Programmatic Fire Sale


This episode features Jason Roberts a guy who knows recruitment, process, and technology. This podcast is jam-packed with topics like:

- It's a "Programmatic Fire Sale" - I Frankensteined a tech stack - Monster matching works? - Hatin' on Indeed's Be Seen - Everyone wants a chatbot! and much much more...

Big thanks to Sovren for making this podcasting exclusive magic possible. Enjoy!

PODCAST TRANSCRIPTION sponsored by:

Disability Solutions is changing minds and changing lives through disability inclusion.

Sovren: Sovren is known for providing the world's best, and most accurate parsing products. And now, based on that technology comes Sovren's artificial intelligence matching, and scoring software. In fractions of a second, receive match results that provide candidate scored by fit to job, and just as importantly, the job's fit to the candidate. Make faster and better placements. Find out more about our suite of products today by visiting sovren.com. That's S-O-V-R-E-N.com. We provide technology that thinks, communicates, and collaborates like a human. Sovren, software so human you'll want to take it to dinner.

Announcer: 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, brash opinion, and loads of snark. Buckle up, boys and girls. It's time for the Chad & Cheese podcast.

Chad: All right, guess where we're at?

Joel: Oh yeah.

Chad: Where are we at? Austin, dude. This is fucking awesome.

Joel: This is the quintessential, like Texas-

Chad: Yeah.

Joel: Front porch-

Chad: Out on a balcony-

Joel: Really nice-

Chad: TA tech-

Joel: Weather-

Chad: Oh, shit.

Joel: Whatever kind of tree this is.

Chad: Yeah.

Joel: That's shading us. This is nice.

Chad: That's a Bur Oak.

Joel: Thank you.

Jason: That's a Bur Oak. Yes.

Chad: A Bur Oak.

Joel: Yes. Now I just need some sweet tea, and some key lime pie, I think.

Jason: Whoa. No. Pecan pie, thank you.

Chad: Pecan.

Joel: My bad. I had a Florida moment there.

Jason: You did.

Chad: That voice you hear is the voice of Jason Roberts. And, let me set this up real quick. So when we get an opportunity to actually get some of the cream of the crop on the show-

Joel: Cream of the crop. I don't think he's ever given anyone that label before.

Jason: I'll take it, 100 percent.

Chad: Adam and Quincy, and I mean those guys, Jason's of that same level.

Joel: He's on a one name basis. He's like Madonna and Prince of the industry.

Chad: In RPO for how many years?

Jason: Oh, gosh. Over a decade.

Chad: Over a decade in RPS. So RPO again, for the listeners out there, recruitment process outsourcing. This is the business of recruiting. These guys focus on margins. They focus on efficiencies. They focus on technology that in most cases, talent acquisition, they just don't have time for it, because they're dealing with 401ks and-

Joel: They're focused on the biz nasty, is what you're saying.

Jason: It's true.

Chad: So that being said, Jason, what do our listeners need to know about you other than all this wonderful setup?

Jason: I think that was pretty good, man. You made me sound pretty good.

Chad: What's your last name for those that aren't on a one name basis with you?

Jason: That's probably good. I'm just one name. That's all I need.

Chad: Cher, Madonna.

Joel: Jason and the argonauts.

Jason: Aren't there dudes?

Chad: Prince.

Jason: Prince. Yeah. I was going to say, there's got to be a guy.

Joel: Just, Jason.

Jason: So, I'm Jason Roberts, and that's right, I've worked in RPO for a long time. I don't work in RPO, right now. So, that's a new thing for me.

Chad: Okay. It gives you a time to breathe at this point though, right?

Jason: It does. You know what I've realized, I had to sort of do a little soul searching, and decide what my next chapter was going to be.

Joel: You should start a podcast.

Jason: I've got one. We just don't record very often. So, yeah. In fact, we've added that to our tagline. We've said, it's like the best in bots, and whatever he says. Then I just tack on, every once in a while, because we are not consistent at all. We just do it for fun. We don't do the business of podcasting like you boys.

Joel: The biz nasty.

Chad: The business.

Jason: The business.

Chad: So, give us some background when it comes to tech, because again, we talk about tech all the time, but behind the scenes we talk about stacks, and all that shit that's not used, and so on and so forth. Your whole focus was efficiencies, and making sure that you got everything out of the tech that you put together in RPO.

Jason: Yeah. So there's a problem in technology in that there's a litany of different techs out there. There's modularization is what I call it, right? So if you want the best in breed, you have to have 10, 15 different pieces of tech. And if those are going to be useful for anyone, they have to be sort of woven together through an integration, a sort of string of integrations, really, to create a stack that makes sense. And we're in an age right now ... I've been thinking a lot about this. We're in this sort of age of acceleration, right now. So, we were talking about the second machine age. I talked about that for about three years, now, and now I'm pretty convinced that we're at this point where technology is accelerating so fast that people can't keep up with it.

Chad: Oh, yeah. Moore's law, right?

Jason: Thank you. That's exactly what it is. But see, people don't think about Moore's law and the fact that there's a downstream effect of that, right? The processing power is ... I used to think about Moore's law meaning, oh, processing power doubles every two years. The last doubling was 18 months.

Chad: Yeah. Yeah.

Jason: So, I used to think about that, and I thought, oh, video games are going to be way better two years from now.

Chad: Exactly.

Jason: That was the whole purpose for me.

Chad: Latency. That was awesome.

Jason: That was my whole thought. And now if you think about it, the reason we're able to order a car on our phone is because the size of that processor has gotten so small that it just fits in your hand, right? So, technology has created new thing that we are having to adapt to, and that ... Uber is a good example of that, right?

Jason: So if you have ride share that you can call up, government is trying to adjust for that today, right? They're trying to figure out how do we legislate around this? What are the rules we need to have on these app based ride share deals that are out there? And, here's the problem. Maybe three or four years from now, they're going to be done figuring out what those rules are, and three or four years, Moore's law has already kicked in a couple of times. Guess what's here? Self-driving cars are good to go. Whatever they just legislated is now completely obsolete, because we don't have drivers anymore. And, I think you guys did a story on the California law that they ... the resolution they passed for gig workers, specifically hitting Uber, right?

Chad: Yeah.

Joel: Will be obsolete, at some point.

Jason: Yes. There's no point. By the time laws get enacted across the board, those people will just be replaced by self-driving cars. And, what do you do then? There's a whole new set of laws.

Jason: So, technology's moving so fast that it's hard for our government systems, and in our world, in HR, it's hard for HR leaders to adapt quickly enough, and to understand what to do with it.

Joel: I think the litmus test will be, is it killing people? So e-cigarettes, killing people, legislation follows. So if it's not killing anybody, let's just let it ride.

Jason: Have you been looking at the watch, and the e-cigarette memes? Like, the-

Joel: No, but I can only imagine.

Jason: Oh, gosh.

Joel: I don't mean to make light of people dying. Like, that's crazy shit. But the point is like-

Chad: No, but that moves the needle.

Joel: Governments at some point, what moves them to action that's relevant to our lives, people dying is one of them.

Jason: That's one of those things. And, they can move faster when that happens.

Joel: When public opinion says, holy shit.

Jason: And with Uber, public opinion is, we freaking love this thing.

Joel: Unless, it's guns. That's a whole other podcast.

Chad: Yeah. And, people don't give a shit about that at all.

Jason: That's the meme that's out there.

Joel: Oh, really?

Jason: They'll put out, three people died from e-cigarettes. We're going to ban them. And then, the gun control goes right next to it. And, that's a hard discussion to have.

Joel: What of the founders would have thought about e-cigarettes wrote the constitution.

Jason: Moore's law, man. It was before their time.

Joel: So, what else in technology has you excited? I know that you just got back from SourceCon.

Jason: I did.

Joel: You talked about fully automated recruiting, which I think we've been talking about is the panacea of where this thing is going.

Joel: So, talk about that presentation. What else has you geeked on tech?

Jason: Yeah, I want to tell you about that. So, we built something at Randstad, right before I left, and it was ... I've got to say, I was pretty happy with the way that it came out. So, the idea is that we built a fully automated sourcer. So, we had a fully automated recruiter on the agency side, even if you're looking at that side. So the idea is, we have jobs come in. We match candidates in all of our databases, and there's a bunch of candidates in the databases, including Monster's, at the time, right?

Chad: Yeah.

Jason: So, we match all of those candidates. We send send matched candidates over to a bot. And, we used bot that ... I think it is sort of flying under the radar, but they've got a killer feature that makes the fully automated thing work.

Jason: And, the bot was Wade and Wendy that we used. The killer feature is, when they got the new job in, that they'd never seen before, they have a knowledge graph, and they build their own chat. But what people don't know about the bots is that, with most of them they work in high volume sort of roles where you're hiring a lot of the same thing, and the reason they've gone to those roles first is that, you have to spend some time building that chat every time. So for each role, there's effort involved in building the chat. When the machine can write its own chat, it doesn't matter what sort of job you get. So for white collar roles, like if you have a big multinational conglomerate with 80 percent uniqueness, month over month, in their jobs, chatbots aren't a great solution, the traditional ones, I say traditional, they've been around for three years.

Jason: Eli is a three year old company with what? 80 million in funding, at 300 million dollar evaluation.

Chad: Yeah.

Jason: That's crazy.

Chad: That shit ton of cash.

Jason: That's crazy. So I mean, and they're great at what they do. They automate their piece of the process really well. But, when Wendy is one that is really good at that sort of white collar worker, where you have lots of unique jobs all the time. So we pumped it through this thing, and we were able to submit candidates to jobs pretty quickly. The stats that came out of on the back end is that it was about a quarter of the cost of having a human sourcer do the work in order to have the machine do the work.

Chad: So a quarter of the cost, and then also, are you taking out the actual head count cost of the person, or did you even take that out?

Jason: Not yet. Not yet. So we were just doing ... We did a comparison.

Chad: So, it was even less expensive?

Jason: What we did is we measured what a sourcer's target metrics are. Was. So, a good sourcer handles this many wrecks. They do this many submits per week. So, we had the sourcer targets. So not all sources are hitting those, but enough of them are, and then we measured that against what the actuals were that the machine produced, and the machine produced about four times as much. It's blazing fast. And, we started it out, playing with this on the MSP side, just to see what we could figure out. And, it was great. It did a good job.

Joel: So, let's break this down, real quick. So, you had an OpenRec, basically matched a database of resumes that included what Monster had in their database, and whatever private database you had. So, there was a matching component that matched whatever the job was to the candidates you had.

Jason: We used Monster's matching, which is ... It's what used to be Trovix. We use their stuff, and it's still good, still good.

Joel: While they pimp videos, Instagram for jobs, they have this awesome matching machine that they're not maybe leveraging as much as they could. But, that's maybe-

Jason: I'm not involved in those decisions. That's a different thing.

Joel: Yeah, let's stay on task here. So, you feed the resumes in the matching machine, the matches happen, an email goes out to all the matches saying, hey, we have this great opportunity. If you'd like to talk to us about it, click here. They click the link that takes them to a Wade and Wendy chat bot. They get chatted up. They basically apply through the chat. They get pre-screened. And if they fit, there's scheduling for an actual interview, from there?

Jason: If they're fit, we've treated it ... The way we built it was, we built it to work just like a sourcer, as if it were a sourcer. Everything I'm telling you is something I just delivered in a presentation at SourceCon. So, we treated it as a sourcer, so it actually sent ... In order for adoption to work, recruiters, we like recruiters to sort of feel like they're in their normal process. We set it up so that it would send an email with a writeup on the candidate, and attachment of the resume, and an attachment of the chat transcript. And basically said, hey, this is so and so, we think they're qualified for this job. Here's their resume. Here's my conversation. They're ready to go.

Joel: So it wasn't as granular as like, you fit what we need, let's schedule an interview. It went to a human, to then decide to whether, or not they should schedule something. But it certainly could get to a point, you think, where it does schedule, and-

Jason: That's actually easy. That's an easy next step. A lot of these tools use the same sort of API call for that with Cronofy. So, that's an easy step. But, we put it at the human side, because we can submit candidates directly to a recruiter, or the next step is we can submit them directly to a hiring manager if we want it to.

Chad: Yeah.

Jason: What we found was that recruiters did the same thing they do with sourcers, right? So sourcers send you an email with a candidate, there's a certain percentage of recruiters that don't even actually open the email. The sourcer has to chase them down for it, so that adoption wasn't what we wanted it to be.

Jason: So, we made pivots along the way, and figured out new things. But I think the bottom line is, the technology there is there for matching. It works like it's supposed to. And actually, that technology is a little ... It's been around for awhile. I was going to say it's dated, but it still works. So, it's not dated. It's still works.

Chad: Oh, yeah. Over a decade.

Jason: Yeah. It's been-

Chad: It was bought by Monster.

Jason: I'm sure they've done something with it. I'm sure they've updated it in some way. I don't have visibility of that. But, that's good. And the bot based stuff is getting really good. Wade and Wendy is a pretty understated bot in the market. You don't hear a ton about them, but I think it might be-

Joel: Full disclosure. Randstad is an investor in Wade and Wendy, or they're part of their ... ?

Jason: Randstad is an investor from their innovation fund. Yeah. And, that's how I was introduced.

Chad: AllyO, as well.

Jason: AllyO, too. Yeah.

Chad: Yeah.

Jason: I like both of those. I think they're both good.

Chad: Put some money on different horses.

Joel: Place some bets.

Chad: Yeah.

Jason: Well, I think you need a different club for a different hole. Right? So, if you've got a ... What? I said, hole.

Chad: He said, hole.

Jason: So, for if you need like high volume, you want to automate deep into the process ...

Chad: Right.

Jason: Oh, man. I keep going.

Joel: Keep going, Jason.

Chad: He's taking the hole deep.

Joel: I am. That just happened. AllyO's the is one for that.

Chad: Okay. AllyO, taking the whole deep. That's the new slogan.

Jason: I think they should go with that. They should absolutely do that.

Joel: And the O at the end of their name fits perfectly for the visual on that, I think.

Chad: Yes. AllyO.

Jason: Man, I'm going to get a phone call from Ankit, when this is all said and done. He's going to ask what the hell's going on. Isn't Sahil here?

Chad: I don't know if he is, or not.

Jason: He's like on your Death Match.

Chad: First and foremost, you're welcome for all of the marketing help.

Jason: Thank you. Well actually you know what, say that to the AllyO guys. That's theirs. They can thank you.

Chad: Yeah.

Jason: So my pick for sort of the non high volume, the white collar stuff is Wade and Wendy. That's where I go for that. And so ...

Joel: So, what else at SourceCon caught your eye?

Jason: Jim Stroud just never ceases to amaze.

Chad: I love that guy.

Jason: And, he comes up with weird stuff.

Chad: Yes.

Jason: And, here's the bottom line. I thought people were going to freak out, because I was showing this automated sourcer, and so I've made a point to point out some of the stuff that's out there. Maybe, we don't need automated sources all the time. Right? The automated sourcer actually didn't fill all the roles, and it had the same struggles that that human sourcers had, but it can't do things like be Jim Stroud. And, Jim Stroud had this thing that just ... It's not a technology thing, but it blew my mind. So simple. He had the Christmas card method of finding emails and phone numbers. You want to know what that is?

Joel: I do.

Chad: I want to hear this.

Announcer: It's commercial time.

Sovren: Sovren is known for providing the world's best and most accurate parsing products. And now, based on that technology come Sovren's artificial intelligence matching and scoring software. In fractions of a second, receive match results that provide candidate scored by fit to job, and just as importantly, the job's fit to the candidate. Make faster and better placements. Find out more about our suite of products today, by visiting Sovren.com that's S, O, V, R, E, N.com. We provide technology that thinks, communicates, and collaborates like a human. Sovren. Software so human, you'll want to take it to dinner.

Chad: It's show time.

Jason: If you're trying to do get phone numbers for an organization, you figure out what their email, sort of template is. You send emails to massive numbers of people in their organization on a major holiday. So, you send it on Christmas. And read all the auto responders back, where people put their phone number, and the who to contact, and all that stuff. Is that not genius?

Chad: Yeah.

Jason: It's super basic. It seems like it's a no brainer, but I didn't think of it. Jim Stroud.

Chad: Super basic. That's why you need guys like ... I mean, guys like Jim Stroud-

Joel: That's evil genius level.

Jason: Right?

Chad: