S4 E59 | eXpert Tease: Sourcing with Boolean Blackbelt Glen Cathey

Announcer 0:00
Welcome to the CXR channel, our premier podcast for talent acquisition and talent management. Listen in as the CXR community discusses a wide range of topics focused on attracting, engaging and retaining the best talent. We’re glad you’re here.

Glen Cathey, Randstad 0:17
I want to welcome everybody to CXRs eXpert Tease. I’m Chris Hoyt, president of CareerXroads, and I am bringing to you a new segment of our podcast that sits us down for just a few moments with industry leaders, to really talk about some of those bigger their biggest victories, their career challenges, and hands on how to sessions. And we do this in less time than it takes for you to enjoy a full cup of coffee, just 15 minutes. And that’s why we actually call this an expert teas. All of our discussions have focused topic. They might be workforce planning, strategic talent, ethics and recruiting. And of course, like today, DE&I challenges. So if you’re with us live, we’re encouraging everyone to add questions to the chat area have a broadcast throughout the talk. And at the end of the segment, we’ll take a question or two if we have the time. And then we’ll move the rest of the conversation online to the talent talks open and public exchange@www.cxr.works/talenttalks. So I’m pretty excited to jump in today for hands on how to session of expertise with none other than friend and industry colleague, Glenn. Cathey Glen is the head of digital strategy and innovation at Randstad. And is also known to some of us who’ve been around a while longer than others as the Boolean blackbelt. Glenn, thanks for joining us. Good morning.

Thank you very much. Thanks for having me.

Now, we have surveyed hundreds of multinational companies this year, and we asked them what was top of mind for them. And what they were worried about most what they really wanted to lean in on over the next 12 months. And one topic that came up a few times. Well, more than a few times was attracting and finding diverse talent. So you’ve got some things you want to show us with regards to finding or sourcing diverse talent today? Yeah,

Absolutely.

Chris Hoyt 2:02
All right, I’m gonna let you get to it. So so give us your thoughts on actually finding diverse talent, and then walk us through some of that,

Glen Cathey, Randstad 2:08
Sure. So I have to always kick off with some foundational elements of search or information retrieval, right. What I’ve often like to share with people is that all searches work. Now, that’s whether you run a Boolean search yourself, or whether you’re using a solution that matches for you. All the searches work, they return results. The problem with that is they’re never going to return all of the available results of people that you actually want to return. I can’t stress how important it is to say that and to understand that, and it will come into play when I go into the demonstrations. But it’s important for you to understand that every search or match that you use manual or automated doesn’t matter what the solution, you’re not only including some of the people you want to include. You’re also excluding some, there’s really no way around that. And if we want to get into discussion during Q&A, at the end, that’s fine. But just know, there’s no perfect solution. Perfection is impossible. If you’re if you’re looking to perform diversity sourcing, there is no way to have a single search be 100% inclusive. And the reason why I say that is sometimes when when I demonstrate some of these techniques, some people will say yeah, but that’s leaving out some people. And it’s important that you know that I know that it’s also important that you know that. So whether you’re recreating your own search, or you’re using a solution, and you click a checkbox, which I’ll show you one, a couple actually, they do that in no way shape, or form or any of those ever perfect or 100% complete, as long as you know that, then you can do something about it and say, Well, maybe I’ll do a number of different things to make it more inclusive, knowing that there’s no way to include all potential people that you’re trying to target from an under representation perspective, asset to throw out a disclaimer that what I’m going to show you, they’re just examples of what you can do. It’s not I’m not telling you to do this, this is up to you, your company, your HR, legal, diversity, there’s a lot of things at play here. I don’t know your company, I don’t know your diversity sourcing goals, I don’t know what you’re capable of doing what your company says you can’t do. So take all of this with a grain of salt. But I am super excited to show you a number of things that if you’re capable of doing these things, you have either some of these solutions now or perhaps in the future, we’ll talk about some free angles as well. That you’re you’re just a little bit more informed about the limitations of these approaches. And I think it’s important to keep that in mind. So just remember, these are just examples. Okay, so the first thing I want to jump into is actually LinkedIn. So I’m going to share my screen. People could just give me confirmation when this works. That would be awesome.

Chris Hoyt 4:48
Yeah, I can see.

Glen Cathey, Randstad 4:50
Okay, so actually, we’re gonna jump to LinkedIn first. I know not everybody has LinkedIn recruiter, and that’s fine. Welcome. Talk about at the end, I’ll have a little closing comments before we open it up to questions. But I have to, I have to talk about LinkedIn for a couple of reasons. Number one, in most countries, they’re the single largest source of people data that you have access to, at quality, professional data. So it’s, you really shouldn’t ignore that. What’s also interesting is that surprisingly, few people know that when you are using LinkedIn recruiter or professional services, you’re already getting gender representative results. So you can search this yourself, just google representative talent search on LinkedIn, this actually came out about two years ago. So anytime that you’re running a search on LinkedIn recruiter, what this means is that whatever the search is, LinkedIn has the date on the back end, so they know what the gender mix is. And it lets say the gender mix for this particular search, which is a software engineer with JavaScript, no particular location, that doesn’t matter. If they know in the back end that about 70% of the software engineers are men and 30% are women. on each page of 25 results, you’re going to get that same representative mix. And I don’t know if anybody wants to comment that Yeah, I already knew this, or Hey, that’s interesting, I didn’t know that. They are working on some other angles towards representative results. But it’s important, when you look through the results to know that before they actually implemented this, based on differences in how men and women write their profiles on LinkedIn, you could literally go pages and not see a single woman’s profile. And so I think it’s important to understand and appreciate the fact that as you scroll through, you’re going to see a representative mix on every page of 25. So you’re getting the best of the representative mix. Hopefully that makes sense to everybody. So again, if you’re already using LinkedIn recruiter automatically, you’re getting represented results, which does help you from a gender perspective. And we can talk more about exactly how that works at the end if you’d like. But predominantly, they’re using AI. First, they have declared demographic information. So they do ask their people that are on the platform for information. If somebody doesn’t provide it, they use artificial intelligence to essentially infer gender. And they have a high confidence interval for male female, they also have people that they’re not sure. And so they’re always going to have a representative mix of those people as well. So if there’s 10% uncertainty, you’re going to get 10% uncertain on each page. So important to understand that. Now, getting into the search of things on LinkedIn, this is something that I’ve talked about. Let’s say back in, see if I can get this tab open real quick.

All right, you guys can see this doc. All right. So about a long time ago, 2012 was the first time that I actually thought about how can I perform a diversity sourcing for LinkedIn talent Connect audience Now, what I’m showing you here will actually work on any search engine that allows you to run relatively large queries. Okay, so all I’ve done is used my one of my best friends, it’s my second best friend and Boolean searches the or, and I’ve just thought of what are all the ways that somebody could mention something? Because the real question to diversity sourcing is, what would uniquely appear only or predominantly on diverse or underrepresented profiles? Or resumes? That is the killer question. That’s what started it all honestly, for me, and it’s a question you all can can ask yourself, and then no one is answered it fully, right? There’s still unexplored territory of what are other good ways of finding people based on terms that would probably either only or predominantly exist on their profiles. So when it comes to gender, you know, one of the things that I thrown out there before was women’s colleges. And I’ve already created a list of all of them. You can look up sororities, you can look up African American sororities and fraternities. You can go to professional organizations. So the list goes on. And you know, believe it or not, you may see the size of some of these searches. And if you’ve ever seen me talk about these before, you might say okay, I knew that you could do that. But I guarantee you, there’s some people on this call that had no idea that you could actually run these types of searches within LinkedIn recruiter. And again, there are some other atss that depending on configuration, you may be able to stuff these searches in as long as you can, they will run as well. And you’re essentially looking to try to return people who are highly likely to be within that underrepresented group that you’re you’re focusing on from a diversity inclusion perspective. And these are things that we can provide to you, you know, after today’s session, like a complete list of all of this typically black colleges and universities, there’s really no limit to the answer to the question that I posed a few minutes ago. All right, which is what would uniquely be present on these persons profiles. This is one that some of you may have come across at some point, I did initially throw this out there, I think in 2012, we’re not there looked at the most common female names in the United States, it’s government data. And I compiled it over multiple decades deduped it, and I came up with 417 names. And I’ll show you some of these at work.

Chris Hoyt 10:19
Yeah, cuz some of these are pretty massive. If you’re looking at the screen, these are some, some incredible,

Glen Cathey, Randstad 10:24
Yes

Chris Hoyt 10:24
Complicated looking maybe even intimidating for those who are not a blackbelt I Believe.

Glen Cathey, Randstad 10:29
Yeah, and they’re they they do look like that. But the reality is, they are very simple, I can show you very quickly how to put them together as well. But just to show you, this is a save search for the top 800 Hispanic names. And I just put them in the last name field, and the last name field was able to handle all 800 terms. So all you have to do at this case is add a few other, you know, whatever else you’re looking for. And again, I don’t know what your goal is. But if you were target, if you were looking at Hispanic population for under representation, this is an easy way to do it. Now, does this include everyone? Absolutely not. I mentioned that before. But it does include many millions of people. When you do this for the female, first name search, keep in mind, we totally recognize that it can never include all women, because it’s impossible, you’d have to have the name of every person that exists on the platform and search for that. But you can’t get over 50 million results just from a single search from a first on a first name basis.

Chris Hoyt 11:30
And Glen the question may go may seem kind of silly. But where were you getting the list of names because I remember way back in the day, when I was doing sourcing and building Boolean strings, nowhere near as complicated as this, I was buying baby book names, like what would name were popular, you know, what names were popular and what year for females?

Glen Cathey, Randstad 11:49
Well in the US I and I know we might have some global folks on here. And different countries have different resources, but I’m using government sources like census, I’ll use a number of different government sources for people that actually are born, you know, your name is registered, and they have the math behind the most common names. And of course, if you say, hey, it’s only the top 100 names per se per decade. That’s not. That’s not including everyone. Again, of course, it’s not no search can. But what we’re trying to do is show you that this is a good way to target up to 60 million women just in the United States on LinkedIn platform. So it is incredibly inclusive, but it’s never going to be 100%. inclusive.

Chris Hoyt 12:30
All right, I want to keep us moving. You’ve got you got another piece that you want to show us.

Glen Cathey, Randstad 12:34
Yes, I do. I don’t know if anyone’s interested in this. But like, how do you actually create these types of strings, if you’re curious, you just go to word you create a list, you can go to find, we can go to replace. And we’re just going to basically say, hey, every line, every new line, we want to replace with this. And again, it’s something we can cover later as well if you need to after the session, but we replace all and we basically turn it into that or statement that you can just copy and paste and put into into LinkedIn.

Chris Hoyt 13:05
See that tip right there is worth the price of admission?

Glen Cathey, Randstad 13:08
Yes, absolutely. I could share it in the comments as well. So moving on. If you’re outside of LinkedIn, there’s some other solutions you may be familiar with, I have no affiliation with seek out other than the fact that I like what they’re doing with regard to diversity. And I want to point out that the things that I just showed you manually that I’ve put together in terms of let’s say, diverse universities, first names, etc. Companies like seek out have actually simply done the same thing and have taken it to another level that a human can’t do with a you know, basic search. So you know, all you have to do is check a box. And they’re inferring things. And you can check multiple tick boxes. And you know, they’re they’re pulling from public information. So it can be really helpful to use these types of searches. But like I you may have even seen there was like one false positive and the first screen that came up 100% Normal. No search is perfect, whether it’s a human created one or an algorithmically created one. But again, what I’m trying to stress is that, hopefully you want to if you’re committed to diversity, you’re going to take proactive action. And it’s better to take action with an imperfect approach than to take no action because you’re worried that the approaches are imperfect.

No, I’m Sorry, because we got time we take just one question. We might be able to squeeze two in here. But first question is like is how accessible is this level of search for everyone is using LinkedIn does it cross for that feature that you showed us? Does it across all levels of the LinkedIn membership and licenses? Can the free folks use it the light folks use it corporate or enterprise doesn’t matter?

That’s a really good question. So I know the answer to some of that, which is you definitely can’t do that with a free version. To the best of my knowledge with LinkedIn recruiter recruiter professional services. Yes. For light. I’m not sure Yep.

Chris Hoyt 15:00
Alright, we’ll test it out. And then the last question that we’ve actually got time for, I want to ask you about a character limitation that’s at least seen within LinkedIn. Are you getting around that somehow? Or?

Glen Cathey, Randstad 15:10
No, I actually don’t know exactly what the limit is. I’ve actually asked the LinkedIn, I don’t have an answer yet. But as I showed you, I was able to stuff 800 names into the last name, field, and it still took it, I’d have to go back and count the characters. If you’re curious, I actually tried to stuff 1000 names in there. And that kind of broke it. So it’s somewhere between the number of characters between 800 and 1000 names, but you can see it’s a relatively large amount. But if I get the information, we’ll definitely share it back with the group to make sure that everybody knows

Chris Hoyt 15:40
Absolutely perfect, Glenn, your time is valuable. We really appreciate that you gave it to us, along with some blackbelt wisdom today. So thank you so much.

Glen Cathey, Randstad 15:49
You’re very welcome. Thanks for the time.

All right. I want to remind everybody that next week, November 5, we’re going to sit down with Sarah smart, she is the VP of global recruitment at Hilton. And she’s going to talk to us about the hardest thing she’s ever had to do as a leader and that is mass furloughs during done recently, during the pandemic, what her team went through lessons learned. And really, I think a little bit emotionally how it impacted her both as a leader and an individual. toughest thing she says she’s ever had to do. So I’m really looking forward to that conversation. I hope everybody will come back and join us. Until then, we hope we’ll see everyone online@www.cxr.works/talenttalks and we’re going to drop some resources and some of the stuff in there from Glenn and Glenn is going to be in there also. So if you’ve got some questions and you want to keep picking it that big, beautiful brain of his he’s gonna be in there how to answer those questions for you, Glenn. Thanks again.

Oh, you’re very welcome. Thanks for having me.

Announcer 16:40
Thanks for listening to the CXR channel please subscribe to CXR on your favorite podcast resource and leave us a review while you’re at it. Learn more about CXR at our website CXR.works facebook.com and twitter.com /careerxroads and on Instagram @careerxroads. We’ll catch you next time

Tagged as: Sourcing.