Breaking Free from the Talent Abundance Trap

Ah, hey everybody, how you doing? You are live on Shelly's Alley. It is the last episode of January, the last day of January, January 31st, and here we are ready for another month and ready for a new president here in the United States. For those of you that tune in from, from overseas, we had a change in president. The middle of the month. Well, not in the middle of the month, but you know what I mean. So lots of things going on, lots of stuff going on in the news. And, um, today I have a special guest, Ravi Tandra. Ravi is the CEO and founder of a company called Provenbase, which means it's a proven database. Is that, is that kind of what it means? Proven? Pretty much, pretty much, yeah. What is ProvenBase? Well, ProvenBase is actually a sourcing platform that has some really interesting refinements, filters, and other kind of AI-assistive technology that I frankly have never seen before. So Robbie reached out to me and wanted to write a white paper, and we wrote, uh, we co-wrote a white paper on the talent abundance trap. So for those of you that are wondering What in the heck is that? Um, the Talent Abundance Trap basically is a theory that I came up with that explains why we just have this overwhelming inflow of applicants. I know how it's happening, but the technology that's making it possible isn't really the root cause. The root cause is something else. So what's happening is recruiters are being overwhelmed with too many choices of applicants. And it actually makes it really challenging to identify talent. And so, they spend a lot of time going through resumes. They spend a lot of time in there, you know, trying to filter things out. And so, basically, really good candidates kind of just get drowned out. This happened to me when I was looking for a job and applying for jobs last year. Where I knew that I was a good fit for the job. I knew that my name would probably ring a bell with some of the people that I applied because I'm applying to jobs with companies that have recruiters that know who I am. So, it was like, well, how am I not even getting noticed? And I reached out to some folks and they were like, oh yeah, your application's in here. But we already closed the rec. We already filled the rec. It's already— and it's like, oh, I get what's going on. So, that's my theory is that we have too many choices, too many applicants. And it's actually making it harder for recruiters to do their job and do their job well. So, you know, people used to come to me and say, 'Shelly, sourcing is dead. Sourcing doesn't exist anymore, blah, blah, blah.' Well, let me tell you, I beg to differ because there's like 1.5 million people that do this job and it's more than ever before, partly because we have all these different roadblocks in the way. So, that's kind of what the white paper addresses is the reality around what is really going on with the industry and what we need to do in order to change what's going on. So, uh, Robbie has this, this organization that has a solution that I was looking at. I was evaluating it and actually purchased it. And like a year went by, and then I was like, oh, this is really interesting. And he called me and we had some meetings, and then we started writing this white paper. And his technology that's in this application has some interesting answers to this. But I want to hear from you, Robby. What— where do you think this whole thing came from, this talent abundance illusion? Like, I know we have lots of candidates, but companies still have jobs that go unfilled. They still have jobs that go unfilled for months, even years, positions that are posted for a long time. So, you know, clearly, there's something going on here. Where do you think this is originating?

Firstly, thank you, Shelley, for having me. Hi everyone. You know, the problem is with DEI. That's where the problem is. That's where the root cause is. In a light note here, Shelley.

So that's why we got rid of it.

That's the solution, by the way.

Appropriate timing, right?

Yeah, on a serious note, it is the overwhelming influx of applications that recruiters have received for these job openings, right? That is the biggest challenge.

It's— I saw it, you know, even 3 years ago before the bots started kicking in where, you know, my recruiters that I was, you know, overseeing this strategy of this unicorn, this $2 billion startup. And the recruiters were like, I can't get out of these recs. There's just 200, 300, 400 applicants within 24 hours. And even back then, I knew it was a problem.

Now this is actually, you know, I was talking to a talent leader recently that she said that they received over 12,000 applications within just 4 days for a data scientist role.

Whoa, and this is a unique skill set. I mean, data science is not a common skill set.

Exactly. Now you may think this is one of the hot jobs in the market. Even recent LinkedIn study said that, right? AI engineers, data scientists is one of the top jobs. But even the, the interesting thing here is even with the automated filters they have in place, such as rejecting the visa required candidates or applicants or candidates from foreign countries, recruiters still face a massive workload. I want to put it out there, the tough recruiters and sourcers have the toughest job and they are the backbone of the economies. If anyone has any doubts on that, I want to put it out there, uh, because sourcing or recruiting is not going to die, they only thrive. So, uh, the issue with this is they often need to dig deeper into each profile, right? Just, it's not that not only you're getting these many applications, but whenever, when they're going through these profiles digging into each profile, verify the details, and conducting additional screening. So that itself becomes a daunting task at scale. So some companies are even considering to hire additional staff just to screen resumes, so which, which adds cost and complexity rather than solving the root cause. So now the shift from this talent scarcity to data overload has created like a, a new problem To me, you know, as a paper suggests, sourcing is no longer ability to find the unfindable talent or about finding enough candidates, but about cutting through the noise to find the right ones efficiently. So what do you think, Shelly?

Cutting through the— well, I, yeah, I think that the problem definitely has shifted. It used to be candidates were hard to find, and then we had this influx of tools and technology that came in and helped us find people. And For a while there, folks were saying, "Oh, it's really easy to find people now. We don't have a problem finding people." Well, that never really went away. It's still difficult to find people, but there's just so much information that the challenge now is getting to the information that we need. It used to be we had a hard time finding information that we needed. Now we have too much of it and we still have a hard time finding information. There's a reason why data science is a job. Library science is a job. Research is a job because sure, you may have all of the knowledge in the world inside of a database, but that doesn't mean that you can interpret that knowledge or you can take action on that knowledge. So it is definitely true that our technology hasn't really provided an answer yet. But what I'm wondering is where the problem really comes from. I don't think it's because some people say, oh, there's AI chatbots now that make it possible for candidates to apply in volume, in bulk. You said 1,500 or whatever that was, number— I can't even— 12,000. Wow. Yeah, like, I'm just like, what? What? I mean, it's incredible. So clearly, you know, there's some technology out there that's allowing people to do that, but the technology is solving a need. I'm curious as to where the need comes from. Why do we have candidates applying for you know, thousands of jobs instantly with an app? What's, what's driving that need?

So one of the, one of the things I think is the, the automation is coming so much, even the job application system, there are robots that are kind of applying the jobs, right? That's one. Two is, um, there is definitely some awareness in the market right now with the recent, uh, uh, you know, turbulence in the market. That your company is laying off and whatnot. That's there for sure. Um, I think it's, it's also the people, the shift from being remote jobs to hybrid jobs, now in person. It also comes from, to me, what I really, what I'm seeing is depending on the way job is, and this particular job, the example what I'm sharing is a remote job offer. So when it is a remote job, there's a lot of demand for sure.

Okay, in person So there's fewer remote jobs compared to before because of the shift in the pandemic. You're saying is we had this peak in remote hiring and then that sort of slowed down. And so people really got to, they want the remote jobs.

Right.

So when they see that, so what you're saying could be proven or disproven. I don't know if that's something we wanna do right now, but it could be proven or disproven by comparing applicants that are for onsite jobs versus applicants that are for remote jobs. And would there be a data science job for with 12,000 applicants if it was not remote. Um, we could investigate that. We did not for the white paper, so that's a really good— we should have brought that before. But still, my question, it comes back to, um, why the need to, to have a robot? Like, so let me, let me flip the table, right? There was never a need or there never has been really a recruiting process where they just invited every candidate to apply even if they weren't qualified. But the candidates are doing that to jobs. You understand what I'm saying? Like recruiters never like sent an email to every single candidate in the world. Well, some did. The ones that did, they're probably not employed anymore. But the point I'm making is that recruiters don't just blindly indiscriminately just email everybody and say, hey, 'Apply for this job.' So why are candidates doing that? Why are they blindly and indiscriminately— maybe 12,000 people applied for that job and they weren't all data scientists. Again, we can prove that, but I just don't think that there are 12,000 qualified people within 24 hours. I bet you in that pile, based on my own personal experience looking at the statistics, looking at the data as a consultant and seeing the rejection rate or what we like to call in the industry, resume review stage. So when we go from resume review to move them forward, that means we're reading the resume. The percentage of people that are being passed that gate has decreased. So these are not qualified candidates.

Exactly. I think that is, that is one of the major problems here, right? So one, candidates right now are also, on the other hand, if you think from their side, they're applying for jobs and not, they're not hearing anything. They're like, you know what, let me apply for whatever it comes, more jobs as possible. Let's be 'Let's go.' 'This isn't working, I want to do more of it.' 'Yeah, I want to do more of it.' And they're using system. There are two problems here. That is from the candidate side. And the other side, the system side, you don't have the proper filtering mechanism to really weed out these candidates. It's just kind of passing through all these candidates. In the end, that becomes a burden to the recruiter to really go through it. And I heard, Shelly, over the surveys that we are doing with a lot of recruiters and sourcers, they said over 70% anywhere from 70 to 95% of the time, the candidates that apply for the jobs are not qualified. So with which they becomes, you know, they have to go out there and source candidates on themselves, right?

They have to source anyway because all the app— not all the applicants, but 12,000 people applied, 1% of them are qualified. But the problem is, to get to those 1%, you have to go through 12,000. It's just easier to go out and find people than it is to go through 12,000 candidates.

Right. It's, it's, it's, it's absolutely— that's what, that's where we are going. That's exactly the solution for that is one is having an efficient system to filter through the candidates that are incoming, and also how you— the recruiter itself has to go out there and really find the candidates that could be a better fit for them. So these are the two approaches that could—

yeah, right. Gartner found that 56% of recruiters in a survey that they did, 56% of recruiters recruiters that they surveyed spend more than 1/3 of their time sorting through profiles in the, in the application queue. So, like, more than 1/3 of it. If your job is to recruit people and a third of your time is spent just sifting through profiles, you're not— that's a third of your time you're not recruiting. Now, the other third of your time, another third of your time is probably taken up in recruiting, and then another third is meetings and planning and tracking all that stuff. So, you can see how recruiters just have a lot a lot less time to really do the actual job, right? The actual job, right? And so, they don't have time to respond to 12,000 applicants. So, the applicants should apply for more jobs.

Absolutely. Yeah. Which is a crucial part, like you suggested in the white paper, is the engagement, the personalization. You're taking away all that, which is the most important in the process.

So, the biggest challenge that has emerged from this overload. So we had, we had scarcity of talent and it was because there weren't enough candidates, or, and that's still the case, but we have a different problem now, or because we couldn't find them, et cetera. Now we have this data overload. So we went from not enough candidates to like overload of data. And what you're saying is that the biggest challenge that has emerged from that is that we still don't have adequate filtering in most of our tools so that it would be great if we, if we got a million applicants and we can immediately get to the right people. That would not be a problem. The problem is the more applicants we have, the less we can get to the right people. That's what you're saying?

Uh, yep, the, that, that's, that's pretty much it. And, you know, I think what I think is also important here is especially with the leveraging AI and automation can definitely help. And then on top of it, one thing I thing that can be pretty effective is within the job post itself, having some questions embedded like in the middle of the job post or at the bottom, almost at the bottom of the job post, you know, like what is your favorite color? So now the systems should be able to categorize people that answers these questions versus not. So now you know which one to go first.

Oh, you don't make it a requirement.

Yes, so someone is actually reading a requirement before applying, right?

Right.

So the other thing is careful crafting, like we suggested in the white paper, Understanding the needs of the job and the company culture, skills, years of experience in niche areas and other things, right, that are key, which can tremendously help to narrow down once you have that understanding as you're going through the list. And the other thing is human insight is definitely important here. Don't just automate everything. Oftentimes people confuse with AI and automation, and there's a clear distinction between them. So I was— this is an interesting example. I was talking to one of the talent leaders.

Do you actually think that people are automating everything? I mean, I've Oh yeah, this is a real example, Shelly.

So I was talking to one of the talent leaders from a major corporation. I don't want to give the names now, but he told me that he really liked AI. And this, this gets interesting, right? That can shortlist the candidates, create projects, send messages, and schedule the calls. Okay, you want— so now he wants a robot to do everything, right? So the thing is, here he's talking about AI. He wants AI to do all that. The issue here is, or the flags I say, first, the lack of understanding, clear understanding between the AI and automation, one. And then the lack of understanding of compliance. So that is completely taking the human inside away. It can be a missed opportunity, right? Like, and a bad candidate experience because there is no personalization, which you write, which you wrote in the white paper, which is crucial. And there is no genuine engagement there, and that is very important, uh, and, you know, part of the, you know, important part of the job. So to me, these are some of the really important things, uh, to be able to overcome this data overload, to be able to sift through the candidates in the most effective manner.

Okay.

And be careful and mindful about some of these things, you know, having that clear understanding between AI and automation, and also going through the going through the filters that are necessary within the— within like whether it may be the ATS systems that may be using that are passing through the candidates, having those filter criteria to be able to only send the candidates that really make sense is one thing. The second thing is, as you're searching the candidates, if you're getting given like 2,000 results for the search criteria that you're looking at, but then having the ability within the tools that can really Narrow down to expand, like narrow search, like, you know, you really want the bullseye, or if you don't have enough candidates, have the ability to expand that so that you're not leaving anyone behind that could be a good match or a potential match, right?

From within that, from within that, right?

Having the ability so that now you're giving the professionals, the talent professionals, the opportunity for them to really— how you want to use this, right? And one of the things which you really liked, or we talked about, is kill skills used within a time frame. You know, not every job you need someone to do 10 years of using a particular skill, right? Maybe you want someone to use 3 years of using a particular skill. So having that level of flexibility within the tools that are using to really get to that needle in the haystack at a faster rate can also help you to cut through this noise. So I feel like there are a number of opportunities here as you are to kind of narrow it down to it. Like in the precise manner. I think the big thing is AI obviously is one thing, the filtering criteria, how precise you can go about it and how flexible they are. So these can really help tremendously apart from the other things we discussed.

So yeah, so far I wanna point something out. There was a comment by Natalia Romanova that she says companies have made it super easy for applicants to apply. So this could be contributing to it. So in other words, maybe, you know, there used to be, and I remember this, you have to fill out 45 minutes of, uh, you know, information in order to apply for a job. Now they're making it really easy. So there's all this like, oh, we got to make it easy for the candidate, easy for the candidate. Now it's too easy for the candidate. So that's a, that's a very, um, that's a keen observation. I think that's a good, a good point that she made there. Um, but then I want to come back to what you were saying, which is we were— you're offering two 2 possibilities, 2 solutions. One is to source the candidates, which has always been my passion. And I still think, you know, today more than ever because of this apparent, you know, overwhelm of candidates, I think it's even more important that we source and find the right people because it's going to become harder and harder as the technology makes it easier and easier for people to apply, right? And even if technology helps match candidates and everything, there's still going to be a need to source. But you're saying the second side of that is to modernize our ATSs to create the ability to filter better. And one of the things that you— I just want to point out that, that you put into your platform that is very different is targeting skills within a time period. So in other words, someone that has used this skill according to their resume or their profile in the last 2 jobs. And that's important because if you just do a keyword search, you'll find that keyword might be present on the resume anywhere in the resume. It could be in the summary, it could be 10 years ago. So the ability to pinpoint that and tie it to a year, as in freshness, right? And, and if you're thinking, oh, you know, that's going to discriminate people with experience, I understand the concern, but this isn't saying I don't want— can't— I don't want to find candidates that have more than 10 years of experience. What you're saying is if a technology like, say, ChatGPT is only 2 years old, I want someone who's been using it for at least a year recently, because if they haven't been using it for a while, it's probably changed so much that they're probably not very good at it since they only used it, you know, when it first started, right? So you're giving us one, one more angle. What's another similar angle like that that we can use that is not in most ATSs that we can use to filter out candidates in a kind of a smarter way?

Yeah, so the, the one is the narrow search to the wide search, or let's say that is one thing that comes to mind, right? How you can narrow down and wide searches or the potential searches so that you have the ability controlling the optimization, the unprecedented control that we're giving to optimize the search. The other thing is putting a cap on For example, education, right? Sometimes you may be looking for candidates, you know, I'm okay with having someone with bachelor's, that may be the cap because you want to really tap into skills are more important for you than the degrees. So having that level of flexibility can also help to put a cap on degrees or degrees in progress along with the optimization And then the other thing I think could be pretty important is for you to have the ability, for example, I was talking to one of the healthcare customers recently. So they want to find candidates, like I want to find, they are hiring in South Dakota and they wanted to see in other states where they could potentially pull the candidates from if they don't have, I mean, they have shortage of candidates in this particular case, right? Okay, they want to know, hey, someone that went to school in South Dakota but worked in Minnesota, working in Minnesota currently, right?

You want to filter for that, right? So, so there's no way to filter people who— and this is a really good technique that, you know, a lot of sourcers use, is to find people who formerly lived in the location for, for local jobs. So in other words, um, like, I'll give you an example from, from, uh, from my own personal experience. There was a position here in Atlanta with— you might be familiar with the Cartoon Network.

Yep.

It's part of a, part of a bigger company, but they were trying to hire somebody to develop apps for the App Store. This was a long time ago. I mean, there was not a lot of people that had experience with monetizing apps yet because it was early on before— a lot of apps used to be free or whatever. There was not a whole lot of monetization at this time. It was early stage. And they wanted to hire somebody here to work in Atlanta because this is where they were based. Unfortunately, most of the people that had this experience were in California and New York, other cities, and they were having a hard time getting people to move from California to Atlanta, et cetera. Not everybody wants to relocate, even, even back then, right? Even now, I would say even worse. So when I went in there, the solution that, that, that I implemented, which worked pretty quickly, was we started looking for people in California who went to school in Georgia.

And I think that if you think about the traditional way, a lot more open to relocation, which Exactly, they did. Yeah, right. And if you think about the traditional filtering criteria, that goes to another topic in the white paper, is if you go into the traditional filtering criteria, this particular way of searching is not possible. So now having that level of flexibility to be able to get into that can be pretty helpful. And which is— I appreciate the wisdom of the recruiters and sources looking at ways to find candidates in this way. It's amazing to me, right? That also gives us opportunity to see what else we can do. How can we make the systems better so that it's actually solving the problem instead of adding additional, you know, headache to the end user, right? So that is one thing that comes to my mind. I mean, like you said, Shelly, you know, I just want to add one other point is the strategic pipeline building is so key with this because one, yeah, you're getting this plethora of candidates for the job posts that you have out there, but imagine that This is a secret weapon, right? The effective weapon that you have, the sourcing. Like, you know, as the white paper suggests, that according to LinkedIn Global Recruiting, uh, uh, the transport companies with the robust pipelines are 33% more likely to fill the roles, uh, faster than those relying on ad hoc recruiting.

Right. And to be clear, in this case, what they mean by a pipeline is we're talking about like candidates that are interested in your company that you've, you know, reached out to, you've contacted. So a pipeline doesn't mean a list that you've never contacted, right?

Right.

So having the pipeline is one thing. Finding the right people using filters, like you said, who used to live here but went there, um, you know, have they used this job in the last couple of years? Those kinds of things are, are key in helping us filter. Just Paul, um, is, uh, got cut off here, but he was It was asking if part of the problem is that it's just, um, the, you know, job portals like LinkedIn and whatever are really not supporting the process. And I think that's actually a pretty insightful comment. I'm interpreting that to say, you know, a site like LinkedIn, they want more activity, they want more people, so they make it super easy for people to apply for jobs. They give you lots of options. And so that's actually increasing the problem because it looks good for LinkedIn to have lots of candidates, but that doesn't necessarily translate into hires. So they're looking at the wrong thing. They're looking at how can we increase the volume, but they didn't really focus on quality. And I think that's actually a pretty good, pretty good.

That's a very good one, uh, for just Paul, uh, commented on. I actually, uh, encountered one of the, uh, from one of the talent leaders, the same thing they said is, you know what, when we are Searching for candidates, oftentimes we are getting not the best results because one is flooded with the data, but also the— this also goes to the search accuracy, right? I'm searching for apples, but if you're giving me oranges, it's not going to work out for me. But how effectively you're going to have a lot of oranges? Yeah, so it's like, you know, I want to look for a psychiatrist and, uh, and a medical, medical director, but if you're giving me totally like irrelevant people that could be it's not going to cut out, right, if you're going, going to give me a relevant result. So I'm looking for basically having candidates with these two, two, two, two skills or two jobs, job titles. So simple examples like that, even the systems that are sort of, uh, so popular are not providing these results, which is kind of interesting. That, that kind of, you know, echoes to what, uh, Jaspal is commenting here. So again, going back to having the systems that have these flexibilities, able to, you know, to really get you what you want, gives you that additional flexibility, is super important as you're evaluating, or, you know, whichever the systems you're using.

Hmm. Okay. It occurs to me that we've been talking about this white paper, but we haven't told people where to get it. So I'm going to put up a QR code here. Oh, that's interesting. Let me, let me Um, try that again. There you go. So there's a, there's a download link somewhere over— I'm trying to point to it. There we go. There's a download link right there if you want to scan that or take a picture of it and, um, get the white paper. I just realized we have— we're talking about this white paper and we actually haven't told people how to get it. Um, if you go to theprovenbase.com website, you'll find it there as well. And that is just provenbase.com.

I'm going to share the link.

There is Provenbase right there. And then you can, um, you can see the— there'll be a resources page and the white paper will be there, or you can scan the QR code. Sorry about that, I just realized, um, we've been talking and talking about it and like White paper, white paper. And people are like, what are you talking about? So we talked about the outdated sort of fragmented systems that are not allowing us to filter. We're talking about an interesting revelation, which is that maybe it's in the best interest of some of these job boards to flood us with resumes because it makes them look good. And that's always been, I mean, since the days of Monster and Online Career Center, that's always been the case. But there's still, you know, there's still jobs with no applicants. I mean, even if, even if we have positions where there's a lot of candidates applying, there are still positions where very few candidates apply. And that's a different problem. So we have both problems. We have too many applicants that are unqualified. It's overwhelming. And we also have the opposite of that, like zero applicants, you know, reqs that get posted and go with zero applicants for months. And that's also a paradox of the noise. I think, you know, if you're looking for a job, I went out there recently. I was— I used to, I used to like, you know, go more frequently, but just recently I went out there to look for sourcing jobs and there were like thousands of them. And I'm like, wow, where do I even— how do I even filter this down? Like, it's, it's— there's too many. So even as a job seeker, I'm overwhelmed with too many jobs. And I can see how some of those jobs that are way down in the bottom of the search results, because of whatever reason, they're not getting any clicks. The ones that come up in the top of the search results, they get all the clicks. And the ones that, you know, page 3, right, of results, they never get any applicants. So we have that problem as well. Maybe it's too many choices. But we also have another problem with too many choices, which is all of these tools out there that all of them— no offense to you, I mean, I know you're the founder of a of a sourcing platform.

It's true.

But there's so many of them out there. They're all like, we're the best, we can solve all your problems, one click, AI finds everything, we match all your candidates, we have 18 billion candidates, and we can find— you know, it's like always the same message. And it's just everybody is offering the same flavor of toothpaste, you know what I mean?

Yep.

So it's too many choices. How can we make smarter decisions as buyers of this recruiting technology or evaluators or recommenders of this technology, how can we make smarter decisions without getting lost in all of this overabundance of hype, right, of this technology?

That's a great question.

See through the clutter here.

Oh, great, great, great, great question there, Shelly. I think first thing is education that comes to my mind because Education to the sourcers and recruiters that are doing the job about first them in about their company culture, tech stack, hiring goals, and about roles, compliance policies, and then staying on the top of the market innovation, right? These are some of the important things about the, from the, from, from the company side of things. But then, uh, learning and getting trained on what matters the most because there's a lot of, like you said, tools and technologies that is innovation is happening, but what matters to you the most, kind of staying on top of it is crucial. So this will give, you know, y'all, the recruiters to be dangerous enough to exactly find what you're looking for. So, and then, you know, when you are really being reached out or looking for tools or technologies, point out what you want and ask those exact questions of pain points. And see what makes sense, right? And choose. And don't just take their word, because that is what happens. You know, you take their word, you see a lot of noise out there. There are a lot of these big players out there, the market noise, which we discussed about it.

Yeah, yeah, yeah, I get what you're saying, but you know, I'm going to challenge you a little bit here because, um, I know you didn't say this, but it kind of, it kind of sounds like recruiters know what they're doing, and I don't think that's the case. And I know that's not what you said, But like, it's not that we can't, it's not that we can't evaluate the technology. It's that everybody has the same flavor.

Correct. Everyone.

But yet, let me, let me finish. Everybody has the same flavor, but yet all of those are the flavors that we don't want. It's like they're, they all have this. They all do the same. They all do the same thing. They all use ChatGPT. Now everybody's like, oh, now with ChatGPT, they all have, you know, matching. They all have this, they all have that. But the things that, like you were talking about past 2 years history, the, um, I don't know if anybody caught that, but you were talking about the maximum education. That's brilliant because sometimes you don't want candidates that have master's or PhDs to apply because that they're, it's, it's, they're not going to be in, they're not in a good position for that job. Right? So if you, if your requirement is a bachelor's and you have, you're sourcing people that have bachelor's but also master's and PhD, that's not a good fit. So that kind of thing, the ability to determine whether or not they have more than a bachelor's, doesn't exist in any of these platforms, right?

I think, I think, yeah, what I'm saying, thanks for that, pointing that out there, Shelly. So what I'm saying is Having this understanding, I think, helps to really figure out what really is important to us, right? When you have presented these number of tools, you know, we all know that there is not really like something— everything is already done at some way or the other. We are only using what is existing out here, repackaging it and making more sensible and useful to the users right now. That's what is happening. The more innovation that we are seeing, making it easy, how we can make it more meaningful. All the data that we have available. And within that, how can we— how can you choose the best out of that, right? You have a lot of this noise. That was my question.

I don't know. That's— I evaluate every tool. I mean, you know, people reach out to me with this new tool and I, and I'm like, all right, I'll take a meeting. It has to be— let me put it this way— they have to be absolutely awful in their messages for me not to evaluate because I'm always looking for the next thing. But I will tell you this, I am the decision maker and I am the influencer in many cases. However, I also don't suffer fools. I don't have the patience for a demo that is a bunch of PowerPoint slides where they literally start talking about the same thing. So, I take all the meetings and I do all the demos. But after the first 5 minutes, if they're not going anywhere, I'm cutting them off. Not because I'm being rude, but because they're not showing me anything new. And that's the problem is How do— from their email, I can't tell that they have anything new because they're all promising that they have something new, but they all don't.

Unfortunately, what you're doing, you have to go through that. There are only two things. One, they have to trust someone like at your capacity who did this work and have unbiased recommendations. I'm just saying unbiased.

Oh, and don't call me.

And then two, you try yourself on for— I mean, having one of the things I'm saying is having the level of understanding. Like you said, I want to cut it off if I just saying this or that. So you have some level of— you have your own filtering criteria, what I want to listen, what I don't want to listen. And that comes with your own learning and experiences, education, whatever the number of years. That's all I'm saying. That is what I meant to say, education and getting yourself up to speed to be really evaluate what matters the most for me, what Do I have time to entertain this or not?

And if you don't, then, then just go to the—

yeah, go to what matters the most and give it a try. And don't take the word for it. Like, like, I think it's a—

it's an important point that you're making here, which is pick a couple of things that are really valuable to you. So for example, if you're in a business where maximum education is really not a thing, then don't look for a technology that allows you to filter by maximum allowable education. Because that's not important. If you are in an organization that has a big priority to hire people with disabilities or veterans, then look for technology that has that. For example, Provenbase has built-in technology for people that are hearing impaired. It's got built-in and free, by the way, translation transcription or whatever that's called. So that's if you're not hiring that for that people, it's not, it's not a criteria that you then, then Ignore that, but you have to pick something, one or two things that are really important in your workflow because if you just go out there and start shopping around, they all sound the same.

Exactly. I think that is one way to filter out the clutter and then get to what you want, but you have to go through the pain one way or the other. That is there, but you can minimize it at least.

Either go through the pain yourself or go through the pain that somebody else— Somebody else. Somebody else's pain. Great. You know what? Use my pain all you want. I'm here. Yeah, I will. I will take the calls. I am not unbiased. I have a bias, but my bias is very much biased towards precision and accuracy and action, which has always been the case for me. Um, you know, people, uh, people used to think, oh, Shelly's all about the technology, he's a technologist. Um, actually, no, I think I'm really technology agnostic. What I like is I like to see it actually working. So if I get in there and you literally give me the keys, I want to test drive it. Remember when we first met? I'm like, just give me a 24-hour pass. And you're like, I don't know. And I'm like, yeah, just, you know, I, I'm gonna get more out of that than I will out of a demo. And you're like, okay, fine. And you gave me the pass and I went in there and I kicked it around. I was like, oh cool, there's all these things that I can do. So, you know, that's what I do, but not everybody has the patience for that. So feel free to take advantage of my already having invested the time. Just know that my bias will be, is this precise? Is this, you know, efficient? Um, I'm not looking for, you know, something that's going to be the Holy Grail here. I'm just looking for, can it do certain things well? And you know me, you know, after we've, we've gotten to know each other for, for a while now, and you know that I use 5 or 6 different tools.

Yeah, absolutely. That is right.

Because I think some of them do things better than others, period. It's—

yeah, yeah, there is no one-size-fits-for-all There is no one solution that does everything, right? So I think the key here is one of the things I also think it's important is being open. Like you said, you're technology agnostic, tool agnostic, or, you know, being that type of open mind can also help, help you. When I say you, means the users, uh, to find the better solutions. And, you know, if you're thinking of, I want to do the things the same way that I'm used to doing the things, that is one of the adaptability issues I have seen. If you're trying to find something better that you want to do better—

and you wanted to do the same things you already know how to do, that's a good point—

and you have to be open enough to, to really break the system or really get out of it. Or maybe it works, it doesn't work, but at least be open enough without any biases. Then, you know, try it out. Then I think that may open new doors, and, you know, you become irreplaceable at that point because you're bringing something new that, that is more valuable to the company end of the That's interesting.

So you want something that's different and better, but you're evaluating it based on all the things that you already used to do before. So you're not willing to, to change how you do it in order to use the solution that has the new offering. Maybe that's why everybody goes back to the same, you know, tired and, and same thing, and it's just comfortable, proven, or they don't want to.

I mean, one of the things is obviously if you have a name brand and if it doesn't work out You don't have to— I think it's also not— they don't want to be taking the blame. Hey, it's not working out because I referred it, right? If it is something like already named brand, then, you know, everyone knows it. No, if it doesn't work out, no one— I don't have to take the blame for it, right? That's also another challenge. I think that is a big challenge. I think the talent professionals or the users have when they're opting for something new, right? So that is something you have to also The management has to be open to it too, as well. Hey, you know what, we're trying it out. If it works, that's great, but if it doesn't, that's okay. We can't just blame, you know, even the greatest systems that has all this marketing noise may not work or may don't work like we are hearing right now.

Yeah, okay, that's good advice. Um, another question I wanted to ask you is, um, in the white paper, we, we, we talk about a shift from a fundamental shift from transactional hiring. So like, you know, fill recs, which is, which is how a lot of recs get filled. But we're talking about the need for moving away from transactional hiring to a more value-driven approach where recruitment is actually adding value. Now, search firms have always added value. That's why they demand the higher prices. But once we go into these high-volume recruiting roles, or once, you know, we, we start doing the transactional hiring, that value goes away. So, what does it mean to shift from transactional to high-value recruitment? How can we— what can we tell people to look for in order to make that transition? You know, like you said, they're used to doing it the same way. They're used to doing the same thing. That reminds me of that, you know, I've always done it this way. So, I'm going to look for more tools that allow me to do it this way. But you're still going back to transactional. So what is the, what advice do you give them?

This is my favorite question because you have, you provide your personal experience there in the white paper. I love it. And I want you to share if you don't mind, but I mean, companies should look for what impact that hire can make, right? And the cost of that unfilled position rather than just being transactional. So it's all about the bottom line.

That shift in the mindset is cost not to hire or the cost of leaving it open. Those kinds of things, right?

Yeah, the shift is— the shift in mindset is crucial to have a thriving company. Um, the, the, the example that you shared in the white paper about the cost that one of the top 20 retailer companies, like, the cost of not hiring is— was $13 million per month.

Imagine, and $13 million.

Yeah, $13 million. If you don't mind sharing that example, Shelly, that itself answers that question. I really love that example. That itself, that kind of sums it up, basically.

Well, yeah, so I can't give you specifics. Obviously, it's sort of redacted a little bit in the white paper. But the example is this. There was a hiring manager leader, a head of a department that I spoke with as part of my diagnostic that I was conducting with this organization. To try to get to the root cause of why some of these jobs were going unfilled and how they can improve. This is all about process improvement and optimization and making their TA function better, which is what I do. And I asked the question of— I made the mistake of asking the question, do you know how much it's costing you not to have this position filled? And the reason I'm saying I made that mistake is because the way I asked it, it sounded like I was saying 'Can you count to 3?' You know, this person was, uh, it was basically a quantitative analyst, a statistician, pretty much. So, you know, of course the answer was, 'Yes, I do. Exactly. Like, I know exactly.' And I was like, 'Okay, um, so, you know, can you share with me?' And he said, '13 million.' And I went, 'Wow, it's cost— it's cost you $13 million not to fill this job?' And he goes, 'Oh no, no, no, no, no, it cost me $13 million per month and had been open for more than a year. So I was just like, I mean, I was— at that point, I mean, I tried not to show my shock, um, but I started digging up other questions like, how come, you know, have you told anybody? I mean, do people— why is this, you know, why has this been open for that long? I mean, it's an incredible amount of money. And the reasons why the position went unfilled were mostly a process issue. It wasn't because the candidates didn't exist. They were out there. It was just a process issue. It was too many applicants, not the right candidates, not the right applicants. Um, you know, it wasn't prioritized correctly because it wasn't given the— one of the things that, that I advise companies about is prioritization of requisitions, right? Not every requisition is the most urgent, and the absolutely worst way to prioritize a requisition is FIFO. FIFO is an accounting term for first in, first out, right? If you work on the newest requisition that came in and your sense of urgency is which one is the oldest, like time to fill, then you're doing it the wrong way. Time to fill is an absolutely ridiculous metric. It means absolutely nothing because how long it takes you to fill a job means nothing to the hiring manager. They want to fill it with the right person. And in this case, increasing the time to fill from a year and a half to a year and 3 months would not make much difference. So time to fill was not useful. They weren't looking— they were looking at once, once a rec got beyond the average time to fill, they were like, oh, that's an old rec. It's, you know, we're gonna work on the new ones. Well, it's, it just, it was, it was, it was an issue related to workflow at this organization. So yeah, thanks for bringing up painful memories.

No, that's, that's a beautiful example and why this is so essential, right? The sourcing itself and the impact, because this question is so, so important. Transactional to a value-driven, that's exactly what we're talking about. So understanding the cost of unfilling instead of time to unfill, what is the cost? That can change and what value it can bring to the company and organization, right? So I think that's what needs to happen in every position. I think then companies are going to thrive even more.

Yeah. Here's another, here's another bug that I want to pick on here. Um, scaled personalization outreach efforts that are, I mean, for, for the longest time, ever since I started doing this, almost 30 years ago. I've always crafted well-thought-through, well-researched messages, good content that was impactful. I always did. I still do today A/B testing. You know, I do A/B testing on the subject. I do A/B testing on the body, just always constantly trying to increase the response rate of my outreach messages. It doesn't matter what platform, text, email, phone. It's all about improvement, right? But with the volume that we need to do now, um, how can recruiters scale personalization if they need to send 100 messages a day? They can't— they can write well-crafted messages because it might take you hours to write a really good message, but then personalizing it for each candidate, if you have to do that with 100 people, um, that takes you know, that takes a lot of time. I mean, if you have to do 100 personalizations and it takes you, let's say, I don't know, 5 minutes, right? Uh, that may be a little bit too much to personalize it, but I mean, you got to read a profile and kind of, you know, so maybe 5 minutes is not— it's not ridiculous, right? So we're talking about, um, what, like 500 minutes? What is 500 minutes? That's like almost a whole workday just personalizing 100 messages. So it's impossible. How can recruiters scale personalization in this day and age? And I will tell you this. I have had— I hear all the time from customers where I'm doing consulting. I hear all the time about their response rates being in the— they're like so proud. They're like, oh, we get a 30% response rate. That's awesome. LinkedIn said, this was last year, I don't know about recently, and by last year I mean 2023, so 2 years ago. LinkedIn had this thing that said if your response rate is 18% or less, then you kind of get like a warning because it's too low. So 18% is really low. I mean, 18% would be, if my team had response rate of 18%, they would be looking for work. Let me just put it that way. That's really, really low. And that's the bar LinkedIn is setting. So when clients say, oh, we get 30, 35%, and they're so proud of it, and I just look at them, go, but that's, that's awful. That's— it took you 8 hours to find 100 people and you're only getting 30 of them to respond. That's just not— that's not a very good, efficient— so the reason I'm telling you this is because I still every day still work on RECs. I work on RECs every day, no matter how much consulting I do and even as a global head, I still worked on RECs because I think it's very important to stay up to date with how things are doing. If I can get the things to work, if I can get a tool to work on an actual REC, then I know that it's something I can recommend to other people basically. And just last week, I had a campaign where I used heavy personalization. Heavy personalization. The message went out to 32 people on LinkedIn, and I know that's not 100 like I was just talking about, but it is 32. However, it didn't take me 5 minutes per. I used a technology that allowed me to automate the personalization, and I think— sorry, did I say 37? Yeah, it was, it was 37. And out of that 37 people, I had 26 responses. So that is a response rate of about 70%. 70%. And to me, that's low. I'd like to get to 80%. That's kind of where I'm— my happy space. That means that if I send it to 100 people, 80 of them reply. I send it to 100 and 70 replied, but it was actually 37. I did use automation for that. Basically through ChatGPT and generative AI, but it was very specifically crafted. And that's not easy to do. I mean, it took me a good chunk of time to just set it all up. How can recruiters do that? I mean, if they're not as technical, if they don't have the time, they don't have the patience, how can they do that? There's nothing out there that does that, is there?

It doesn't for you. I think that's— we have— I can speak to what we have to exactly address this problem. So it's essentially what you did, right? It's an AI-powered messaging system, and you— we have system where it looks at the candidate profiles, at their core skills, and it matches with the job search criteria you put in or the job description. It creates a personalized message to the effect that, hey, I came across so-and-so, I like your skills, I like the companies you work for, we work in the same industry. So kind of making it so personal that candidate actually response, right? So that's—

but see, that's the problem. See, is when, when messages are like, I came across, or whatever, they don't get 70% response rates.

You have 10 different variations of it though. So it's one of the examples. But all I'm saying is making it so that you're going into the profile, actually reading it, to picking the things that they did, so that you're highlighting those things. So not— that's a generic one.

That's not— that's not enough.

So, um, one of the—

If you're saying, I found variable, variable, variable on your resume, that's still not personalized.

One of the things is, you know, you have like 10 different messaging, right? So it's also about the prompting. You have the prompt option too, like I said, having the question. So you need to put a question like, these are the important criteria for me when you're sending a message.

Yeah.

So what I'm saying, you have the ability to put your own customized message within that to ask, to frame how you want your messaging to be, taking that into consideration and taking what the candidate has in there, in the data that they put in, combining that is another way to do it as well. So one is you have different options that you can choose from where you can also tweak and customize. The second way is customizing it with your own way, things that matters to you most. And then the AI can generate the message based on your input. And then get to a point where you feel good about it and kind of repeat and rinse, like for the candidates. Because you—

I definitely did that. I mean, I did a lot of A/B testing with it to get to the point, but I'm telling you that it wasn't just injecting keywords from their profile. Have you seen, or you probably haven't yet, but have you heard of the new LinkedIn Assistant?

Yeah, and I'm hearing a lot of complaints, by the way, also.

Why?

So one of the things they said is it is kind of taking away their power, what they are, doing before, like the flexibility. It is just forcing them to use it and they're not getting the results the way they are looking for. And this I'm hearing again and again and again.

The generative AI that it does is basically doing exactly what you talked about. It compares the, the resume to the job and it just inserts a bunch of like— so what happens is that message doesn't demonstrate why this job is important to that person.

So I mean, I'm not giving the whole thing, I'm just saying like there are 10 different variations The way we are approaching is we are not letting them to pick just that or nothing, right? We're giving them different variations of it.

No, I understand what you're saying. That's my question is how is that different? How is it? Because scaling by, by just scaling, by just saying keyword from job, keyword from resume, Mission Mob, ChatGPT doesn't work. I mean, it is, it's an improvement over not personalizing, but people still can see that they— when, when you get a message written by a robot you know that it was written by a robot. It is so obvious, right?

I think what, what the other thing that we are also doing is we are also taking other data points that are within, within that. So you're taking one is the available information, also anything their achievements or significant, you know, like a journal they wrote or anything else that you have on them. Okay, so you're doing some other using them as part of that also. That is where we are also incorporating some of these additional data points within that, right? So that it's not going to be generic, but it also goes much deep dive into it. And also, you know, we can only operate with the information what we got. We don't want to manufacture it, right?

Right, right. We— right, yeah. I mean, there is a— there's a built-in limitation with LinkedIn's AI that it can only use whatever's on the profile. And if the profile doesn't have anything in it, um, you know, the message is going to be pretty bland because there's nothing for the AI to really match, right? Um, what I did with my automation, which was really not automation, it was assistive technology and I want to be very clear about that. I didn't just set it up free. I didn't just say, here's 37 candidates, send them all a message. I did go through each message and review it, but it took me only a matter of seconds to review the message because I was just checking it for quality. The generative part of it put the messages into a queue. And so I had all these variables that were in there, but I also had my own language that included some of, you know, a little bit about me, a little bit about them, a little bit about the company. And the prompt that I wrote helped with the translation of their trajectory to the job. So for example, if they were open to work, it took advantage of that. If they were in a similar company, I took advantage of that, those kinds of things. So there's some if this, then that kind of rules that were built in. By the way, once you build that once, all you have to do is change the content of the original message and it works really well. But before I hit send, I didn't send all. I hit send into a queue and then I went send, send, send. That's a lot faster than 5 minutes. I still had to read through because every once in a while there was a couple in there where the message didn't make any sense because there wasn't enough information.

Right. That is where the human insight comes into play. But that's what I'm saying. The custom prompt option that you said about me, that part you can inject into it as you're comprising and then kind of adding.

Yeah, I think we're on the same page. Yeah, I think we're both saying the same thing. So the answer to my question is, to be able to achieve personalization at scale, it is not— the answer is not to have, you know, source 100 candidates and then spam them all with an AI that does all the talking. That's not the answer. The answer is to Create a well-written message first, then consider the right variables and make the query, make the prompt be very real about you and your company and the motivation for them and the motivation for the company. For example, I used something that I knew about the organization, about their mission, because a lot of these candidates were mission-driven. They were PhDs. And sometimes, you know, often PhDs have, you know, they're more mission-driven than they are money-motivated in some cases. Pretty frequently. So I use that. But what I'm trying to tell you is that I think it's not a matter of set it and forget it, and people think that that's the answer. The answer is to have the tool pre-generate 80%, 90% of the stuff, but you still need to hit that send button. You still need to supervise it. It's a big accelerator. In order to write 37 messages, it would have taken me much longer to write them and personalize them manually, then I was able to do it. It still only took me a few seconds per message, but that's a lot faster. So I'm just kind of trying to drive at the point of like automation doesn't mean a robot does everything. Automation means a robot helps you, right? Like driving assistive technology. The difference between your car has assistive technology to help you drive, like stay in the lane, radar for collision detection, all that stuff, The car doesn't drive itself. It doesn't decide where it's going to go. You decide where you're going to go. But you can set the cruise control and it will stay in your lane. It will hit the brakes. It will turn on the wipers. That is automation that makes it easier to drive and reduces driver's fatigue. But it is not a self-driving car. People think that automation is a self-driving recruiting system. That's not the answer. Yep.

I totally in agreement with you on that, Shelly, because that's exactly what we do is we don't automatically send it, but we create the message where the user have the opportunity to tweak it, but at the same time incorporate additional messaging through the custom prompt so that they're getting the best of both worlds. So yeah, yeah, that's kind of an important thing to have that human insight. And in the white paper, we discussed about the Salesforce example with the AI-powered personalized messaging, the impact of it. So that's also kind of important thing.

Yeah, I mean, granted, it's a good example to use. I know I cheated a little bit there because Salesforce has a ton of money and they spend a bunch of time on this AI technology. Not everybody has the luxury of doing that. You know, not everybody can be the top company, but it's still a good example because that shows you what you can achieve.

The idea, right?

Yeah, it's achievable. It is achievable. And sure, it costs them a lot of time and money, right now, but the next time they do it, it'll be faster, cheaper, just like that new deep, what's it called? Deep Six? No.

Deep Seek.

Deep Seek. Yeah. It's not that great, but it's a lot cheaper and faster. So that means that better ones will be coming around, right? So the big deal about something like Deep Seek isn't that it's so much better. It isn't. It's actually worse. It doesn't do many of the things, doesn't remember things that you tell it, stuff like that. But it's a faster, cheaper way of building something. So what that means is that the next generation that uses faster, cheaper ways will actually get better. So first you get better, right? First you invent something and it's hard. Turing, when he invented the calculator that used punch cards to make— to create possibility statistical analysis algorithms, math, he was using advanced algebra and calculus. On a machine that took a huge room and so much time I can't even imagine to put together. Now that thing can be done by a chip that fits in here, right? So first he did it the hard way, then it got faster, cheaper, and then there was another evolution. So this is another evolution in that step. But I think that we lose track of the fact that these are all steps. ChatGPT didn't just arrive on the scene and all of a sudden solved all our problems. There's no instant fix. There's no, you know, super, super button that does it all. You know what I mean?

Right. I think it's an evolution, right?

So it's a— yeah, it's an incremental— incremental innovation eventually leads to a breakthrough transformation. And like Deep Six, but that breakthrough transformation may not necessarily be the best one to use. It just is a breakthrough that makes other incremental steps. So, the breakthrough makes other steps possible. ChatGPT 3 was awful but 4 was much better and the new 4.0 or whatever, 5, again, much better. But there's a lot of steps that go in between that. So, as a recruiter, I want to remind people, keep track of what you're doing, keep experimenting, but always fall back on what you know works, which is your quality control, your value add, That's the key. If we continue to add value, automation will work. But the moment that you set it and forget it, that means you're no longer necessary. Yep.

So yeah, make yourself irreplaceable.

Yeah, by adding that value.

Yep. Yep.

Yeah. Anyway, uh, we're over time. I know we all— you and I always do this, man. When we get together, we like talk more than the appointed time. Um, so we talked about adaptability, training, We talked about how to be more efficient. We talked about smart ways to use technology. Is there anything that you want to leave us with, like one last thought? And make it a question. Ask a question that we should ask ourselves as we rewatch this episode. What should we be asking ourselves?

I think, do you think education and adaption is key? In order to achieve the, the precision that we need with this talent abundance and the engagement? What do you think?

I mean, that's a yes or no question, so I'm not really going to think about it too much. How about we take that and we say, how are you educating yourself? How can you educate yourself? Right.

Educate yourself. And is adoption a key thing? What do you think about adoption? Because that is something we wrote on the white paper, right? So adaption is—

yeah, how can you adapt? How can you educate to adapt? Because if you're not— I think if you're not adapting, uh, I mean, that's clearly—

yeah, but if you don't adapt, then it's obvious, right?

Get out of your own way. So yes, I think it's pretty obvious that we need to adapt, um, but I like what you're saying, which is, you know, the question to ask ourselves is, how are you adapting and how are you educating yourself so that you're capable of making that adaptation. Cool. Good way to close the program, man.

Hey, thanks a lot, Shelly. Thanks for having me. It's always a pleasure talking.

Always a pleasure talking to you too, Gavin.

And, you know, get ourselves better in general.

Don't forget, folks, the white paper is out there, uh, just scan to download. It's, it's free. It's just a white paper that we were talking about here. And, um, stay tuned for next episode. We will have Steve Hardy from Fiserv, a former colleague of mine who's a very well-regarded TA leader, and I think it's going to be a really interesting conversation. So please join me with Steve next Friday. Thanks everybody.

Thank you. Thank you all for your time.

Be kind to yourselves. Take care. Bye-bye.