Episode 57: Kshitij "KJ" Jain - Unleashing the Power of Data with Joveo

Speaker A: Thanks for trenching in. You're listening to TA in the Trenches, the podcast for HR and TA leaders making workforce decisions with real financial and business consequences. I'm Matt Reimer, co founder and COO at Rogue Hire. Rogue Hire runs the largest healthcare TA benchmark program in the land, where we know one thing, uh, acutely that top performers don't have more applicants. They waste fewer of the ones they have. Each episode I sit down with CHROs, VPs of TA and recruitment operation leaders living at the front of that problem, Operator to operator. Less opinion, more evidence. Let's get into it.

Speaker B: Talent Acquisition in the Trenches is brought to you by Rogue Hire, the team behind the 17 year Healthcare TA benchmark program and the makers of medics. Your ATS tells you what happened. Medics tells you what to do next. A patient monitor for every open wreck. Continuous vitals, early warnings, one clear signal on track, watch or critical. Your team works from the same truth. Your dollars go only where hiring risk actually exists. Learn more@roguehire.com

Speaker A: thanks uh everyone for trenching in. I am your host Matt uh, Reimer. I'm the uh, co founder and a COO of a company here we call Rogue Hire. Today, uh, on the pod we are continuing in our series to tackle uh one of the biggest challenges in our field which is unleashing the power of data automation and advanced AI. Um, we have the perfect uh guy uh today to help us bridge some of that intelligence gap, uh between your marketing spend and uh, what's happening in your ATS as reality. And so an industry icon known as kj. Uh, he is recognized uh globally for his passion for solving the most uh difficult uh problems in the talent acquisition ecosystem. He founded a company called Jovio, uh, which we're going to learn a little bit about here in just a second. Um, started though his journey with over 20 years of recruitment marketing experience, founded a company called uh, Mobile. Um the interesting thing about Mobile is that it got acquired by uh Indeed and it's actually the technology that runs um today, uh, what we know and love and what is called ah Indeed's Easy Apply. Um and so in addition to all of his professional experience and the work that he does in our industry, he's also a trail runner. Um, and so I'm going to dig into that one a little bit. And a deep sea diver. Um and so uh, joining conversation today is the co host of uh here of this segment, uh, Ryan Affilter, and so partner of mine here at Rogue Hire. And so Ryan as well focuses ah on Top of the funnel, talent attraction, uh, with a deep expertise in candidate experience, employer brand and programmatic marketing. So Ryan, what are the learning objectives here, uh, today?

Speaker C: All right, first of all, welcome kj Always good to see you. Uh, I've been a long time jovial follower and customer during my run as uh, head of recruitment marketing at Lifetime. So I think today uh, we're here to kind of address a uh, problem that both uh, the rogue hire community and Jovio are aiming to solve. And that's really the reality in recruitment marketing. Often TA teams are flying blind, spending budgets on programmatic, uh, for recs that might already have deep pipelines or even worse, uh, pouring expensive candidates into a black hole workflow where hiring managers haven't moved a resume in uh, a few weeks. To me this conversation is really moving past that gut instinct, uh, the vanity metrics, uh, of the past to have a strategic conversation about bridging, ah, the intelligence gap between recruitment marketing spend, uh, and analytics. So, so great to be here. Welcome uh, kj, Excited.

Speaker B: Thanks for having me. Ryan. Matt.

Speaker A: Kj. Um, so the first question that I want to get to is the name Jovio. And I think uh, as we lean in here, just uh, maybe understanding why you called the company Jovio and then just for everybody on the call, maybe a little bit of ah, an introduction to what Jovio does and its expertise. And then really I want to dig into this idea of ah, wasted spend, uh, in the programmatic space and begin to get your point of view on how you look at um, this community, uh, who we're talking to here today, which is primarily healthcare talent acquisition, practitioners optimizing the limited, in some cases recruitment budgets that they have. And so Jovio first. And unless you and I lean in hard on cost, uh, efficiency.

Speaker B: Efficiency, absolutely. It's a great question. Um, I just didn't want to name a name for the sake of naming a name. You know, people try to find like some very quick, easy to remember name. I think there has to be a purpose, a purpose that unifies my entire company. Right back in the day when we started we were just a team of three people, but the vision was it would be 300 plus, which we are today. And it represents a modern age. The acronym for a, uh, job for everyone. Now what does it mean? Uh, it means is we believe that there is a right job for every person in the world. There's nothing, uh, irrelevant or a bad quality applicant. It's all about not just putting the right job in front of the right person who has the highest likelihood of getting hired, but also helping that individual through the process where the conversions can take place and the right person on the other end of the stream can look at that person and say yes, I got the kind of I'm looking for and gets hired. So our job does not finish till that person gets a job. That's how we believe it's a job for everyone.

Speaker A: So Jovio is a job for everyone in the end. That's kind of the meaning of it. Yeah. I love it. So, um, maybe talk us through um, once you left Indeed. Um, why this space? What is so intriguing about the opportunities, um, for a job for everyone? Um, and then fundamentally um, you all um, at your foundation are programmatic job distribution. But as I understand you also have other kind of key features in the funnel that you focus in on. And so, um, talk to us a little bit about the post Indeed experience and then why build this company now?

Speaker B: So I think uh, as a part of the product leadership team, I used to report to Chris, uh, Hines, uh, who uh, later became the CEO of Indeed. And I learned a lot from him, uh, uh, looking at data in a different way, right? Looking at the data at the God level and to see where the inefficiencies are. And it was always a purpose to deliver hires from an Indeed standpoint. And I started realizing that in the process of delivering a hire there are inefficiencies at every step of the process. Uh, right from the time you decide where to even post the job, uh, how to reach to the right candidates, then you bring a person into a candidate experience realm, uh, which is person comes to the job postings, they've seen the job postings multiple times. One in the career side, then on the ATS and then apply process starts. And it's not conversational, it's a dumb form that you're putting blockers every step of the way from creating a login for that matter, right? 30% people drop off right there. Just to give a small example, which means that 30% of all your advertising, recruitment, marketing spends dies right at that one step. Uh, conversational experience, which is very rule based, not intelligent. Um, you're not able to have a right chat with the right person. There's no screening that is happening out there. Uh, you're not even for example, like one of the things I believe that 70% of the people and I've seen with data actually end up applying to the wrong job in the company.

Speaker A: Say that again, 70% of people end up applying to the wrong job Is

Speaker B: that what you said?

Speaker C: Correct.

Speaker B: So uh, if you see the data, you realize once we have the candid data and we have the jobs data and we said this candidate apply to this job, they are not analysts, they are not uh, data scientists. Don't expect a job seeker to be an expert on how to search for a job. So it's our job to deliver that right job in front of them. But M means is they may have started applying to a wrong job and that's okay, but we should surface up the other jobs which they have the most likelihood of being hired in terms of skills, match location, uh, competencies and so on and so forth. So to be able to put that in front of them, uh, uh, if there is an interview that can be used, AI interview that can be used to basically make sure that very quick level of screening can happen in real time so that uh, you can get all the answers you want to get intelligently. So the AI interview is a part of what we do, uh, conversational. As a part of what we do, we build uh, uh, whole career sites. So entire process and then the re engagement gets automatically built into the entire process. So by the time you finish, we want to make sure that you spend less than half of the money you're spending today in advertising. Advertising or optimizing advertising was just the starting step. But we know that if everything else was done right, you have to spend a lot less. And that much money you can put in a lot of new A.I. initiatives, uh, you can experiment a few things, maybe give hikes to your team, uh, make life a bit better. I've started seeing that there are three times to four times more applications. These are AI applications. Mhm. Real applications. What can we do to make sure that the right job, the right person, the right tutor is getting the right results? That's the microcosm of everything that we're doing. We believe that we just don't want to be a commoditized job advertising, programmatic job advertising firm. We are married to a concept called 4 3, 2 1. A lot of people initially made fun of me even in my company. And the 4,3 21 philosophy was I wish there was a day that um, comes that four clicks, three applications, two shortlist and one hire can happen. So everything we do is from the mindset of is there a world that exists out there where we can ever get to 4321? So that's always the pursuit that we have.

Speaker A: So 70, 70% of the waste. I mean this is an interesting number for us to maybe anchor on um, you know, 70% of you know, humans, if you will, are applying to potentially the wrong jobs. Right. Uh, in the end, like let's just start with that high level premise. Um, and so the waste in our systems and so like we as recruiters, we as talent acquisition leaders, right, we know that the process that has historically been built um, is a uh, requisition based design process, meaning that it's built to fill a requisition. It's not necessarily built to optimize for the human that we're trying to push through the processes. So we've obviously talked about this idea a lot. Um, and so what you're, I think evangelizing is that you know, hey, if you get the matching right or if you get the right person at the top of the funnel and chasing through on that singular job, it makes the full funnel more efficient. If it's 70% today is mismatched, ah, is there um, a goal or is there a place that we need to be driving to? Is it 20%, is it 10%? Obviously 100% seems a little bit far fetched but where do you see this working? Well, um, to where uh, an organization is not filling their funnel immediately with mismatched candidates that they're um, basically just going to waste and throw away.

Speaker B: Yeah, I think uh, you're absolutely right Matt. I think 100% is uh, of efficiency is uh, very hard. I believe in a 2080 rule, Pareto's Law, 20 uh, percent improvements can improve it by 80% in terms of the match, uh, to get to 80 to 95% takes 10 times more effort and 95 to 98% takes another 10 times more effort. So what good is good enough? Uh, I do believe that if we can do a few things right, the personalization on the career side, when a person comes in, you may have just the IP or click id, but there's a fingerprinting technology. You get a sense of who that candidate is. You can have the right kind of a content showing up for the candidate, right. Not just the person who's logged in previously and coming back to it. The second thing you could do is make sure their job descriptions are optimized. So when a candidate is deciding which doc to apply to, they have a clear set of information. Don't make it like 10 pages long, don't make it two paragraphs long, uh, put the right content out there. The third thing I would say is build check ins during the apply process where if you look at apply Flow and say, hey, these are three of the jobs that can be great match for you. You may not have looked at it, so you should not stop a person from applying to a job that person wants to apply to. Because uh, uh, it can be an aspirational application, but at the same time you can surface up the right kind of jobs to that person and say, hey, this is where it makes sense. Uh, you can do it in conversational AI as well. And once that person is applied, the technology can actually identify that. Hey, I'm sending the right jobs after that because I know that these are the jobs that surfaced up and that has the right requirements that match your competency and skills. AI has gone to the extent that it is not a. Um, people don't have to be experts.

Speaker A: Mhm.

Speaker B: To find their right job. I think, uh, you can use contextual understanding and learning historical data. Excuse me. To surface back to them. Serve them, don't make them serve you. If I were to say it like that, right. We need to be servant leaders from that standpoint. So I believe that um, you can get to a place where the wastage on that specific aspect can be about 25%. That's a good number to go after.

Speaker C: Gotcha.

Speaker A: Okay, Ryan, um, and I guess uh, I've certainly got a thread of questions here I'd love to dig into with KJ here. Is there any place that you'd like to maybe see us go here first a little bit as we lean deeper in?

Speaker C: Yeah, I mean we can start here or circle back to it, but you know, kj, from your perspective and knowing the jovial platform that, uh, you know, for um, most customers you have the ability to kind of see end to end from um, the initial attraction stages, from our employer branding and our social posts, you know, down funnel. So we'd love to kind of lean into, to where you see kind uh, of the most common bottlenecks and where you kind of see the most common areas, uh, that are holding up, whether it's your social media budget, your programmatic dollars, uh, at the top of the funnel or the middle of the funnel.

Speaker B: So if I understand your question right, uh, where is the funnel inefficiencies surfacing

Speaker C: up in the funnel, the common bottlenecks that TA leaders should just kind of be on the aware of.

Speaker B: Yeah, I think just. I'm going to make it very simple. Dumb distribution of jobs. I don't want to use that word, but it is truly right. I get a job and it gets automatically distributed to these three places and that's about it. There's no intelligence. One source could work for one location for a certain type of role better than some other job in some other location. Right. So sources, the efficacy depends on so many things at the same time. Right. Some places there are not enough candidates on a job board ecosystem. So you understand job board ecosystem represents a third of all the job seekers that are out there. So now is there a passive audience through social that you can reach out to for that particular geography? And wasting money on the job boards which are not going to be relevant does not make sense. Uh, so I think that is one, the distribution aspect of it has to be improved. Uh, that's a very important thing. The number two I would say is, uh, uh, I think the optimization part of it. So we believe in leading a lagging indicators. Hire is a lagging indicator. Uh, Companies can take 45 days, 60 days, 90 days to actually hire a candidate. Now I'm going a little bit down the funnel, but the screening could be a very immediate feedback loop. It can be two days, one day, three days. So if you have to wait 60 days to correct your advertising budgets for a certain job, the job is already filled. You can only fill fix it for long term future. So understand which part of the funnel represents quality. I believe that once the candidates are shortlisted, uh, their path to hire is pretty much certain. The third level of inefficiency is understanding the funnel and the quality of the funnel a lot better. For example, I've come across all my life. Uh, see, you don't understand a recruiter's mindset. You're not investing in that. A recruiter needs to provide 10 shortlists because they need to deliver one hire. They're going to look at 50 applications, they get the first 10 good shortlists and they're not going to look at any application that comes after the first 50 because they got attention and you continue to spend on that. So understanding what the shortlist to a higher ratio looks like so that you can know where to stop the funnel. Similarly, all jobs are not created equal. And I can go on and I'm going to uh, say the last point. Some jobs are very hard to fill. Some jobs have remained unfit for a very long time. Some jobs have a huge financial cost to the company. Right. Imagine if there is a way we can see how long has this acquisition been live and what are the billable dollars that attach to it, that the company has lost a hospital system for those billable resources or for places where you are having contingent resources, you have to hire. There's a huge cost attached to it. So why don't you prioritize that versus a job that you're going to get it. I'll say one more last thing and I know I'm putting a lot of things out there. We don't even have the basic intelligence to be able to predict that when a job goes live, how many candidates am I going to get from a career site or how many will come from a CRM system? Companies wait for a month or two months before they realize oh my God, there is a panic button that we need to hit and now we have to hire. But that data exists. We know for a certain type of a job the native organic traffic will deliver certain amount of hires anyways or certain number of applications or shortlist. So plan for that upfront and advertise right from the beginning and not after the fact. So many things can be done if it is an intelligent system which is looking at the right kind of data sets, providing the right kind of insights, reacting the right time. A human cannot look at 4000 jobs no matter how you say it. And a rule based system will always go wrong because it's not intelligent. M.

Speaker A: We see um, and so we just had um, our um, annual uh, customer event, right called Basecamp. And um, we had a little over 150 healthcare TA leaders from across the US um together uh, for two and a half days. And uh, we talked um, about this um, very real distribution uh challenge. And I like the fact that you started there because I believe that in the end that's where um, the waste begins. And some of it is very simple, right? Um, simple in such that we as recruiters leave our jobs open too long. To your point about how many candidates do I actually need to fill that job? We load these funnels up, we overload them sometimes and we create new waste. Um, I think um, we've been taught at some level, um and I think at least from my vantage point just being in this for a while that that programmatic spend, uh, which really focuses on um, clicks and applies and gets us focused on like a cost per apply is good. Um and that's needed I think in the industry but what we see uh, in our data, right and this is data that's pre funnel data. This is active data with our clients. What we see is that um, you know, upwards of 40% of the jobs that we the employer and so this isn't an indeed problem, this isn't even a jovial problem. We as an employer are suggesting that we need to have advertised, don't need advertisement on them for one reason or another. Right. A, I'm getting enough organic traffic, B it's in my CRM. C the job has an active offer on it. D the recruiter hasn't progressed uh, down the line. Right. And so I think sometimes we get frustrated with maybe third parties and um, the ecosystem that we're tapping into um, to help on the positions that we need flow on. But at some level it's our own root data, uh, and it's the cleanliness of that data and whether we can actually turn the data that we have in our ATS into uh, intelligence and then signal frankly to our partners, whoever that is indeed pick them, appcast, jovial, whatever, signal to them, hey, spend on this and then don't spend on this. And so from my perspective that's really where this conversation of gaining some power, um, and then gaining some efficiency in that budget to then do some other things, which is what I like, like what you're talking about, um, that either point that budget at some of my hardest to fill requisitions or allows me to point that budget to create mid funnel efficiency that maybe I don't have today. And so I guess from your perspective in the industry as a whole, um, you know and you look at, you know and I don't know what indeed is. It's what 30 to 40% of all jobs may be posted or the traffic is, you know, um, you know, is indeed um, the rest is everybody else type of thing. Um, are there um, changes coming to how the consumer um, based upon AI and geo, how the consumer is going to find a job or do you feel like the channels that we're using right now are relatively the channels that will always exist when we need to go to market, if that makes sense.

Speaker B: They will not. Um, uh, I have, you know, I remember back in the day, right uh when people were using their mobile devices for look for jobs, right. That's when the first startup idea came up. Um, I remember every quarter I would see the adoption of people to look at jobs change from 1 person to 2%, 2 to 4, 4 to 8, 8 to 12. Right. I could see that every single quarter the number was going up. How people's find jobs is fundamentally going through a shift. Um, it is not an SEO anymore, it's an aeo. It's an answer engine optimization that needs to happen. We um, need to be able to do a better job of providing insights on what is it that would make people find your Jobs where your competitors don't have an advantage, where they are not investing in AEO and what will make that brand stand. And the SEO game is dialing down. It's the first time in the history of Google for last. If you see that actually traffic on the, on the search engine is going down and the AI engine is actually going up where people ask questions and they're getting answers as AI answer.

Speaker A: So just uh, to make sure everybody that's listening heard that. I think it's an important point especially for our industry. We're seeing for the first time ever SEO, so search engine optimization, things we've been taught that we need to be good at. We see that traffic winding down or come or having um, a downward trend. And, and then we see the likes of, and I'm assuming, right, it's the clods of the world, it's the perplexities, it's the ChatGPTs, it's these types of um, engines that we're now beginning to use. We're starting to see the job traffic wind up on those. Is that, is that an accurate and simplified statement?

Speaker B: Very, very accurate, yes.

Speaker A: And so basically like the two plus two there is that, you know, if you're dumping all of your money into SEO, maybe it's fine or whatever for the near term, but as a midterm and long term strategy, that might quickly um, be a disadvantage for you if you don't understand the consumer behaviors that are coming online.

Speaker B: Correct.

Speaker A: That's great. Are there any other major disruptors outside of. And one thing at the conference that folks were talking a lot about and I think there was some energy around this was that, you know, all right, so we're using AI to become more efficient, but then the job seekers, right, or whomever is also using AI to put more applications and more flow. And we even heard um, um, uh, KJ um talks of fraud. Meaning like we see some uh, instances where um, applicants who are not even real making it into the pipeline and making it far enough down to the pipeline that they're talking to a human. They're potentially even getting an offer and an offer potentially needs to be rejected. And so what are some of um, your vantage points on the consumers using AI to maybe flood us or flood the employers with applications.

Speaker B: That's the cutting edge of technology evolution that's happening right now. I don't think anybody has really thought about or try to solve this problem. Let's talk about conversation layer. Person is chatting. What proof is that that on the other end a smart person Is not using AI to answer those questions and do that. Right. Uh, uh, there are companies that have raised tens of millions of dollars because they're just telling candidates, hey, I can apply to 100 jobs in a week for you, whether through a chatbot or through, uh, stuff like that. I think there are a lot of small little parameters that go into understanding that. Is it AI that answering? Right. A human is randomized in their clicks, right? The keyboard clicks. Uh, A.I. is not. A.I. would use a certain type of language all the time. Humans will never be perfect. AI, ah, will be perfect all the time. The speed at which the responses happen. Uh, for example, if you're having an AI interviewer, right. Uh, how do you know that the person who you're looking at, but the answer that is coming is actually coming from the person or something to AI, right. You can still have a lip. So how your lips are moving versus what the answers are coming in, you can start seeing the differentiation versus that. So every single thing, I would call it as a guardrails. And if AI is not built in your company with guardrails, then I'm telling you that you're going to have a lawsuit sooner than later. And this is extremely important. Europeans are far ahead of the game versus uh, uh, uh, US in North America where they're putting a lot of those, uh, compliances and guardrails in place for that. So everything that we try to build is knowing all the possible guardrails that are there and try to incorporate that and knowing that is there an AI that's working behind the scenes to do what a human should be doing. Uh, now that being said, many, uh, a times, these candidates that are coming through system, they're real people, but AI is filling that out. Do you want that? Do you not want that? I think that's a question to ask. Uh, I do not like the idea of, you know, I think, I think it's a full cycle. Right. Initially we thought easy apply was such a great thing because there was a bad apply process. Now the world is going back to. The thing is easy apply is not really a great thing because AI can mimic it and replicate it and just fill out hundreds of applications. So the world is going around in full circle right now. I think every fashion comes around if I want to say it like that. So we can still go with easy apply as long as you're doing it intelligently.

Speaker C: Yeah.

Speaker B: And I think I'm seeing Angela and Leslie asking questions on 432. Oh, I think Leslie already answered that. Yeah. Sorry. Yeah.

Speaker A: 4321. So four clicks, three applicants, two short lists, one hire. That's the vision.

Speaker B: Yeah. If you do it right, we can get there. Right. I've seen examples where it used to be like 110, like, you know, like 2 and 1 or whatever that is that can be reduced. Uh, uh, if you do right re engagement, if you put the right job at the end of the apply process, if you, if you have the right experience on the, on the candidate side of the house, I think it can make a big difference. Right. For a chatbot, for example, it's intelligent chatbot. Just ask a person to upload the resume and say, these are the best jobs for you right there. Don't even ask that person to search. Making a person search is old school. People are not searching on Google anymore. It's Gemini that's leading them to an outcome. I was talking to one of the large, uh, healthcare, uh, customer of ours and uh, uh, she was telling me that we're getting so many applications through this AI, we don't know what's working. Can you guys provide a report that can show us that what are we doing and what can we double down on? And we were able to provide that. You're doing this right. You can do small tweaks and you can get 10x more. Now they know why what is working? And lo and behold, your competition is not doing it. And AI is again, uh, learning all the time. So the, it's like think of it as a greenfield. Right? A land grab. Right. The earlier you can put your flags in the territory, the longer you'll be there.

Speaker A: Ryan, I think we've got about 10 to 15 minutes left here. I, you know, I like to stop around 145, so I want to give you a little bit of space, see if there are some, some questions for kj. I want to come back to KJ and talk a little bit about the human role, um, you know, in hiring, you know, based upon, you know, basecamp and our conversations there. We had a lot of, um, dialogue. Right, about, you know, where the humans at in this process and you know, some of the risk, um, you know, I think, um, to what KJ was saying around the guardrails here and how to set up maybe the right guardrail strategy. But, um, Ryan, is there something that you would like us to maybe get into here before I go down that rabbit hole?

Speaker C: I think we're doing a real good job here checking all the boxes. I love the idea that we're leaning into this concept of geo, um, right, the generative engine optimization and how that's impacting job search traffic, especially on the organic front. That's ah, super interesting. Um, yeah, uh, kj, maybe just a question, either going back to your mobile days to now with Jovio, uh, but looking at this 4, 3, 2, 1 mission, which is really interesting, is there one piece uh, of advice or one fix or one major leak that you've seen that's happening in the modern candidate experience that you know, most of us in talent acquisition are overlooking or one simple thing that we can, we can fix today?

Speaker B: I would say instead of being a very tactical approach of saying that this is the one thing, please go and fix it. I'll break it in two parts I would say. There's so much going on right now it's very confusing. Pick one battle, focus on one thing, solve for it, show the results, claim uh, that win and then add more in AI. Right. Just you can't just take all of AI and do all of it at the same time. Uh, and the one thing in continuation of that is please, and I request you all who's listening and anybody who's not listening, please always have AI used uh, in a way that is always a human in the loop. Do not let AI take hiring decisions. One wrong step and we are going to be paying for it. Our organizations are going to be paying for it. So human in the loop is also an important thing. The last thing I would do is the lowest hanging fruit as you said. Tactically, um, I think just bring some intelligence in your job distribution from some down the funnel metric right out of the gate. The funnel metric which is leading and not a lagging indicator. Right? Not a higher but let's say some shock, some screening where you know that I've got a qualified relevant quality candidate. Uh, so if you were to do one thing, just make sure that you can connect the two ends.

Speaker C: One, one follow up question to that kj, so you talked about you know, the, the leading indicators following you know, potentially quality signals. Do you ever see within your, your platforms optimization of customer spend or uh, healthcare TA leaders looking at it in the inverse may be looking at the reduction of waste. So optimizing programmatic by looking at what funnels, which channels in our multimedia mix are creating the most noise?

Speaker B: Yes. Oh wow, okay. That's a whole segment of wasted spend out there, right? I can talk about so many things that you have enough candidates already. Don't spend more on that. Right? Spend more on the candidates where funnel sufficiency is not there. The jobs which have the highest cost to you as a company in terms of billable dollars or contingency dollars that you're spending on. Make uh, sure that if the process is getting stuck somewhere, bring that insights right to the, to the recruitment, uh, marketing teams. The hiring managers say hey, there are so many people in the interviews process that's stuck right till this process is forward. Do not spend more time on this. I think there's so much more that you can do to stop wasted spend. Um, there's just a lot we can do. But I think if you go to wasted spend and you start doing proper. See I'll tell you the elephant in the room over here is integrations. The atss. It's a one time pain. If you do it right, the gain is long term. Without data, the sanctity of the data, the quality of the data, you cannot really solve any of these problems.

Speaker A: Yeah, I was going to go just there. I mean to me it's about um, getting your ATS data, um, you know, before you start going to your partners, whoever that might be, getting that data, um, tight, um, and getting good data definition there, getting good data rigor there, um, and getting it clean so you can actually create the intelligent decisions without that data foundation. A lot of what we're talking about here is all for naught when you think about the big budget line items and most talent acquisition departments it's only a few things, it's three or four things. It's the team, it's the labor and so that's a big expense slide up. So if I've got, I was at UPMC, I had 130 recruiters. When a uh, CFO is looking for efficiency they want to know how productive are those recruiters. Are uh, Those recruiters filling 212 wrecks per year on average or are they filling 112 recs per year? So that's a cost factor. The other big one that we don't um, always talk a lot about is just background checks and the expense of processing uh, things. Big budget line item typically in that third spot outside of your capital expense on your technology, it is the marketing expense. It's what you've got baked in your budget for top of funnel. Um, and we um, I uh, believe have a responsibility to make sure that that budget which you've got some control over if you've got good clean data, um, we gotta make that budget work really well and really efficient in healthcare um, today. And so if Somebody believes that 40% of your marketing spend could be re leveraged because you're overspending on these filled pipelines. You're underfund, which is often an issue. Right. I'm underfunding some of my hard to fill roles. I need more budget on some of my harder. I'm just spreading it like peanut butter. Um, that to me is a very easy win that you can collect, um, right out of the gates. Like, I think you can collect that win, um, within months. But to your point, it starts with good quality data that can turn into intelligence. Um, if you don't have good quality data, you can't turn it in to, um, intelligence that can actually be used and then ultimately automated. Right. And so like, what I, what I don't want to see happening is that I manually have to tell Jovio and KJ that I need something to have. I want the intelligence coming from my ATS saying, do this, don't do that, do this, don't do that based upon what's going on within the candidate data that's coming at me. Right. And so I guess, um, you know, uh, aj, kj, I wanted to dig into this idea of, ah, human in the loop. And um, you know, you hear people talk about that, um, and then you hear conversations like this, right? Where it's like, hey, you're going to need some AI or advanced automation to become more efficient and to kind of keep pace, get it right. I guess sometimes. And this was what, uh, you know, I think we were wrestling with at, uh, basecamp here. Sometimes it's like, where do I deploy AI? Like, where is it safe? You know, kj, is it simply on tactical things like setting up interviews? Uh, is it on tactical things like helping me write emails? Um, because, you know, when you think about the funnel, the funnel is, you know, basically decisions, decisions, decision, big decision, right? So decision on who I'm moving forward, decision on who I'm going to have interviewed, an interview happens, and then the big one, which is, I'm going to hire you, I'm going to make a decision on your ultimate outcome. I'm just curious, from your perspective, where is it safe and maybe where is it not safe?

Speaker B: Yep. I think what is always a safe place versus the who. Right? Uh, so what, uh, uh, means that where do I advertise jobs? Right. Uh, uh, contextually, uh, if AI can understand that a particular. I'll give you a very simple example. Right? Um, if you're doing the social ads and you're targeting nurses, uh, I can tell you pretty much 100% of your dollars are wasted if you are putting a target ad for nurse on a desktop. That's a contextual understanding that yep, it's wasteful. They are on the mobile devices, they're on the feed all the time. It's the mobile devices. So any ad that you show to them, because if you show it on desktop the conversion rates are abysmal. Like literally hardly any conversion rate. But on a mobile you will see a much better results. It's a contextual understanding. Similarly you can say that uh, ah, a certain time of the day right. Uh is where you get certain better results and then AI can constantly optimize for that right. A uh, certain type of a job description. So it's all about the what you can optimize. The who to hire should be a secondary. That's not the, the lowest hanging fruit the battle. You should fight the fight because the moment you get into who you're getting to hiring decisions of an individual. And this is also I think uh, goes to the security or it uh integration questions is as long as you can deploy as much as I AI as you can without the PII data. Yeah. Identify all that bucket solve for that. Right. Uh, you know uh, I'm talking about spend caps, I'm talking about dynamic thresholds, reactivation triggers, uh uh, uh, funnel sufficiency. They can all be done without knowing about the individual data. So that's the, the 1M thing I would say is instead see once you start, let's say you're having a conversation, right? Um, in a conversation the person gives information you match to the right job that's still perfect use case you're asking the right question, still perfect the moment you have PI data and then you start making a decision while it is doing everything that is doing right with all the guardrails. But make sure that all the compliances, all the logs, all the audits, all that report, all the trigger insights are all there with you at all times. So that if you ever get a question asked, you have an answer right out of the gate. And that requires a lot of learning and lot of understanding and we will get there. And I'm telling you healthcare space, the risk appetite or the amount of risk we carry in the healthcare industry is so much more higher than just working for a department store out there. People don't understand there are two different worlds out here. Yeah. So that's my sense over here.

Speaker A: I think that's great guidance frankly and an easy way to easy construct to kind of think about the who uh, versus the what? And so there's plethora of opportunity to deploy um, AI. And I fully agree, um, you got to stay away from the PII and stop calling it integration and just start calling it connections. Right. And so these are just gentle connections that we're making and expanding the power of your general ecosystem. Kg, this has been a great conversation, I guess. Um, any final thoughts or maybe things that you had hoped to cover today that you thought, um, maybe we should have covered? Otherwise, I'm going to maybe start to wind us down here and give uh, some folks some transition time.

Speaker B: I think the holy grail here is data. Right. I think uh, Rogue hire is also looking at data all the time and we are looking at it all the time. And getting clean data is hard. And I think I want to answer Wendy's uh, uh, point here is, um, I will make a very simple, I'll give a very simple situation. I'm telling you 90% of the talent acquisition teams in the world are actually working off wrong data. Yeah. So what happens is if you think of tracking a candidate to the down the funnel, it's all based on UTM tax, right. And the tying of the data together happens on cookies. If people decline cookies, the tying of data of where was the click? And if this is a person who applied to a job or went further in the process does not get tied. Right. YouTube tags just follow through something. There's another issue over here which is very important to understand that once a person looks at a job tag, passes on applies, but then the person goes and finds another job and applies. There's no UTM tag over there. You do not even know which job board is actually delivered. A quality candidate who has actually applied to a second or a third job where the tags are not going to work, they're all going to get lost in ads. So the way cookie list technology works is and if the right kind of um, tracking codes are put on the site, it can map all of this data. So you can see a candidate who's applied three times with the right kind of a cook release tracking, fingerprinting technology that can exist that still that level of insight will never even exist in the ats. They will never to a job board or a source after the first application, the second or third or fourth they would know. Nobody would know about it. This is my 2 cents. I think one could also see that in the ATS this person applied using this job board, but applied to another 3 jobs in next 1 week M and the candidate ID can be matched on our end, we can actually surface that up for you. So I think there's a lot about data and thinking about the data. So use your partner as a consultant and tell all your problems and the problems that do not exist. And this brainstorm first, before you go dive into and say, hey, I have $500,000 budget a year or million dollars, $2 billion and spend, let's plan the whole thing out first before you do anything. That's my last two cents.

Speaker A: I think it's a great last two cents. And if you believe, if you're listening to this and you believe that 90% of um, the data that we're working off of is potentially the wrong data, then that's opportunity, that's opportunity to impact your organization, impact your team and impact your own career. And we certainly are entering an era in my opinion. And I'm uh, a marketing guy that got into recruiting, that got curious about process. But I am by uh, no stretch of the imagination a data scientist or somebody that understands all of these models. But with frankly the advent of AI, this is becoming more accessible for operators, that data literacy, if you will, is becoming more accessible for the day to day operator. And so don't hesitate to lean in, don't be afraid of it. It does though start with um, your process and your ability to ensure that you've got some standard operating procedures inside of your operation that you're executing and that you're running. And so the data is just a reflection at some level to your operating procedures and how structured those are or aren't once we talk about recruitment operations. So kj, I feel like you and I could go for uh, a little bit longer than the 45 minutes that we have scheduled. Hey, one last question. I said you're a trail runner, so what's the longest trail running trail run you've been on? Are you like one of these ultra marathoners where you do like a hundred miles or you just kind of like 20, uh, miler or you just do it for fun?

Speaker B: Uh, I do it for fun. So when I was at my peak I should do about 20 to 25 miles every weekend.

Speaker C: Wow.

Speaker A: There you go.

Speaker B: I not a single trail exists in the D.C. area that I don't think I've been, have not been there.

Speaker C: Okay.

Speaker A: All right. I just curious if you're an ultra guy. So. Because if you're an ultra guy, then we have to have another podcast about that.

Speaker C: So.

Speaker A: All right, all your uh, information is logged in the chat here. Uh, we appreciate everybody that joined us live today. Um, we know your time is valuable and we appreciated all the great questions and, uh, constructs, uh, here today. Kj uh, it was a pleasure to get to talk, talk with you and meet you for the first time. Uh, we look forward to partnering and, uh, working down the road. Um, folks, um, uh, we'll be back here, I think, Ryan, maybe in a couple weeks with another live episode. Um, keep the, uh, questions coming and, uh, we hope everybody has a great day. Um, thanks again for trenching into this live episode, Ryan and kj, I'm signing off. Thanks, guys.

Speaker B: Thanks, kj.

Speaker C: Thanks, Matt.

Speaker A: See you now.

Speaker B: Thank you. Bye.

Speaker A: Thanks for trenching in with us today. If a benchmark, a framework, or a question from this conversation is worth taking back to your leadership team, please do that and let me know how it lands. You can always find me out on LinkedIn. For more on Rogue Hire's Healthcare TA Benchmark program and medics the decision intelligence layer for talent acquisition, visit roguehire.com if the show is useful to you and your peers, a subscribe and a review always go a long way toward getting it in front of the leaders. Making these calls day in and day out. Until next time.