they're going to be speaking with,' and when they get there, they can scan a QR code and check in and go right where they need to go. So all of that was done manually before, you know, it's just kind of a free-for-all with hundreds of people showing up to an event. Now you can actually use this technology the same way you would if you're going to a concert or anything else like that where you've actually got engagement and direction for the event itself.

Joel Cheesman: Dammit Chad, I thought he was going to say QR code somewhere in there, but he didn't.

Shay Johnson: That's just one option. Text codes...

Chad Sowash: Instead, we're going to talk about ATS, but there's conversational ATSs now. Talk a little bit about that, and how that has possibly changed the game for you guys, or if you're looking for it to change the game.

Shay Johnson: Yeah. Yeah, so we, again, I've mentioned Paradox, and we started our partnership with them back in 2018, and it was more about putting the conversational overlay on our career site, but it sat on top of our ATS. And now we're actually implementing Paradox, as we speak, as our full end-to-end conversational ATS. So, yeah, we've actually been on this 7 year plus journey with Paradox, and now it's finally gotten to the point where their offering is mature enough for our business's complexity, which it is, there's no way to cover it in the conversation we have today, but the level of complexity, customization, configuration, and just the requirements we have, and the integration depth that we have, I think Paradox has done a really good job of getting to that point where not only do they have a great conversational product, but a great ATS product that can serve all of that. So we can say, 'Hey, it's not just this overlay where it's kind of like chat to apply or search for a job or schedule your interview, we want to have that same experience all the way through offer, onboarding, and beyond,' and now we get that end-to-end with Paradox. And it also just makes that experience for hiring managers and recruiters so much better because Olivia's their assistant, too. So, anything that they need to do within the ATS, it's conversational for them as well. Some things they're going to do via just native software and UI, but more and more things every day, Olivia can actually help guide them to what they need to be doing.

Joel Cheesman: How about from a job seeker's perspective? Because if I'm a job seeker, I go to site after site after site, it's looking at a job, clicking apply, filling out the application. I go to your site, it's a conversation. Was there pushback? Was there like, 'The sun has come out'? Talk about that. Change management.

Shay Johnson: Yeah. 5 years ago, there were so many people that were like, 'Nobody will want to do this. Nobody will want to chat,' and, you know, 'chatbot' was such a dirty word 5 years ago, because we were all so used to the airline chatbots and the DMV, things that make you infuriated because...

Chad Sowash: Comcast.

Shay Johnson: Yeah, and I would say this admittedly, when we first went live with that, there was some of that, right? You kind of get stuck in that infinite loop or, 'I don't understand what your question is,' and things have just obviously progressed so much in the last 5 years that now everybody—but then the challenge is everybody is so used to frontier models and leading labs and that experience and how great it can be, that now the expectation is you'd better be serving that up on your career site. So, I think that is the challenge for a lot of the TA tech companies or the HR tech companies, are your products and your models and your conversational products as good as the ones that people are going to be using either for free or paying for at home. And so it's gone from resistance and nobody will use this, to yeah, we have to be using this, but it needs to be like, everybody wants to buy it and making it better.

Chad Sowash: Better be good.

Shay Johnson: Yeah.

Joel Cheesman: Because I think there was a movement of like, 'Yeah, we're going to try this chat thing, but it's going to be in the corner, we're still going to have to search as you know it.' And you guys took the leap into that, and I think you should be...

Shay Johnson: Yeah, and I think we like to give people choice, right? Like you can still have a traditional kind of search experience, but you want to drive as much to the conversation as possible, and I think we've learned a lot of things over the past few years where not everything needs to be the back-and-forth conversational elements, especially if you're on mobile and you get to the point where you probably are filling out more of a form, it's way easier to use the saved information on your phone to like quick fill a form, so maybe in the conversation you're being presented with that form, so it becomes quick fill instead of having to say, 'What's your name? What's your phone number? What's your address?' and typing it all out. I think there's a lot of those things that 5 years ago that's how it was, it was very kind of deterministic, 1-to-1 conversation, and now all of the products have gotten a lot better of saying like, 'You can actually build UI into conversation, so let's start doing that where it makes sense.'

Chad Sowash: Gotcha. So, no, we're not going to talk about QR codes. What about agents? This new cool thing, because we keep hearing about orchestration, and obviously, Workday buying Paradox and being able to see kind of like that technology start to move away from talent acquisition into talent management as well, because you've got that conversation that's happening as you've said, you can have it with employees as well, but there's the agentic piece. Can you explain that a little bit?

Shay Johnson: Yeah. The way that I think about it is, we've gotten away from conversational as an experience layer and bots to handle things that are kind of your very repetitive, tactical interactions, and where agents really come in for us, as we think, it should do what you'd expect the best possible version of a worker to do for certain processes. So in recruiting, if you think, 'Okay, I'm going to be looking at candidates' profiles and resumes, I need to reach out to them and ask them some questions, based on those answers, maybe I need to update some information in their profile and make decisions about what's next.' If you can automate each of those things to a degree, and at scale, you have to standardize it. So if you're thinking about it, like I said, 7 million applications...

Chad Sowash: Yeah.

Shay Johnson: ...you're going to have to design something with automation that's very uniform and it's not custom to people. If you're a recruiter, then you can take all of the context of, 'Here's the job description, here's the hiring manager, here's what's unique about this location, here's what's in this candidate's profile, here's the pre-screening questions that they answered,' and so it can go something like this. An agent, instead of just a bot, can actually call you, text you, email you, engage in whatever way is going to be your preference, and they can say, 'Hey, not 'Where are you located?' like a bot would, they would say, 'I see you're in Santa Clara, California. This job is in downtown San Francisco, that's about an hour commute. Is that okay with you?' You might say, 'Oh, I actually just recently moved to downtown San Francisco.' You'd say, 'Would you like me to update your address in your profile?' Like these are the things that a good recruiter, if they actually had the time to do it, could do. And imagine if you had the time to have that care and that detail and that thought all the way through the process, 24/7 with 7 million different people. That's where when you start to build agents, you say, 'What's the best possible version of a workflow look like, and what are the different points that could be brought into it? Let's build an agent that's really, really good at a screening conversation. Let's build an agent or, in light of a recruiter, that's really, really good at that second level deeper interview once you have more context. Let's build an agent that is really, really good at nuanced scheduling for a certain team that does this really crazy interview process.' So, that's just the thought process behind it, but we can get more technical if you want.

Chad Sowash: Yeah, yeah, yeah. I would love to, we don't have enough time. You said interview. Are you guys thinking about the prospect of agent interviewing? Because...

Shay Johnson: Yeah.

Chad Sowash: ...for me, the types of jobs that you guys have, if they're qualified, instead of scheduling the interview, you just go right to the interview. Skip days prospectively.

Shay Johnson: Yeah. Yeah, we're working on that right now and it can look a little bit different depending on the role, right? Frontline versus professional. But we're truly, yeah, we set out about a year or two ago to say, 'Hey, how can we use a lot of the new AI capabilities to have better screening,' is how we thought of it. Is saying, 'Hey, you could still have candidates answer these basic screening questions: how old are you, or are you available for these shifts, or do you have this experience?' But what we were finding is, let me say for frontline hiring managers, they're still going to have to pick up the phone, call the candidate, and ask them some other questions that are specific to that role or their location that we can't just standardize in screening format. And so we were like, 'Let's have an AI agent that could go do that first interview, that 15-minute phone call that a manager or recruiter would have to do.' And once we started looking at the products and the capabilities, now it's like, oh wow, it could really just do the interview significantly better than, especially a hiring manager that's not a trained recruiter, could do...

Chad Sowash: On demand.

Shay Johnson: ...Yeah, and for everybody. I mean, we can make it so that, yeah, if you're a candidate and you're pre-qualified with the pre-screening questions, you can actually have a very nuanced interview that is asking you, revalidating and verifying some of the basics, but also is getting more into what eventually becomes more like an assessment, actually like probing the candidate and asking about certain skills or experience that they might have, so you're building a more robust profile about them. And this is not just for the purpose of that one interview, you can apply that to any job now that they might be interested in. We actually...

Chad Sowash: That's the feature, right? You're gathering the data.

Shay Johnson: Yeah, so this is the big thing is what we've noticed in the pilots that we've done so far with the AI interviewing is that a common thing that happens for us is candidates will apply to multiple jobs at once because we've made it so easy. So, the AI agent, instead of saying, 'I'm just going to reach out to you and interview you for that cook job,' they'll say, 'Hey, I noticed you applied for multiple jobs today. Let's just have a general interview that would cover any...' like if again, if I were a real recruiter and I reached out to you, I'm not going to interview you three times, I'm just going to have one interview, apply it across all those different roles. And then all of the data that's gleaned from that interview now can be used to enrich your profile. If we need to re-engage you 3 months from now, we have all that information that we wouldn't have had if a person got on the phone and talked to you, and that knowledge only stayed with them and never got passed on to anybody else.

Chad Sowash: Yeah, yeah, because you know they're not putting it in the system.

Shay Johnson: No.

Joel Cheesman: A common complaint that Chad and I hear is the flood of candidates that are coming in, the lazy applies, turning agents on the other side, agents for the job seekers are coming into your funnel. Do these tools you're talking about solve that problem? How do you guys handle the flow of fake candidates, of bots? It's a huge issue right now and I think a lot of people would love your perspective on how you tackle that.

Shay Johnson: Yeah. I mean, we're putting in place bot detection and fraud detection. So what's good about that is both the systems and the agents are getting better about that where when they're actually reaching out and contacting a candidate and getting their responses back, whether it's via text or via phone, they're getting very good at being like, 'What percentage of likelihood is this a bot or is this potentially fraudulent?' So, we have that signal now, so that is helpful.

Joel Cheesman: Is that built into the software that you're using?

Shay Johnson: Yes. Yeah, so then what you're getting is, you don't have to necessarily make it a knockout, you could just be like, 'This is good signal to have that, if maybe we won't automatically schedule a candidate with a person if there's over a 10% likelihood that they're a bot or fraudulent, like maybe they need to stay in a more of a manual review state for someone to take a look at them or call them themselves.' But I think that's the biggest thing is we're doing a lot with top-of-funnel volume optimization where we don't want 7 million applicants, we want to get back down to the point where we're truly making it so that sometimes we have applicants before a job is never posted, and there's some things to talk about that from a legal perspective, but when it comes to internal mobility, definitely, we do a lot of this. So, we're just saying, 'Hey, if we already have this huge candidate database and from previous interviews we have this knowledge about them, I could just say, 'Hey, this job just opened up, I could reach out and text you and say, 'Hey, I know we just talked last week, I just had a job open up, are you interested in this one?' I could do that with 50 different candidates that were engaged last week and have a pipeline of 10 candidates, 3 scheduled for interviews without ever opening the job.' So, that's just another way that we can go with this, but in that, yes, we want to have fraud detection and bot detection. But the point is, we want to be smarter about engaging our candidates and re-engaging them. So, the more signal that we have that someone is authentic, that's who we're prioritizing engaging with versus prioritizing just speed and we're letting a bunch of traffic through that is fraudulent or bots.

Joel Cheesman: When you say we don't want 7,000 people applying or those kinds of numbers, why not? If you have a system to detect who's real and who isn't, why would you not want as many people in the funnel as possible?

Shay Johnson: Let me put it that way. Yeah, if let's say 10 to 20% of that though, is fraudulent or bots, you don't want 7 million of those applicants to—you don't want hiring manager eyeballs on them, you don't want recruiters wasting time trying to sift through them, they would never even see those applicants. The only ones that are actually going to be under review or land on a calendar for an interview are the ones that are legitimate, they are verified, they meet the minimum requirements, they've had an interview, they've gone in-depth, we know there's some good skill or experience matching here, pretty much everything that you need to know to say like, 'I just need to have a conversation with this person or bring them on site and make them an offer.'

Joel Cheesman: Got it.

Chad Sowash: Yeah. You're one of the guys out there that's always looking on the horizon, and there are a lot of people that aren't because it's way too hard with all the work that you have to do day-by-day.

Joel Cheesman: Shay's a visionary.

Chad Sowash: Okay, he's a visionary.

Shay Johnson: I'm crazy!

Chad Sowash: He's in our face. When you're looking out, what is the newest, coolest thing? Agents, we just talked about that, that is cool, orchestration is amazing. What's next? What's that next thing that...

Joel Cheesman: ...bring your crystal ball...

Chad Sowash: ...that you want to get into? Doesn't have to be something that you come up with, but the next thing that Compass Group wants to get into past agents.

Shay Johnson: So, agents begets this, but I love how agents build upon themselves, and so it's like once you have an interviewing agent and you have access to this data, teams... Oh, yeah. Well, what I just love is you can start to do things that you've never done before and not just at scale, but just in general. So, I love the idea of AI helping to find the solve for transparency in the candidate experience.

Chad Sowash: Oh, yes!

Shay Johnson: And so what I mean by that, I kind of talked about like if you have that signal, like if it's very obvious from the moment that that person applied to that job that they probably aren't going to get hired...

Chad Sowash: Why don't you tell them?

Shay Johnson: ...engage them immediately, tell them that, have a conversation with them, say, 'Hey, we've got a couple of other roles though, let's have a conversation.' As soon as they have that conversation, we've done skill mapping, it's understanding how well they might match, and it could be honest with you. It could say, 'You know what, Joel? Based on our conversation, I don't know if this role would be the right fit for you, but can I give you some feedback and some coaching on things that you might do to develop to be ready for that role?' You can start to deploy coaching directly to candidates and this can be used for internal, too. You can use all of that information so that in succession planning, the moment you know that there's going to be a promotion happening on your team, you can basically say like, 'Go into our candidate database, go into our internal employee database, start to look at who might be the right fits for this role, start to look at what the skill gaps are, and go ahead and engage them and say like, 'Hey, just wanted you to know there's some role opportunities coming open, let's talk about, from what we know from a data perspective, these might be your perceived skills or your assumed skills, let's have a conversation and get a better sense of what you're really interested in, what kind of work you've done over the past year and start to build up that profile that then is immediately going to feed into, okay, maybe these are your internal candidates that you want to prioritize.' So, it's a lot of the talent things that we would call talent intelligence, but before now, you've been so reliant on it's only as good as the data people put in, and it's a lot of candidate like self-selection and forms, and how do you create a really frictionless, easy, intuitive recruiting process? You try to make it short for people, but then you're missing out on data that you desperately want to inform talent intelligence, introduce agents that come in and say you can bring the power of natural conversation and speed at scale and it just enables so, so much. It's just unlocking a ton of possibility, and right now we're trying to decide like, where do we want to focus? Because I think there's probably 20 use cases on the table right now and we're like, 'What are probably the top three that we want to focus on in the next six to, you know, 12 months?'

Chad Sowash: Yeah, you got to have your own roadmap.

Shay Johnson: Yeah.

Joel Cheesman: The last sporting event I went to, I grabbed a beer, couple hot dogs, and I went to a little kiosk, put it on the table, and it told me what I bought and how much it cost, and I credit card and I was or tapped and I was gone. As someone who's on the front lines of people and providing people to facilities, but also facilities wanting to hire less people and have more solutions like that, you're on the front lines of this. Are we headed for an all-automated game experience? Is there a balance between, yes, there's some people, but fewer? Like you're on the front lines of this. What are your thoughts?

Shay Johnson: Yeah. Well, we always have people involved in key decisions, so that's going to be, you know, who are the final people getting interviewed, who's getting an offer, who's hired. So, there's always those human elements to it, and there's a lot more than that, but for us, it's we're automating things that today we simply cannot do at scale. We have to be so selective, and so many people don't get looked at, they don't get the conversation. So, you're basically just like, 'I can do this with people, but I can only do it for 1% of the population.'

Chad Sowash: And you don't scale well.

Shay Johnson: So, yeah. So, when you automate at scale, then what the people can do is things that we also can't do today. So, for example, we have a team of 20 recruiters that support over 150,000 hires. And so when you do that by automating all of what we're talking about, so 5 years ago, those 20 recruiters, all they did all day long was post jobs and answer emails from hiring managers that wanted to close their job or add openings to their job.

Chad Sowash: What a horrible existence! Horrible existence!

Shay Johnson: That's all they did! And so now, when you automate 90% of that, what those recruiters get to do is we say, 'Hey, we're opening up a new Delta One lounge at JFK, it's going to be a really exciting opening, we want to send five of you there on site, we're going to build the recruiting strategy, we're going to do some canvassing, we're going to do some media, you guys are going to meet everybody, you're going to welcome them, you're going to help them with like their benefits transition, you're going to make sure all the tech is running smoothly.' So, they're facilitating the tech experience and the people experience, whereas 5 years ago, they just sat there and clicked buttons and referenced Excel sheets all day.

Chad Sowash: Isn't that a better existence? I mean, seriously. I mean, because we talk about how, again, AI is taking jobs, and literally it's taking tasks, and they're really the really crappy routine tasks and they're administration heavy, where what you just described, again, that to me, is the discussion we should be having. What are we going to do next, right? AI's going to take the—what's the job going to look like?

Joel Cheesman: Yeah, is Compass Group hiring recruiters this year? Are you adding to the team?

Shay Johnson: Yeah! Because, yeah, especially when it comes to the salaried professional roles, like a lot again, we used to outsource a lot of our exec recruiting, we used to use a lot of agencies, our dependency on that goes down because now our recruiters actually are able to level up and use those skills to deploy that.

Chad Sowash: Cost for that is high, so you're actually going to be saving money and then getting more, yeah.

Joel Cheesman: Yeah, your chats with the executives must be really positive with all the money you're saving and all the efficiencies.

Shay Johnson: Yeah, and everything you save just gets to be reinvested into something that is more purposeful and it's all for scale, right? It's, 'Hey, we used to pay all this money for something that, like I said, only had a 1 to 5% impact, we can reinvest that into something that's going to give us 20 times the scale and the impact.'

Joel Cheesman: Social media. You hire a lot of young people, that's where they are. I know you've got the job posting thing down pat and everything else, how are you guys leveraging social media, if at all?

Shay Johnson: Yeah, so I would say we have taken a very non-corporate or enterprise approach to this, and for one reason, and it's because we do not have a lot of consumer brand awareness or presence. We're not a Taco Bell or a Wendy's where people are like, 'Oh, I see this all the time and it resonates with me.' So for social media, we rely a lot on what you would consider kind of your crowdsourcing, like our employees. We have a lot of associates that have a social media presence, some of them are musicians, some of them are influencers, some of them do magic, or have travel blogs, like whatever it may be. So, we've actually found like, and we do this in pretty small scale, we're not pushing this from a corporate or an enterprise perspective. In each of our sectors or brands, we have marketing and social media teams, and they work more intimately with those specific people to say, 'Hey, if you're okay with it, would you mind using your channel to promote some of this content? You can or you don't have to.' So, we've found a lot more success with going a little bit more organic in that way than having this big robust corporate social media strategy where we're just spending tons on ads or something else. We stopped doing that years ago. We don't spend a dime or touch social media like from a TA perspective.

Chad Sowash: That's more organic, the way that you're doing it. It's more organic as opposed to, you know, trying to do something that's more polished and corporate, and they will sniff that out. They'll sniff that. Are you guys using any like video platforms to be able to manage that so that you can send out links to your employees and say, 'Hey, if you don't mind...'?

Shay Johnson: Yeah. We again, we don't from the corporate TA perspective we don't, we've put it in the hands of the direct marketing teams at the sector level because they know their brand, they know their people, their culture. We never want to push it just as Compass Group, it needs to come through as Levy or Eurest or Chartwells Higher Ed. So again, that's where having a good partnership with like our comms and our marketing teams are good, because that's part of their role and, you know, what they're asked to do is support with recruiting and staffing. So what we do is try to enable the technology and what's the downline process, like, 'Hey, once you actually get someone that's interested, let's make sure that experience from search or apply to hire is as frictionless as possible.'

Joel Cheesman: Chad, I didn't even know Taco Bell had a social media presence, so that's news to me.

Shay Johnson: Sure you didn't. Your algorithm isn't as refined as ours.

Joel Cheesman: No clue, no clue about any of that. Guys, that's been Shay Johnson at Compass Group. I'm Joel Cheesman, he's Chad Sowash, and we are the Chad & Cheese podcast. This is the AI Series. We out.

Chad Sowash: We out.

Outro Voiceover: Wow, look at you. You made it through an entire episode of the Chad & Cheese podcast. Or maybe you cheated and fast forwarded to the end. Either way, there's no doubt you wish you had that time back. Valuable time you could have used to buy a nutritious meal at Taco Bell, enjoy a pour of your favorite whiskey, or just watch big booty Latinas and bug fights on TikTok. No, you hung out with these two chuckleheads instead. Now, go take a shower and wash off all the guilt. But save some soap, because you'll be back. Like an awful trainwreck, you can't look away. And like Chad's favorite western, you can't quit them either. We out.