Chrissy Wilson! Live on Shally's Alley

Hey, all right everybody, hello, how are you doing? You are live on Shelly's Alley. That's right, this is not a recording. Now this is a recording. No, it's not, just kidding, still live. Welcome to Shelly's Alley. It is Friday, October 11th, and finally we get a break from the oppressive heat of the summer. It looks like the weather is turning. Maybe all those winds actually did something. I hope you, uh, you stayed safe, Chrissy. Chrissy lives in Florida, and I guess there was a scare for the hurricane recently, but it missed you, or it missed me.

I'm in West Palm Beach, so I'm on the Atlantic side, so it missed me. But we did get some rain and, you know, uh, some, some winds, but we're good. We're good to go.

And we finally got some cool weather here, although the hurricane before this did not miss us. There was a lot of damage. But definitely changing seasons. So everybody, glad, you know, thank you for staying safe. I'm glad you're back. Our guest today is Chrissy Wilson. Chrissy was actually introduced to me by David Marr. You guys know how I ask at the end of the show for recommendations, and that's how her name came up. So we reached out and connected and learned a little bit about her work. She's been— she's done sourcing at Uber and Disney, led sourcing teams, originally from Toronto, but as she said, lives in Florida now. And she's currently doing some consulting work, but open for new opportunities for a contract or full-time. So if you are looking to hire someone who's got some strong sourcing chops according to David Marr and myself, then you've got the right person. What I wanted to talk with Chrissy about today is AI tools for resume screening. Many of you know I'm a skeptic when it comes to AI actually really being AI. And I know that machine learning has a lot of applications. So I just kind of wanted to have a live conversation about her experience with resume screening AI tools. That, you know, I just haven't really been trusting a lot lately. So hopefully, I can learn a little bit about that and you can too. So, Christy, tell me about how you got into AI recruiting tools. Like, how did, you know, did this get forced on you? Like, you got to use this? Or did you go out and look for it, seek it out?

Well, so let me ask you this, Shelly. AI has been a part of recruiting for a really long time. If we think about the ATSs that we've been using and the CRMs, it's been around. Like, this is how we know what resumes are at the top in terms of in the ATS or CRM. It kind of filters out. So I feel that it's been around. I might be mistaken, but it's been around for a while in terms of how we kind of sift.

I mean, you're absolutely right. There's really nothing new about what we call generative AI or what we call, you know, the AI tools of today. They were by and large invented in the '40s, a big part of defeating the Enigma cipher that the Axis and Allies were using to, you know, in the war. So honestly, the technology hasn't changed much. What has changed is the scale. So it's, it's still very much the same old decision tree, logic tree. You know, what happens if this, then that, you know, make decisions, label those decisions. But because of the scale and the amount of data, it has much more practical applications nowadays than it did back then where, you know, it was much harder to get the data in. Now, now that data access is pretty broad. So you're right in that it hasn't changed much other than the scale, but that is what's given it the power nowadays. It also is why I'm skeptical about it is because of the risk involved with the training data and the, in many cases, gaps and absences of data that human beings would be able to detect, but machines trained to do the same thing over again, over and over again. That's right. So that's where my gap is. So, but you're right. I mean, you know, sorting, sifting, grading, scoring, that's been a big part of dealing with inbound applicants for a while.

That's what we do. That's what we've been doing. Now, I joined a company, I believe it was 2009, called Talent Drive. And they were trying to automate and do this on a bigger scale and create these spiders. And I was a part of the startup. Of that. And what we found in the pilot period was that there was a, like you said, that these machines or the AI cannot read gaps. They cannot, they can't fill that in. And therefore when we were getting back, these resumes or profiles, if we were going to rely on AI, this person would've been an absolute no. When in fact, so we still needed human eyes. Right? So it was almost like, okay, I don't get the point of this, but I get the point. It's going to assist us in making better decisions.

That's right. It's a tool to help us do our job.

Exactly.

But like every other tool, it has its applications and it has its limitations, right? I mean, you know, I've always said— people ask me, um, what's my favorite sourcing tool? And, you know, facetiously, like, the internet, right? Like, it's— there's not a single tool. There's just so much that the data is out there. And AI is really just another tool. You know, you can't rely on job postings alone. You can't rely on, you know, passive candidates alone. You can't rely on telephone sourcing alone. You've got to combine these and apply the appropriate amount for each lever for the mix of candidates that you need in your organization. But there is definitely some truth to the fact that we have become, I want to say, complacent to the algorithms. Yeah, people are becoming more, you know, recruiters are becoming more, um, accepting of these rankings, which I think is the— that's the danger, that's the problem.

It's very dangerous. I'll say this, at, at, um, Uber, we use iSims and Beamery, um, great tools by the way, not knocking them, great tools. But I would be one of the recruiters that would actually try to empty the box because I realized people that were ranked further down, absolutely a match. There was just something maybe off. Again, if you're reading buzzwords, and you're not putting transferable words or skills in there, you might miss out on a golden applicant because of that. And absolutely, it takes me so much time because I will go to the bottom of the list. And I'll be like, this guy is bang on, but If you're writing your resume and you don't know how to put those buzzwords in, but you actually have done the job, and us as recruiters and consultants and thought leaders, we understand you can put software engineer or software developer or Python engineer. AI is not going to pick up on that, right? So I'm skeptical too. I see the hype and I don't buy into it. I'm like, you still need sourcers and recruiters and TA people to execute and deploy. And right.

Yeah. And then you do need human oversight. And, you know, we're in a space now where we really— it's our responsibility to hold the government accountable, to create the oversight that doesn't exist yet, because we— this really is something that we need oversight for because primarily the vendors— and this is— I've been writing and talking about this for a while now. The vendors are all trying to sell you something. Hey, if I was a vendor, I would try to sell you something too, but they're going to sell you this technology and make you feel like it's going to solve your problems. And so unfortunately, there's not a lot of transparency that comes with that other than like, well, you know, I had this one vendor pitch me because I evaluate all these tools, right? All the time, constantly for my customers and for my own use. And this one vendor was pitching me like I asked about the bias and they're like, well, we've evaluated 1 billion profiles. And I'm like, you hold on a second, there's 7 billion people on the planet. So you're telling me that you have digested 1/7 of Earth? Like, I just am not buying that. But wait, let's assume that you have. Did you happen to get those 1 billion profiles from LinkedIn? Because I know for a fact, at least half of the profiles on LinkedIn are duplicates or straight out fake.

Absolutely.

So you're evaluating, you're feeding your learning algorithm with fake data.

So yeah, so the data is not correct.

The data is not correct.

It's not correct. And, um, it, it feels like, yeah, there is a big push to get these, um, these platforms out and get everyone hyped. I think us OGs, we can see, we see this, we see right through this. We're like, this is really not so much different than what we've been doing. And you still need that human touch. You still need someone to go in and view these resumes. Talking about bias, I mean, they're saying that recruiters, we can have unconscious biases, but of course, who is learning, they can also establish bias.

No, the problem is that the AI doesn't know that it's biased. Right. So With humans, it's something that I learned some time ago. There's always going to be bias and we are all biased. Everybody is biased. It is actually, it is necessity. It's a necessity. We need it to survive. Bias exists because our brain needed a way to avoid the lounging tiger that could eat us in the jungle. Now, there's no more jungle, but we still live in an environment where we need to detect these patterns. And so, bias helps us make these predictions. But you have to be aware of the bias. And if you're aware of the bias, you can manage the bias. So, it's not about erasing it. It's not about removing it. It's about working with it and around it to make sure that it doesn't make you— give you bad information or make you make bad decisions. It's more about like, okay, I know that this algorithm tends to lean towards a lot of like profiles from people that are already working here. And, you know, we wanna hire people that are successful and people that are here are successful, but at the same time, we also wanna hire people that are diverse and that bring a different opinion. So they are probably not gonna look like the people that are already here. Therefore, we need to make room Yes, we can use the AI to stack rank and get ahead and, you know, get some work done, but we still have to go back and, you know, allow some of those manual, uh, selectees in. Otherwise, we're ignoring the ones that are— I keep thinking about Sesame Street, which one of these four doesn't belong, right?

Right, right.

It's up to us to still kind of do that.

We're missing out on huge talent if we just solely rely on AI and then we use these excuses as to why, uh, it, it is better. It is a tool that should be used again with the human touch and with, with professionals and, and, and TA, TA specialists. And it, it aids in our job.

It is our job, but you have to be aware of— I don't know what can it do, what, what should it do, what should it not do, what to look for if you see this, right? And my biggest, my biggest piece of advice with the AI is no matter how skeptical you are about it, you can always do a random sampling. I mean, there's— that, that's what I like to do. It's what I do when I test these tools out is I do a random sampling. I just toss something in there and be like, all right, let me, you know, let me see if I'm gonna try this because I know it's a, I know it's a fringe possibility, but let's see what happens if we throw it in there. And if the algorithm doesn't pick up on it, then I know that there's some bias which I can then work against. So, you know, you can still work with a biased algorithm because you're aware of it and you can then, you know, implement procedures to work around that bias or defeat that bias. But I think there's also a big misconception about AI that it's, first of all, that it's thinking. And secondly, that the same thing that is AI is also automation. AIs don't think. Automation is different than AI, right?

Right.

Yeah. So how did you end up kind of— what drove your passion around understanding these AI screening tools? What was the onus for you?

The onus was the nonstop rhetoric of AI is taking over our jobs and we need to jump in and kind of learn this right away.

So you're like, let me peel the onion and get right into it.

I'm like, that's not— It's not happening. Like, let's be real about this. So just, you know, there was such a hype around it and I don't get excited about much, to be honest. Anything that's really—

show me kind of person. I mean, to— okay, show me. Yeah, sure. Okay, show me. Yeah.

So what is— I've gone to Copilot.

Here are the keys. I'll test drive it for a week.

Exactly. So I've gone to Copilot and I've said, build me out a Boolean search for diverse talent. Da da da da da da. So whether it's veterans or whether it's female in technology, and it did a basic job, but it's kind of the job that I would have done. So we're expecting these miracles and this magic pill, and it's like it's not doing anything other than what we would have done. It might speed up the process. Right.

I was going to say maybe it is what you would have done, but it can do it 10 times faster. So now you can do what you would have done 10 times.

And that's the only thing that I see that's great about that is that it does cut down on the time. Right. It does cut down on the time of us manually doing it. I need to go in and refine the search. I still need to go in and see, are these people— is this true? And a lot of times it's pulling stuff that's not 100% true. Then I got to go and twerk. Okay, let me put this in. Let me put tweaks or not twerk. That's awful.

Hey, I wasn't sure if that's what you do to get candidates.

I was doing that before this call, by the way. I was listening to some candidates.

Gotcha. All right.

But I got to go in and tweak the string just to ensure that it's pulling what we need it to pull. So it's a tool and it's great and it helps, but I do not buy that it's going to take over and it shouldn't. It really should not. And it has to— there needs to be kind of guardrails around this as well as we're pushing this. I feel like there's not—

the guardrails are missing.

Yeah.

Well, how do you feel about this? I've been kind of mulling this over in my head and people have asked me speaking at conferences and for, you know, to present about this topic about like what's the, you know, what's the next evolution or what's the future of recruiting, etc. Um, and is sourcing going to be eliminated, right? Is this— comes up a lot, especially when it's a vendor trying to sell tools. You're like, oh, we're going to eliminate sourcing. If my job— my job's been eliminated so many times since it started that, yeah. But what about this? I keep going back to this idea. What made someone a good sourcer back in the day when you had to, uh, constantly tweak your searches is also what makes someone good or adept at using ChatGPT and prompt engineering, because it's basically the same thing. We didn't call it prompt engineering, but we were Google buffing or whatever. And it's kind of the same thing. So instead of Google Boolean engineering, we're now prompt engineering, but we, we think that way, right? It's that trial and error. So isn't— I mean, isn't it possible that maybe, you know, whatever happens, our jobs maybe, you know, just shift and become more of like we're, you know, the sourcing or the function of sourcing gets more involved with making these, leveraging these tools, right?

Right. I believe that. I don't believe— I believe there's a reset and there will be a little bit of a shuffle, but I don't believe for one minute that sourcing teams will be eliminated. I feel maybe we won't have these huge teams like we, we did because now we have a tool that we can use, but we still need to go to a different, like if you're like me, you're a little ADD and we're going to different sites at different times and pulling this up and going to GitHub and Overstack and just all this stuff, it still requires a professional and it's, it's not gonna be eliminated. I don't buy that for one minute.

Discerning eye. Right. And what, so what about the next step? Right. So resume screening, profile screening, applicant screening, applicant ranking. I mean, you know, before, any of this machine learning model-based stuff. We had questionnaires that we would ask candidates, you know, when they applied for the job. We would ask them a bunch of questions and rank them based on their answers to those questions. So, you know, we've been trying to do this for a while. But that's not the whole process. You know, there's this other step that has to do with once you find somebody, getting them to actually talk to you or engage with you. And is AI helping us with that?

You're asking the wrong person because I'm so, so kind of like on the, on the fence. But I feel—

why are you on the fence? So your answer is AI is trying to help, or people think that it's helping, but you're skeptical. So why are you skeptical?

Well, it, it is helping, but, um, the whole thing of it's replacing us, I can't wrap my mind around that. I know that it's assisting with Sure.

Forget that though. Is it actually assisting us? Right.

It's assisting us and vetting people out. But even with that, if someone does not put the correct answers on, say, automated screen, they're out of play. And this person could be perfectly fine. Now, if they spoke to either me or you or David or someone, we could say, yeah, this guy is really good. He might not have answered it you know, use the buzzwords, or, um, you know, I, I'm skeptical of anything that is screening kind of like human intelligence. Does that make sense? Like, we— you still need—

that's what's— yes, it does make sense because, you know, these are chatbots, right? ChatGPT is a chatbot. And before this sort of blew up, there were already chatbots in recruiting, like, um Olivia, and, you know, and they were chatbots. They were answering candidate questions. And did those actually work as well as we thought they did?

I don't believe so, because I would loop back and have to ask the same questions and then get a real answer. And this is what they actually meant. And it kind of doubled the work sometimes. Right. So what I'm getting is, are we going to get there? Right. I think we I can't answer that. I don't know. I feel like we're still dealing with your— if, if it's just compiled data, can we really get there?

That's my question, is can we really get there? And people are saying, a lot of people are saying that we can get there. I'm thinking, well, it's still all based on a logic tree, so unless we've predicted every answer, we really can't. But, but they're saying that we can because of the context, or we can get to Most of it. So my next iteration after that is if we can automate a chatbot on our company's career pages that talks to candidates before they apply and helps them apply, helps them find the right job, then why can't we also do that with email when we're contacting candidates? And if we're able to send out email messages and automate the response and actually have an engaging conversation with a candidate via email, at the hand of a robot, right? Is that what candidates want?

No, that's not what I want. I certainly don't want that. And I want—

so if we can do it, that's my question. Let's assume we can. Let's say we solve the chatbot thing and it's working, and then we move to email and we start doing it that way, and we're able to do it. It's possible. So, you know, I'm not, I'm not even playing the skeptic anymore. I'm not saying, well, I don't think it's— I'm saying, let's assume that is possible. Do we want to do it?

I personally wouldn't. I have turned— I have like stopped filling out applications when I see there's like, there's a video that needs to be done and it's not to a person. It feels so— there's no human touch. It feels like you're not connecting. HR and TA is about connecting. We're connected. So to have a chatbot or AI answering questions, it feels very impersonal. And I feel like I'm not getting my story out there. So I'm really— again, I'm okay with using it as a tool, but I'm not really interested in going back and forth. I've even done it when I've called the IRS and they're like, okay, we can answer your questions. And I'm like, no, I want to talk to a person. Like, I want to talk to a person. I want that human touch. But that's just me. Uh, I'm an OG, right? Um, we can get there. I'm sure we can. But do people want to have it automated, or do you want to speak to a real person and, and, and, and tell your story?

This is my concern. This is my concern, that it, it's sometimes— and I've seen this already happen— misused. In a way that it is obfuscated. So they make it look like it's a real person. And so that kind of creeps me out.

It does.

Let's assume that the AI gets smart enough to be able to answer most emails and, you know, the AI finds the candidates, sources them, automatically sends them messages, automatically responds to the messages. Schedules interviews, all the stuff that, you know, we think that it, you know, that people are saying it's going to do, right? And along the way, you're very clear and you disclose it. You say, hi, you're getting a response, you know, hi Chrissy, you're getting a response from an automated, you know, system, right? I'd like to ask you some questions. You might be like, okay, expedite the process. Yeah, hey, you know what, I'm okay with that because normally I don't get a response from a recruiter. So at least I'm getting a response. It's a robot, but at least I'm getting a response. So I'm gonna go along with it. So we go back and forth with a couple of things like, okay, well tell me a little bit about this. What's the compensation? And then all of a sudden the robot comes back with, you know, some weird answer or not an answer. And now I'm like, I'm at the end of the rope. Now I want a human. I go back to the black hole.

And so you've wasted all this time to go back.

And I got nowhere. And I got nowhere.

Yeah. And then they have the audacity to say, did this help? No, because I need to call now or I need to email directly. So it's in that pilot process and what we're seeing, and it's not even a pilot, this has been going on. I'm always going back to reverting back to, can I speak to someone? Because my question was not answered. And I almost feel when I find out that it's not a person, like I've been deceived.

You know, yeah, well, that's what I'm saying. If, if let's assume that we're, that we get that good and also that we're disclosing it so people don't feel that, because, because if I feel deceived, then that's when I'm like, this relationship is over, right? You know, uh, if, if I'm okay that you're a robot, at some point that it ends, at least I knew that you were a robot. Uh, David Marr is saying that, you know, email is slower. I, I don't want to disagree But I want to state that what I was talking about is a different option, right? If somebody is communicating with me via instant messaging, chat, text, I'm already engaged with them. But I'm talking about an AI sourcing a candidate and sending them the initial outreach, which, you know, may have to be via email. Maybe the AI then says, may I text you? And then you switch to text. So I get what David is saying.

Right.

But I still think that at some point there's got to be a back and forth. And even if that ends up being chat, Messenger chat, or, you know, some kind of unified messaging, I still need to disclose it. And I'm still going to run into situations where I just, I want out of the loop, I want out of the robot.

Yes.

And then what? Like, there's not going to be enough recruiters. So think about this. We're trying to reduce the amount of recruiters that we need to get the job done. But we're gonna, we're gonna need recruiters to answer the questions when the robot doesn't answer them. So we're gonna need to hire those recruiters back again.

Exactly. It's, it's not going to replace us. It's just not. I, I've seen my experience with it, and again, it's not just because I'm on the fence. I've tried, and I've— it's, it's showing to me that I'm always going to revert back to, I need, I need this answered, and it's not, I'm— it's not being answered. So let me go back to the recruiter and I've just wasted 20 minutes. And again, it's not, it's not a waste of time. There's things that I think we can allocate to it to do.

Right. Like, okay, hey, how about this? We just go to like the minimal qualifications.

Yeah.

Hi, Chrissy. Thank you for your interest. In order for you to qualify for this job, you must have a Bachelor of Science degree. Do you have a Bachelor of Science degree? Yes. Okay, great. In addition to that, you must have 5 years of experience in, I don't know, data analytics. Do you have that? Yes, I do.

Yes, no. Yes, no is fine.

Right, right, right, right, right.

And then, like, you need to answer something.

What if at that point the robot says, well, thank you for answering our questions, our preliminary questions. What I'd like to do now is book an appointment for you to speak with a recruiter. And then the robot puts them on the calendar That I think can happen, but again, that doesn't happen until I've already made contact. So I'm going— this keeps me bringing, bringing me back to, you know, Olive or Olivia or whatever, those chatbots. Those chatbots work because somebody landed on the page and asked the question, right? I don't think with, with sourcing we have to send the initial response. It— so it's kind of like the difference between an automated teller helping you with customer service versus, um, somebody actually coming to your door, door-to-door sales, or somebody inviting you to the dance. That's a robot. That doesn't quite fit.

It doesn't translate well. So you're right, there's areas in which it would be effective and it would work.

So I think yes/no pre-qualification questions that we would ask as sourcers to knock it out, or choose one of the following, or even interpret like Tell us a little bit about this and interpreting. Yeah, right.

If you're looking for $150,000 and this is only $100,000, then you're out. Like, things that are clearly yes or no, and then we can move on, absolutely. Anything that requires, I feel, more of a dig, no, absolutely not. Because there's no subtext.

Subtext instead of—

right, right. And we may get there, Um, but we're not there right now, and so I'm not worried about sourcing teams being eliminated.

Yeah, me neither. I just wanted to kind of explore, pull the needle and thread out of this, you know, this conversation. I want to— let's turn to using data to improve hiring decisions, which is not— so moving away from kind of the automated part of resume screening, more like using data to inform that. So tell me a little bit more about how you feel about that and what, like, what tools or category of tools do you recommend for that?

Well, the best tool that I've used to date, and I don't know if you're going to agree or not, is Talent Neuron in terms of getting data and marketing talent, market intelligence. Right, right. So that the talent intelligence where you need to tell a story, you need to go to the hiring manager and say, hey, In this specific market, I know you're looking for, say, I don't know, 30 big data or Scala developers. Unfortunately, in this market, this is what they're going for. These are the competitors. This is what you're looking at. I think data tells a great story, but I don't know if TA has bought into how much we can use data to kind of curve our hiring or be proactive in terms of what we want to do and what markets we want to burst into. So data is powerful, but I'm— and I know some great data-driven recruiters, and you're probably— you're one of them. And I know that data has saved me going into these cadences. And where's our candidates? Well, based on this location and this, you know, these skill sets, it's tough in this area. So data will always save you, but are we using it? Are recruiters and thought leaders using data to really set up? I use it. I have to in these cadences that, you know, when I'm not finding— right now I'm looking for a quant trader in New York, and I now have to come back with some data to let them know, listen, the comp that we're looking at, you got to see what these guys are getting paid out of New York for.

Okay, so let's pull this. Let's take this example through. I get compensation. We've done that before without, you know, we've used sort of stale salary services that are all, you know, by the time we get them, they're already 6 months old. So in that regard, having much more current data, much more like, you know, today's market for compensation is kind of an established thing. But besides that, what other data intel are you gathering from these tools to help you in not just in your recruiting decision-making, but in advising the hiring manager and perhaps the entire organization? What, what are the other pieces of— besides comp, the other pieces of data?

Uh, skill set.

Uh, so the skill set of the market, like what's in the market, what's available, that specific market.

So if we're looking for specific skills, this skill set in this geography, right, it's not working, and so we might have have to scale back. We may have to look at something else. We might have to go remote. If you're looking for someone that is on-site, we're not going to find that. So skill set, even—

so you're going to advise on let's change geography, or let's, or let's reduce the requirements because it doesn't exist, or—

okay, exactly. Specifically, I remember at, at, um, sorry, at Uber, they were looking for ISO developers and Android developers and we were just having a very hard time and they wanted them on site. And we— not we, my colleague, because we discussed it, convinced them to do remote. But that was because we had to, to provide the data as to why. So data will always— good data will always save you. We, in fact, even at Disney, had a consulting firm that would pull data points for us so we can go to the hiring manager. So they hired an outside firm to do it. So they realize that data is powerful. So absolutely. So for skill set, even job descriptions, like, hey, look at what, what's being posted in your industry, what the competitors are doing, and what the competitors are doing, we need to get on board because we're not getting anyone.

So it's informing you with the what? What compensation, what skills, what job scripts, right? It's informing you of the why. Why are we not getting any results? How about the how? Does it give you any how information? Like, does it help you actually change your sourcing criteria? Just you, like, are you using market intelligence to use to find different keywords? Is it helping you with finding hidden talent pool pockets? How is it helping with that?

So even like with where I source, obviously will depend on what I'm sourcing for. So I'm finding for developers, I'm going to specific hubs. And I only kind of figured that out through data. Like, they're not all sitting on LinkedIn and Indeed and ZipRecruiter. They're actually not. They are specific to GitHub or Stack Overflow or other, um, software-specific sites. So absolutely, my, my strategy changes based on what I'm working on, and I will go to get data to kind of quantify that and, and to confirm, okay, you are definitely in the wrong market in terms of what you're looking for.

Gotcha. So it's helping you build a— it's helping to build a case for how and adapting your search and adapting your, um— does it ever help you with the approach?

Yes.

With your messaging?

Yes. I feel like, um, on GitHub, there's a— my approach is different in terms of my messaging than on LinkedIn. LinkedIn, it seems a little more casual. Not casual.

I get that. I get that. Right.

And GitHub, it's a little more casual. I'm speaking to these developers and they don't have time for all the, you know, all the corporate stuff. But then also, like, I've used data to ask the corporations to open up different platforms that they wouldn't necessarily open up and pay for. So I'm like, okay, we don't have an account for this and we're wasting time. And again, not knocking any— this is not knocking.

No, no, I get it. But you're like, we're not getting results, we need to buy something else, paying a crapload of money for.

So let's open up this platform. Okay, show me the data. Well, this is where they're all housed, and that's helped in terms of opening up a platform for me to be more successful in terms of my sourcing.

Okay, okay. Have you ever used this kind of data analysis to change the, um, adapt the, uh, what you're pitching, your EVP, right? Like, for example, would it be sentiment analysis? You know, here's something that I'm thinking. What if we're going after require, you know, we've got a requirement for a particular type of talent. It doesn't matter what the skill set is, but let's just say, you know, financial planners or something like that. And we're able to determine through these kinds of talent neuron and these other tools that are out there, Public Insight and so on, that there's like a negative sentiment towards a particular employer or a particular set of employers or whatever. Can we use that knowledge to our benefit when we're messaging candidates? And I'm not recommending, by the way, that you say, hey, I hear you're really down on your employer. No, but maybe, you know, counterpoint, be like, hey, people love this company. It just kind of hit them where it hurts in the wind. Here's what you're missing, you know?

Exactly. So, the information that you're getting, essentially, you're getting the weaknesses and the weak spots. Right.

And then you're picking your strengths against that. Right.

And you're playing up the strengths of, okay, do you want a healthy environment? Do you want an environment that compensates? Everyone does. But you're playing that up without saying—

I get it. Yeah, it could be benefits. It could be, you know, there's a lot of compensation, non-compensation benefit related choices that people make that could help, you know, you attract a specific kind of talent. I know generationally— I was going to say geographically. What I meant to say is generationally, there's some differences, right? People from different generations value different types of benefits more than others. So, if you know that, if you know that for this target skill set in this target geography, the generation tends to be this group, then your messaging can be adapted to the voice that they want to hear.

I'm sorry, guys. Turn this off. I thought I turned it off.

I can't hear anything. So you're good. All I see you is bobbing up and down.

I'm throwing my cell phone over there. I'm like, who's calling me? No, no, you're absolutely right. So data is powerful and it can be used in messaging. It can be used to play up what you want to convey. And I won't use poaching because that's so old, but essentially you're trying to—

Yeah, right.

It is. But you're trying to go after talent. So you need to arm yourself with data, period, end of story. Like, what does the company that you're working for, what do they stand for? And if there is something that's going on in the background that's negative, just like, okay, for instance, Amazon is saying everyone needs to come back to work on site. And I've seen recruiters straight out be like, hey, if you want to work remote, come over here. They're using data essentially to get that buy-in, which makes a lot of sense. Like, there's a lot of people that want to work remote. So absolutely, I believe in data. Data tells a story and it helps with, with your job. It helps with what, what you're, you're trying to do as a sourcer. And I feel like sourcers are like kind of put in this box of, okay, we just find people. No, that's not the case. We, we do more than that. I know at Uber we were finding, but it's competitive intelligence.

It's analysis, it's market data, right?

It's very strategic. Like we, we're planning out our day where we're in those, those on those cadences with the hiring managers getting this information. And then because we're so close to the ground and so close to the market, We're the ones coming back with the data saying, hey, this is what we're seeing. It's been 60 days and there's not someone in this role. We know you're losing money, so let's figure out why. So we are just more than headhunters. We are strategic thought leaders and we know how to find people. And we're essentially, we're all full cycle recruiters, but we've decided to go into sourcing.

Specialization. Yeah. And in my case and the case of others that I've you know, where I've built teams and stuff, it's, it's common to hear that there's aspects of, uh, the recruiting that we don't enjoy as much or we prefer not to do. So, you know, it's a focus on that. I want to do more of this, less of that, which happens in a lot of industries. Um, you learned a lot of this by yourself. Where— point, point us to what resources we could use to learn about the data, auto, you know, automation, data, um, analysis, things like— where did you— where, where do you go figure this out? Because online, like Google. Google. Okay. Yeah, you're just telling me the internet. Yeah, that's—

yeah, I, I Google everything. I, I, I walked into that one. I attend things like this when you're doing other, you know, with other people on your show. I attend conferences like that are online. I network, I talk to you, I talk to David, I talk to people within the industry and get tools because there's so much out there. It's so overwhelming. It's very overwhelming. And there's so many tools that can— are probably better than the ones that I'm currently using. But if you're not constantly upskilling and constantly knowing what's going on, you can be left behind very quickly. So I feel like these conferences, uh, are great, uh, networking, getting on LinkedIn. Obviously there's a ton of information there, people always share. Um, I'm, I, I'm AIR certified, I have like 6 certifications, um, through AIR. So a lot of the stuff that I've learned comes from there as well.

So you're paying for training, you're listening to chats?

Yes.

And what I'm hearing here, the pattern here, is that you've got you've got an overwhelming drive for, you know, curiosity. And maybe that's what some folks are missing that are, that are not learning this stuff on their own, is they just need to kind of, I don't know, take a curiosity pill or something.

Yeah, I always feel there's a better and faster way to do things. So I'm like, let me, let me find, let me get on the call.

Start by asking, right?

Yeah, exactly.

That's good advice.

You know, what's going on? How can I do this better? Like, for instance, the Boolean searching that I did in Copilot, I didn't know it would work, but I was like, okay, if AI is such AI, because it was me being cocky. Yeah, show me, show me. And then I'm building out my string at the same time. And absolutely they were faster. And some of them, they nailed it, the more basic strings. But when it got complicated, but it, you know, me just being curious, okay, can, can Can this, can this platform do this for me and help me cut down on the time that I'm like, and, or—

So you just took the risk and tried it. Yeah.

Yeah.

So just, I'm always just curiosity and, and, and take it, take a risk and just experiment, you know, try it out for yourself. Don't just kind of take it for granted. And David makes a good point. There's only so much time to test out new tools alone, but you know, you've got to start somewhere. So at least do one. I think maybe what's missing is the, is the drive to ask the questions. So folks out there, like, you know, you heard it here first— get curious, take some curiosity pills, get curious, and then join groups.

Like, I've just started a bullying, uh, bullies group in Florida, Florida chapter, but I've put it online where we can share information. And there's tons of groups like that out there where, um, there's bigger platforms than mine where you can go and there's things people are posting and sharing. That's how it should be. You're kind of learning on the go, but you don't always have to do the research. There's people that have done the research and like, hey, this is a great tool. So you really have to just take control. Like David said, you have to take the initiative and take control of your future. Upskilling and being and staying on top of stuff costs money. Like I, I've had a year worth of, you know, a database access to ADP errors. And all I did was take as many certifications as I could just to be on top. So military veterans, I know how to source for them. Technical, obviously I know how to do that. Diversity recruiting, I know how to. So it's literally just the, the need to want to get better and want to improve in your, in your craft. I want to be the best, the best at what I do. I've been doing it for so long. I don't want to be mediocre. Who wants to be mediocre? I want to be the best, right?

That's that. Shoot for the middle and you'll always get it, right? Right. Cassie says, uh, a mentor kept telling the interns to stay thirsty. That sounds oddly like a, uh, Mexican beer commercial. Stay thirsty, my friends.

Stay thirsty, my friends. Yeah, you, you have to be thirsty. And I think, like, I started recruiting, say, 2004. And I'm still thirsty. And that's why I know that I have a passion for this. It's been a long time. I've been in this business for a long time, and I, I'm still growing. I'm still—

that's a long time.

Yeah, I, I just aged myself. Don't look at—

don't look at my background.

I saw your background. I was like, oops. Yeah.

Anyway, hey, I promise you I was only going to keep you 30 minutes, and of course, in pure Shelly tradition in Shelly's Alley, I always go over except for that one time. But, um, I do want to make sure that I have time to ask you the two questions that I always ask everybody. The first of those is how I met you, which is who should be our next guest?

I have— can I give you two people?

Yeah, because one or both of them may say no.

Exactly, exactly.

Just because you recommend them doesn't mean they're going to want to be on here.

There is Liza. I believe her last name is Valentine or Valentine. I'm sorry. I work with her at US Cellular. She's an amazing person. She's a talent leader. She's a diversity champion. Great professional. I learned so much from her. I think she would be the next person I can send you over. Yeah. And then Glenda Giles. I worked with her. I was actually her manager at Disney. And she is in the legal space and she's in Florida as well. And the legal space is just something totally different. And I think maybe she can shed light on some stuff that we've had.

Anybody? Dials. D-I-A-L-S. Yeah.

Yeah.

So awesome. Great. Well, I will definitely reach out to them. And of course, the last question is, you know, I pride myself on being a bit of a thought provoker. So, I want to give some folks, something to chew on over the weekend. Give us something that you want us to ask ourselves and think about over the weekend.

So I've been going through a little bit of a tough time, and I went to singing bowls. I don't know if you know what singing bowls is, but it's a sound bath yesterday. And there's something that was said that was thought-provoking, and it said, not all storms are here to destroy, but to clear your path. And so we might be in a storm right now, uh, from the team.

Listen to a question though, we got to ask ourselves the question.

Yeah, is this storm here to destroy or is this to clear my path?

All right.

And if so, how do I get ready? How do I get ready to walk that path?

Nice, I like it. Is this storm here to, to make a mess or to clean up a mess?

Or clean up a mess.

Yeah, that's great. Oh, that's a pretty good one. Yeah, deep thoughts. But, you know, the answer is for yourselves. Um, so cool. Well, thank you so much for being on the show. I hope you, uh, you have a nice long weekend if you're able to take Monday off, which is not probably not a holiday for everybody, but it is for some folks. And, um, for those of you that are coming back next week, we're going to have Aaron Brown as our special guest. So once again, thank you, thank you for listening in, and be kind to yourselves. Take care, everybody.

All right.