Alex Marcus Live on Shally's Allley

Hey, hey, hey! All right, party people, welcome! You are live on Shally's Alley. I need some kind of special effects. This is the only one that I can come up with. Wait, that's not the best one. Maybe this one's better. There you go. You are live on Shally's Alley. Awesome! Welcome, welcome, welcome! It is, uh, the last episode in April. April 25th, Friday, April 25th. and we're shifting things a little bit today. We are going to have a conversation about, um, something I don't talk— I don't get to talk about a lot, which is, um, within the space of HR tech, it's, uh, system administration, sysadmins if you will, in the HR tech space. So I was referred to Alex who's an innovator, recognized innovator in the TA/HR technology universe and a founder, founded the organization called System Admin Insights, which is part of Integral Recruiting Design. I think I got that straight. That's right. Yeah. So, he works with community-driven solutions for systems administrators and helps advance recruiting strategies. So, everything from sourcing performance models to leveraging data to market changes and everything. So we want to talk a little bit about performance and sourcing and industry performance and how data can help us reshape our landscape and how basically systems administration can be a key to being successful with enterprise HR technology. Something that we don't talk a lot about. I mean, usually, what would you say, Alex, in an organization, a typical organization, there's like 1, maybe 2 people that do this job?

Depends on the company size. Yeah. So, you know, I was just hanging out with NYU and Fujifilm this morning. We did our first SAI live session here in New York City. And some companies have multiple system administrators. Some companies—

One for each system or what?

Well, no. So let's say a company under 1,500 employees typically does not have a person who's dedicated full-time to system administration, and it lands on the plate either of a recruiter who's great at spreadsheets or even a TA manager who has many other competing priorities, right? Once you get into like the 3,000, 4,000, 5,000 employee range, you get companies who have a person who is full-time dedicated just to running their ATS. And then once you go global, companies with 10,000, 20,000, they typically have a primary system administrator and then they have subsystem administrators who could fill in for that person and who manage the functionality in different regions globally.

Okay. So, there'd be like subsystems or other systems, alternative systems or regional. Got it. But it is pretty much one of those kind of like, you know, the unsung hero, if you will, right? You started Systems Admin Insights to lift HR sysadmins out of isolation, which, you know, It's the, in your opinion and your estimation, it's the single biggest blind spot that executives have about their sysadmin and their HR tech function. So, this was kind of like, you know, your, I don't know, your call to, your calling, I guess, to elevate that. So, what measurable action would you have them take? You know, people that are listening, what measurable action would you have them take this quarter to prove that they finally get the importance of this role, this blind spot, if you will?

Wow, that's interesting. In fact, we were talking about this this morning. It's like, how do you message up the value that you're bringing to an organization? I think the real prize here, the thing that these folks should really be focusing on is quantifying ROI. I have talked to so many TA leaders who cannot tell me their cost per hire, and it's because although it's a simple concept, there are so many areas of nuance and ambiguity that actually figuring out what your cost per hire is can be quite daunting. These system administrators are extremely well positioned to assist in that conversation because they understand the data on a very deep level. I talked to one system administrator recently, we were looking at his ISIM dashboard, reporting. And he said, you know, my manager wants to bring down our time to fill to 30 days, and currently it's 58 days. And I said, okay, let's look at your reporting. And so we dug under the hood, reconfigured their reporting to remove outliers, right, some other nuances in the system, and discovered that he was actually almost already there. Yeah, he was within 2 days of meeting that time to fill goal, just because the reporting hadn't been configured correctly. So if we're building up that internal role of the HR system administrator, and keep in mind, most of these people do not have formal IT training, right? They came from recruiting, they came from sales, they happen to be tech savvy, and somebody at some point had the bright idea to say, hey, why don't you run the ATS, right? So if we build up this professional role, we can get them to the place where they can actually transform the conversation around HR tech investments from being a transactional thing. A lot of leaders are just happy with the thing running and not on fire to getting to the place where they can say this is actually a driver of ROI. And if we make these optimization changes, we're actually going to bring down time to fill by X number of days and it's going to be equal. It's going to result in this specific dollar amount of estimated cost savings. That for me is like the holy grail of this profession. And it's a lot of folks are not there yet, but we're trying to get them there.

So I appreciate what you're saying with regard to, you know, getting, getting the garbage out of the data because that's basically one of the biggest failure points in calculating anything. I mean, I worked with an organization as a consultant. I don't just do HRIS or systems. I also work with strategy and leadership and everything. And I worked with this organization where they were confident that their time to hire, time to— well, what they were— and by the way, people define time to hire different ways. Yes, they do. Time to offer accept, time to start, time to you know, but the job posting, let's call it job posting to position, um, filled because the person actually starts that day is, is one of them. And so they were confident that they had a number for that, which was 75 days. And in this industry, it's not unheard of for people to give a little bit longer notice. So hence the little bit lengthier time, which is, you know, calculated in there. And when you report on this, you definitely want to report on the fact that you have This is how long it took us to post the rec. This is how long it took us to get an offer. And this is how long it took us to get the person to start. Because from the offer accept to the start date is kind of not necessarily something that you can really change all that much if people need to give notice periods, right? And they were convinced that it was 75 days, but after looking at the data, there were some calculation gaps. That turned out it was, it was actually a lot more like 190 days.

Yeah. Yeah. Well, you know, at each one of those milestones, there are about 5 or 6 different ways that you can measure it. So for example, time to fill—

you really need to break that down. I think it's important that you say, here's the, you know, you guys wanted, you, the leadership team, you wanted time to hire. Here's time to hire. However, the, the bar that says time to hire is broken into these 3, 4, 5 pieces because we may not be able to move these, right? You know, there's the, you know, in, in other countries outside of the US, there's mandatory, uh, notice periods of 2 months, 6 months. So you're not going to be able to get in Germany a time to hire of under 2 months ever, right? Ever. Because it takes 2 months to give, you know, they have to have 2 months notice.

Yeah.

Right. So it's interesting that you say that and the systems don't really take into consideration. They're just looking at the raw data.

And let's talk about time to approve, right? Approvals can be very different depending on where you are in the organization. And we can get more and more granular to get to every single action point and make a time-based metric out of it depending on how thorough a deep dive you want to do.

So where I was going with that, though, I was, I was going to ask Is it really— I mean, I think personally, and this is— I'm going to obviously ruffle some feathers here, but I really think that the time that it takes to hire, however you measure it, really is not useful in any way because it doesn't help us make any decisions if it's not correct, right? So what I would love to— what I would like to, which I don't see very frequently, is I would like to know how long it's going to take us to fill a job so that I can prepare my leadership to essentially let them know, if you guys want to hire somebody for this job, you need to start thinking about that at this date because that's how long we need. That's how much lead time we need. That's what I would love to know. But the reality is that this time to fill metric doesn't really do that.

I would argue that it can.

It can, but that's my point.

As long as it's configured correctly. And I think—

It doesn't. That's the problem. Yeah.

And I think a lot of TA leaders, they can kind of smell that and they know that like, hmm, I guess that looks right, but do I really feel confident in it? Like to feel confident in those figures, you need to have somebody in that role. Again, the system administrator who you know has a level of training and sophistication and experience. To really deliver metrics that are bulletproof. And that's just— it's a matter of professional development.

Yeah. Okay. That's what you're saying is the key there.

Yeah.

Okay.

And community.

So you have a community that you run. This is a voluntary thing or is it a—

So there's, there's two tiers. There's a general membership and there is a free call for ISIM system administrators every Friday at 1:30 PM. And we've had over 700 ISIM customers come through that free call experience. Over the last 5 years. And then there is a private community that includes unlimited small group office hours with our full-time team of ISIMS consultants.

And that's the Systems Admin—

Yeah, it's all Systems Admin Insights. There's an ISIMS community.

Got it. Yeah. Right. So this is for, you know, HRIS, HR tech systems administrators to share compare notes, learn professional development, troubleshoot, ask for problem recommendations, and not recommendations on what problems to have, but how to solve them.

Yeah, no, exactly.

Got it. Okay. That's pretty cool. I'm glad that, that I didn't even know that existed. So, you know, you have built, I guess, two user communities and user communities, by the way, just for folks that are out there and listening, it is not easy to build a user community. No, you know, companies that have large install bases have a hard time building user communities, and they have all the population they need. It's still not an— you don't just— I guess I'm getting stuck in my own words here. If you build it, they will come is really not true when it comes to user communities. So you have the, you know, the SAI, and then you have the iSIMS ONE. Um, when you When you reverse engineer, or if you could reverse engineer, the success of these communities, can you give us maybe 3 design principles that can really get a user group to become self-fueling and not— obviously, I mean, you could be in there administering and, you know, participating, but how do we get it to self-fueling?

Yeah. So, you know, we hired a consultant who had successfully built a community called foundingmoms.com. This is for mom entrepreneurs. And without— her name's Jill Saltzman. And without her guidance, I don't think we ever would have gotten this over the finish line.

So she's a community building expert.

She's— yes, exactly. And at the end of the day, it all comes down to trust and authenticity, right? Everybody wants a community. And just like you said, they want to slap it up there and expect folks will come and hang out and interact and thrive. It is its own full-time job, right? And being the community manager, managing the community, building, architecting the community, strategizing the community, it is a big lift. And I would advise people not to do it unless they're— I'm really into community building.

I love it.

I've been doing community building my life in various ways throughout my career, my personal life. It really is a passion for me. I would say you need somebody running your community who is passionate about community management, right? We are in there constantly providing value, connecting people, providing extra value. We have a special guest consultant coming in to do office hours. We're just— we're doing these live events now. We're constantly pouring value into this community, really, because it's what I'm— it's what makes my heart sing. And I think the whole team would agree. This is really the things that are going on there and the fun that people are having. And another thing I would say is that we have been very, very intentional around keeping it non-salesy. Of course, we have some, some courses and services that we mention, but we don't allow vendors in the product-specific communities. We want to make sure that this is a safe place for authentic interaction. And what we're noticing on the back end of the metrics— we use Circle.so, it's a fantastic platform, I would highly recommend it. They're constantly shipping new functionality, they're very AI forward. We can see the metrics and we can see that Half of our— first of all, we're in the highest engagement tier according to Circle in terms of like monthly active users, but that activity is not just posts and discussions, it's DMs, right? So people don't necessarily want to be on LinkedIn, right? You're on LinkedIn, you're constantly getting bombarded with sales and various things. Like, you want to be in a safe space where you can just nerd out about iSIMS or whatever the product community is that you're in, right? And we're expanding to other product communities as well. You want to be able to nerd out with people that you trust and you can DM them and get the answers that you need. And there's also a functionality that we built in where you can search for other members based on what platforms they use, what industries, what company sizes, what specific modules they're using. Right. So all of that architecting, you know, it really took, I think, a year before I was able to realize a vision that I originally had about a year and a half ago. So Yeah, it's really coming along.

Takes time.

It takes time and effort and passion. Yeah.

All right. So those are, that was my answer. Those are the 3 things. I just kind of recap here. Basically trust and authenticity, which is, you know, parentheses, keep it non-salesy, right? It's a full-time job for someone and that person needs to be someone who is a community manager, either knows or is passionate or is going to learn it. And that's going to be a primary function. Otherwise it's not gonna just run itself, right?

Yeah.

Then the third thing was the, well, I guess the part of that sidebar on that is keep it active. Like you have, yes, that person has to, you know, you continue adding content. So I guess that's really two. This is like two, one part A and B, two part A and B. Yeah. And well, I guess the third one is to have a platform that's actually a community platform.

Well, we looked at a few. We looked at Mighty Networks, Heartbeat, a couple of others, and I felt like Circle, just from the UI, you can just tell that it is very current, right?

Clean and simple.

Yeah. And after the first 3 months, they just— I feel like every month they ship some news. This is not an advertisement for Circle, but every month companies do that.

Yeah, I have an LMS that does that, and every couple of months they launch a feature that I'm like, oh, I didn't, I didn't know we could do like— I, I I could use that. It was, it's, it's not features that I'm like desperately waiting for. Like, when are you guys gonna do this? It's more like I already have what I need, but they'll come up with this feature and I'm like, oh, I know how I can take care, like I can implement that right now. So I love that. That's usually pretty good, um, product design and, and kind of staying on top of things.

Understand, understand their customers very, very quickly.

Yeah. And they're definitely adding value a lot because I mean, people can switch. It's, it's It's not that hard to switch nowadays. Yeah. Platforms, it used to be a lot harder, but not, not as much anymore. So, all right, well then, the, the next generation admin playbook.

Yeah.

AI, workflow automation, low code. Let's take a, I don't know, crystal ball, magic window, whatever, magic wand, project out maybe 3 years.

Yep.

What new capability are sysadmins and the leadership teams that will need to hire sysadmins need to understand and master or be replaced or automated into obscurity?

Yeah, so I think the full maturation of this role will be a title of HR systems orchestration engineer.

Okay. I think what is that capability?

So I think SaaS as we know it is going to fall further and further underneath the surface of what you're actually experiencing as a knowledge worker, technology worker in HR and other areas as well. So you're going to be interacting with AI and the ATS, the HCM, all these systems are going to be below the surface of an interface where you are the orchestrator. And so this is a metaphor that I learned from Josh Berson. He talks about the super worker. right? So you are the conductor of a symphony of agentic AI bots who are taking orders from you, right? And you are assessing the impact of what they're doing, and then you are taking that information and you are going to town with analytics in a way that we would all love to but don't have the time to do it.

So, so the durable skill sounds idyllic, but I don't think that that's a special skill.

Oh, I would say it is because it requires an understanding of 4 or 5 different domains that are— these are skills that are, even if we don't get there, these are skills—

I'm just feeling like it sounds like you have the agents to do the work for you. So your orchestration is really more just understanding the business.

I think it takes— I think it will take a nuanced understanding of integrations. Of APIs, of—

even though the APIs are supposed to be handled by—

I can't tell you how many times I've heard, oh, we had an integration between this system and this system and something happened.

But isn't that what Agentic is going to fix? You don't need integrations anymore.

So it will— I mean, supposedly, maybe someday, right? But every technology innovation I've seen, it is effective to a point, and then there are edge cases. That can completely derail something that nobody anticipated. That's where human judgment can be critical.

That's always going to be the case. Yeah, there's always edge cases. Edge cases. I mean, it doesn't matter how well, you know, with any automation, the only thing that you could really automate completely and unsupervised is when it is 100% repetitive. Yeah. And the outcome is the same each time. The moment that you have any variability, you really need to have someone supervising it. Because there's going to be edge cases. But I mean, if it's literally hammer a nail into a piece of wood and you have to hammer it at the same degree, at the same speed, the same strength, and it's the same nail over and over again— not the same nail, but the same quality manufacturing nail— yet yes, you can fully automate that. But if it's varying pieces of wood or varying lengths of, you know, now you have a robot that can handle most everything, but every once in a while— so yeah, I get that. I guess what I'm saying is Um, it isn't the— and I'm— I am— please understand, I am playing devil's advocate in that I get this question, yeah, from a lot of people that are not aware of exactly what agentic is and are being sold a bag of bullshit, yeah, from vendors that are selling something that is really not agentic, right? So the, the seller doesn't understand what agentic is, and the buyer doesn't understand the economic— the economic buyer doesn't they don't understand what Agentic is and they purchase this and then the user has to use it thinking that this is what the economic buyer told them, hey, this new system is gonna do it, but doesn't. So the reason I'm asking this is people are under the impression that Agentic means it's a robot that can log into a website and do stuff for you and you don't have to train it.

Mm-hmm.

So if that's the case, why do we need APIs anymore?

Well, your system, they're not just going to be logging, logging into a website and doing something would be the least inefficient way for it to do that task. Yeah.

Granted. But right now, let's just use a real life example, right?

Okay.

CRM, ATS, and HRIS. Let's call it, you know, so you got your CRM where you have your people that you're trying to get in touch with. They're not candidates. They haven't applied. They may or may not be interested. And you need to do relationship management and they're not associated with a rec. Then you have your ATS. They're now associated with a rec. There's a hiring manager, there's an interview process, there's a system that is in place, and then you hired them and perhaps maybe even onboarded them. And now they need to go into the system where they get laptops issued and their payroll and benefits and all that. So these are 3 different systems, right? Yep. It's pretty common to see CRMs that don't talk to ATSs other than one way, like sending stuff. Right? But not the other way, right? It's common to see ATSs that once again feed HRIS, but again, one way. And it's almost like, it's almost like a batch export. You know, here's a dump. So what a lot of times happens is a user is basically going in and cleaning things up and matching fields and, you know, and they're using a browser and they're very inefficiently, as you just said, very inefficiently copy, pasting, dragging, drop, whatever across. So the vision is that the agent is going to replace that meaningless, tedious, tedium job that is, you know, it's very valuable and it's very important because it keeps the hygiene between the systems, right?

Yeah.

But it's kind of a really mind-numbing job. So that's what agentic is supposed to be. But you're saying that the real solution of agentic is not like copy-paste type stuff. It's more what it's explain it to the people that have this misconception. What is it really?

Okay, so let's say for some reason when you implemented your ATS, you put, um, uh, hiring managers in the wrong login group down your data hierarchy, right? And, and you're a global organization, and it's— and there's certain downstream implications of that.

This never happens, by the way, ever.

Never, never see this. Right? And you decide, hey, really, I want this login group to be over here with the other hiring managers instead of with the recruiters or wherever it was built. Right? There is a just incredibly complicated spaghetti of things that are connected there, right? And to disentangle all of that, right?

I only know this from personal firsthand experience. So sorry, I'm laughing. But yeah, yes.

A leader might not appreciate the amount of effort that's going to go into that. Right. Your HR system administrator, meanwhile, is this like, like burning the midnight oil trying to figure this all out and try to get this reconfigured. Right. There are security rules attached to this, search locks attached to this, right. Visibility issues, right. That is something where I can imagine that AI at some point could search through all of the configurations in the backend stuff that we're never going to see. Right. And reconfigure that for you in a config site first. So then you can— then the human can then verify that it was done correctly and that is having the correct downstream impact. That's one thing that I feel like is a uniquely valuable thing about this type of professional role. We understand what it's trying to do, and then we can go look at what it did and give a thumbs up or thumbs down or give it feedback like, you missed this subgroup, or now our background checks aren't working the way that we want them to, or now this report over here on this dashboard isn't working right. Like, so that type of awareness of the impact on human beings using these systems, I think is something that requires a deep level of technical understanding without actually performing the work itself. Does that make sense?

It does to me. Yeah. So this skill then is, uh, no longer just kind of having the know-how to use ISMs perhaps. As an example, but actually learning about how these systems are interconnected and understanding the business side of the environment as well as the technology side of the environment, you know, things like security and user roles and also understanding the capabilities of whatever, you know, robotic process automation, RPA, which is what we're talking, you know, agentic. I don't, I don't know why this happened, but agentic is basically RPA. Right. Robotic process automation. So whatever that— and that's been around for a long time, by the way. It's just the birth of this new agentic for folks that are out there listening. So you used to have robotic process automation and you used to have machine learning. Machine learning sort of evolved into what we now call AI, which is really not actually true AI. It's not artificial intelligence. It's really large language models with lots of machine learning that are able to you know, trial and error and determine things and help you work faster. And then you have robotic process automation that is very difficult to set up and configure. And somehow those two are sort of coming together, and that's what Atentic really is— RPA with GenAI on top of it. So you're saying that we need to understand the business, the technology, the rules, and the AI, and that's the key skill, is that orchestration is actually really that complicated.

I'll give you another, um, uh, example. Um, I had a young person came to me and asked me whether it was still worth it to learn how to code, right? This kid was like 15 years old. Um, and I said, absolutely yes. And I can hire somebody for $7 an hour in Bangladesh who will code for me all day long, right? So it is very important to learn the skill of coding right, so that you understand the architecture. But is there going to be a job for you as a coder as we conceive of coding? I don't think so, or you shouldn't bank on it, right? So yes, understand how these things work.

Yeah. Should I really learn the multiplication tables because there's calculators?

Perfect example.

Yes, you should learn multiplication tables because you probably don't, you know, you're not going to do multiplication tables for a living, And even if you become a mathematician, you're still not, you know, but mathematicians today use computing and all kinds of stuff, but you still need to know the fundamentals. However, I would say this. So going back to the edge cases thing, when it comes to software engineering today, still today, we have embedded devices and software that's written, that's coded to that device. So there's always going to be specializations. And, you know, in software engineering, there's going to be robots that don't know how to do it because there's not enough examples for it to do. In other words, if the language model, if the, if the training data isn't large enough, the robot can't learn. So there's always going to be situations where this is software that only exists on F-16 fighter jets. Let's not use an AI for that. But, but again, that's not— so you don't— what you're saying is don't bank on that as a career. So for us, that means in HR systems administration, it's don't bank on knowing how to configure ISIMS as a job, but learn how to configure multiple systems and how they connect to each other, because that's probably more valuable.

That's what I'm calling the, the, the durable skills of this profession. Yeah, yeah, exactly.

It's not how— it's not what are the fields in ISIMS, or Workday or whatever. It's how do we make these things connect? What's important? What's not important? What are the things that could go wrong? Right.

I mean, look, I really built a business around being an ISIMS reporting savant.

Right.

I love ISIMS reporting and I would do it all day long for fun. But should ISIMS make that go away so that it's just a natural language interface where you can say, I need such and such reporting and it builds it for you? Probably. It hurts my heart a little bit, you know, like, yeah, right.

But that being knowledge is still going to make you an expert, except you'll be the expert that is training the AI. Yes, you'll be the expert that is developing this, the behind-the-scenes stuff. Everyday users don't need to be Alex Marcus, the iSIMS expert, just like everybody asks me, oh, is sourcing obsolete? No, absolutely not. Sourcing is not going to go away. Sourcing is becoming more difficult because of the edge cases, right? And because we have a different problem now that I don't know if a lot of folks really considered or understand, which is there's too much information.

Yep.

And too much information is just as much a problem as not enough information. Yep. Absolutely. In systems administration, data that is unclean, that has gaps, you can work around it by filling the gaps, or you can work around it by creating models that take into consideration the gaps and have margins of error. But it still has giant gaps missing, right? So you have the problem of not enough information, and in the future, 3 years from now, too much information. Because now we have 27 different systems— Beeline and Workday and, you know, this and the other— and they're all talking to each other. But, um, AI might be really good at doing things like, for example, here's a really fun one that you might like. So, uh, 7 different systems at this very large Fortune 10 company where I was working and, you know, in advising. 7 different systems in the whole HR ecosystem, TA and HR, and 3 different nomenclatures for names.

Sure.

Right? In one system, I was Shally Steckrell. In another system, I was Steckrell, comma, Shally. And in another system, I was Steckerl, comma, A, and my full name. And my full name isn't Charlie. So for example, I don't know if your full name is Alexander or not, but let's say you have Alex Marcus, Marcus Alex, and Marcus J. Alexander, 3 different systems. And matching the information between those was impossible because one system used the Social Security number as a key. The other one used the candidate, right? So even the, even the key point or the data, the key, the data keys are different. Different. So like, oh my God, right? So Agentic might be able to help with that.

I think it certainly can. I mean, the, the use case that we see over and over again is with merging duplicate records, right? So candidates—

because of Shelly and Shelly A in the same system, right?

And, but also the same name but different email addresses. They applied with their Gmail, they applied with their Outlook, they applied with their Yahoo. Even with the same name, I've seen 25 different email addresses, the same person, right? And so you know, you might think— and here's a good example where human discernment is important— you might think that you should just turn AI loose and merging those duplicate records and gather enough information to be able to tell. But there's also a compliance issue there. Do you actually want to do that? Are there compliance implications for putting all of those records together? Right. And that's a conversation that needs to be having your general counsel.

I did not ask that question.

What's that?

I did not ask that question. I just realized I did not ask that question.

That's something I would ask.

Yeah, I did not ask if there's compliance issues because, yeah, well, actually, now that I think about it, there wasn't, but I didn't ask and I should have.

Yeah, well, this is something I learned from Vivian Larson, who joined our team a few months ago. She spent 10 years at ISIMS on their global strategic accounts, and she brought that up to me because I had a very similar conversation like this, and it hadn't occurred to me either. So don't feel bad. Yeah, yeah.

Thankfully, I'm thinking back and going, no, there wasn't, but I can easily see how there would be. Basically disclosure, privacy, you know.

Did I consent here and not here?

That's my point. Yeah, exactly. Consent and privacy and also reporting regulations and what's, you know, what the government requires you to not disclose and what the government requires you to disclose based on where the data is housed is actually really different. Like candidate data, for example, You know, you are allowed to know the gender of your employees if they are employees in your HRIS system for a variety of business reasons and governmental reasons. But in your application process, it's not a good idea. At least not a good idea to let everybody know, right? Like, you might not want to. So, and now all of a sudden you have— and that's just one of them, right? I mean, there's medical and there's so many. I just, I just realized that. But we were literally just trying to correlate. So yeah, glad you picked that up. So that's a really interesting— yeah. And I want to go back into how the hell did you get into this, man? I mean, you were like a program manager way back in the day and program director, I think. And then you did org development. And then somehow you ended up getting bounced into TA. And then you went into founding your own company. And now you're sysadmin. Like, where did Where did, where do things go wrong for a musician from Austin?

Well, if you've ever tried to live on a tour bus with 5 other musicians, uh, it's, it's not pleasant for an introvert like me who would rather— who does— I'm a guy who does spreadsheets for fun to relax, right? Yeah. And so the, the musician's life was not for me. I had a lot of fun doing it, um, but after that, um You know, I, I really enjoyed teaching violin, and so I moved to New York City and did a program called the New York City Teaching Fellows, and they paid for my master's at Pace. And I was a classroom teacher for a few years, and then that turned into an opportunity to be a social work supervisor. I don't have a social work license, but I was supervising a team at a night school program in Brooklyn.

You were a manager, not a social worker.

Yeah, working for a nonprofit where we were helping overage, undercredited kids get their high school degrees at night. I was really interested in— I mean, I really love being a team leader. I was very interested in helping my staff understand their benefits. And HR sent out a communication regarding 3 or 4 different healthcare plans that you could select. And I remember looking at this, and I couldn't make sense. I spent like 5 hours trying to make sense of this. My staff certainly couldn't make sense of this. And so, since I do spreadsheets for fun, I decided to create an app essentially with Excel using VBA that it was essentially an app that asked you for various information and tells you which plan you should choose in natural language. And how much you should put in your HSA or your FSA. Right. And it was super, super fun.

I have lots of questions about that.

Yeah. And I showed it to the HR director, and at that time he was thinking about getting a new applicant tracking system. And we started talking. He started inviting me to these HR meetings where they were bringing in staff from the field. And he said, look, you know, you want to come join HR, you have the interest in people. You're interested in helping people use our systems more effectively, and we're going to do— we're going to implement an ATS. How does that sound? And I didn't know what 90% of the words in that sentence were, what they meant, but I said, how does that sound? Implementing a who?

ATS.

But I was like, sure, I was ready for a new challenge, and I really was excited to move to the central office and kind of see what was going on there and and understanding the big picture of how the organization ran. So that's how it got started. And then in 2020, COVID happened, and they sent us to work from home. And, and that's when I founded this online learning community because I wanted to talk to other people.

Because you were teaching and you're trying to understand. Yeah. Okay. That makes a whole lot of sense. Yeah.

Yeah. Yeah.

And so you, yeah, you used to play— well, I mean, you probably still do, but this was your job is you played fiddle. Yeah, right.

Yeah.

Um, so what is the difference between a fiddle and a violin?

Oh, that's a great question.

I have, I have an answer, but I don't think it's the correct one.

Uh, the person playing it.

That's the difference. Okay.

I mean, I could go so, so, so the more technical answer is a fiddle, um, that music is folk music, and there's a long folk tradition of people playing who did not receive classical training. Structurally, it's the exact same instrument, right? Right.

Um, well, except for one difference, which is what I, I thought— that's at least this is what was explained to me in Nashville, um, at the Rock and Roll Hall of Fame. I forgot the name. And that is that a violin has strings, but a fiddle has strange—

that's a much better answer than the one I gave you. I'm gonna use that instead.

Yeah.

Oh, that's great.

That's great. Yeah. Um, excellent. So yeah, so you've got the creative, you got the teaching bug. I can definitely relate to that. Yeah. And then you ended up, um, you know, kind of in an in-between world of like tech plus teaching plus creative. And that's how you ended up in this world. So, wow. Okay. We talked about the Next Generation Playbook. We talked about the, you know, quantifying the ROI. We talked about community as a force multiplier. What else is out there that people really don't understand or need to understand better, or think they understand but don't really understand, right? About HR systems administration. Like, you know, people say, oh, I'm an HR sysadmin or whatever. And folks go, oh yeah, yeah, I know what that means. But what is it they really don't know?

Just how impactful a person in this role can be. In driving business outcomes.

Okay.

This is not somebody who is just clicking stuff all day long. This is somebody who has their hands deep in your data, and effectively using your data is something that I find a lot of folks in TA aren't totally comfortable with, right? And so they know that they should be generating metrics. Do they get to the place where they are actually driving conversations forward? From an analytical perspective, your best friend is a well-trained system administrator, right? And the extent to which you can have a productive dialog that stays out of the weeds for the most part, that's one thing that's very— so that is a blind spot for a lot of folks in this role because we're techie and we're proud of our deep subject matter expertise. And sometimes if you get us talking on something, we kind of get fixated and can't pump the brakes. Right? So you have to build— I would suggest that TA leaders really get to know them well and build a relationship where they can give you just enough information and you can tell them what you need the system to do and what's important for you. I think there's a partnership here that when it works well can be incredibly fruitful. But often, often that role— and you see it when folks are just like, oh, we bought, you know, we bought the Lamborghini of ATSs and I guess it's going to run itself or Maybe, you know, Joe Recruiter is going to run it in his spare time. No, this is, this is if you invested in sophisticated purpose-built enterprise technology, resource it appropriately and have realistic expectations about what that's going to take, what it's going to cost.

You know, this sounds a lot like sourcing. It sounds a lot like sourcing, a lot of, a lot of the same misconceptions of it. So what role can be? Because, and I preface this by saying that in a lot of ways, the information that we deal with in sourcing is pre-HRIS. It's not in the HRIS necessarily. I mean, it sometimes can be, and there's definitely overlap. But by and large, if you are trying to identify candidates that you don't already know, you don't already have, then they're not in your systems somewhere. So what is the role of modern systems administration in regard to sourcing? How can sourcing leverage the sysadmins?

So it's a huge topic. I would say the thing that should be front of mind is silver medalists. You may not need, you may not need more candidates. You may have entered somebody 3, 6, 9 months ago who was perfectly suited for this other role. And if you are not tagging them appropriately in the system, that is a huge missed opportunity.

Got it.

Yeah.

And the sysadmin can help with that?

Absolutely.

Absolutely. And that's in the, in the ATS?

It's in the ATS, and it's a matter of creating the, the, the, the protocol for tagging them correctly and then surfacing that information through a customized, uh, report that you put in your dashboard and show to the right people at the right time.

So is there like components of kind of like library science in here? Because, you know, one of the things that you, um, when you have, when you have nonfiction there's like, you know, the, the, the ready, the, the ready and proven Dewey Decimal System that catalogs and categorizes the sciences and, and all that. But the entire world of nonfiction, there's really— I mean, basically you go to a bookstore and it's like sort of tentatively by subject matter. It's not even really definitely by subject matter. So how can we categorize information like that for, for sourcing? I mean, is, is that something the sysadmin really kind of comes up with? Like, how are we— what are the, what are the tags? What's the tagging structure? What's the nesting of the tags? Like, you know, the, the, the— what some people might refer to as the taxonomy.

Yeah. I mean, that's something that I think each company needs to determine on their own, but I don't think it needs to be I don't think it needs to be terribly complicated.

No, no, but it's usually missing.

It's usually missing because it's a functionality. I mean, I can speak on a very technical perspective about ISIMS.

A lot of ATSs have the ability to tag, but people don't, you know, they just willy-nilly create random tags. And so you have things in there that have been tagged by one tag that, you know, only one candidate has that one tag. So that really doesn't help.

That's a good call out. So you don't want to deputize encourage any and everybody to create their own tags. I would say you want to standardize them and have a few things rather than 100. And then you— I— when you surface that information, I think it should also be surfaced in a time-based way. So I don't want to see every silver medalist who's been marked as such over the past 10 years. I want to see ones from the past 3 months, past 6 months, who've applied for this category job, maybe in this geographical area, right? So those are all the nuances of reporting that you can do with iCIMS. And then and then put them in the right place, the right user group.

Got it.

Yeah.

Where else can sourcing leverage this role?

Yeah. So iCIMS has a CRM tool and they also have a candidate experience management tool. Those systems are very sophisticated as well. And again, there are some nuances in the way that that information— so CRM was originally a company called Jibe that iCIMS acquired a while ago.

Right. I remember that.

So there are nuances around how that information is passed from CRM to ISIMS and where it ends up. So that's something that your system administration, your system administrator— it's actually a hot topic of conversation in our community, and ISIMS customers are collaborating to figure out the best way to manage those two systems together.

All right, so as a sourcer, it would benefit you to get involved in that, or at least learn from that, because it helps you with how the information gets into the ATS.

Common complaint I hear is, why didn't they just come to me and ask me that question as a system administrator, right? Instead of not asking the question or complaining to their boss or not knowing that you can even ask. Exactly.

Yeah.

If you're— so here's the thing. It's like a lot of our— so we get implementation customers who are going, just starting out with ISIMS, and we get optimization customers. And a lot of times they come to us and they're like, you know, it's a mess. Like, please help. Right. It's a mess because you never realize that you needed to have a competent person in that role, right, to continue to optimize it, enhance it, and report adequately. Right.

Just get a mess because you never put anybody in charge. Yeah. And then you took the eggbeater and ran it around all over the data.

Oh, and you, and you, and you gave user admin access to 20 recruiters, right?

Everybody has admin access.

Everybody's an admin. You know, why not?

Yeah, why not?

Yeah.

Yeah. What could they do?

What could possibly go wrong?

Well, thanks so much for being on the show today. Um, I, everybody that is a guest, especially for their first time, um, not, I'm not just jumping to conclusions and assuming you might want to come back. You may not ever want to come back, but I would love to. Awesome. All right. Good to hear. But everybody does get asked the same thing. And, um, the first of those two questions is who should be our next guest?

Oh, Yeah, yeah. I'm gonna say Brian Sherlock, the CEO of Sync2Hire. Sync2Hire, they're a new company and they're basically building a Slack for recruiters. And I embedded Sync2Hire in our config site so you can embed it directly into iSim so you can have great communication.

Conversation. Okay, got it.

Yeah, exactly. Good.

That's great. I usually don't get like really straight answers. Sometimes like, what? Depend, you know, so I appreciate you giving me that. I might— I'm going to reach out. Of course, they may or may not accept, but, uh, but that's really good. And then the, the, the final kind of wrap-up question, um, which does vary a little bit from time to time, but what I want to do with this is I want people to really have something to think about for themselves. And most folks really do catch the recording of the show as opposed to live. Yeah, we only have about maybe 1,000 people live, but Multiply that times, I don't know how much in recordings, in the podcast distribution. So, I want them to come off listening to this and have something that they think deeply about themselves that they want to— that you think that they should ask themselves and answer themselves. So, what is something related to what we talked about that listeners should think about and chew on for themselves?

And what's the persona here?

Basically anybody that is in the umbrella of talent industry. So talent management, HR, you know, the whole thing, just people out there, you know, what's the one hill that you, that you feel like you're willing to die on and people aren't really thinking enough about and they need to go think about it?

Yeah, I would say AI literacy. Oh, AI literacy. So even if you can't use AI at your current job, you should be using it in your free time and just have fun with it. Just do creative things with it. Just get to know it because there is, you know, there are emotional hurdles to staying up to date with technology that's advancing so quickly and it generates a lot of fear. And I've gone through my own phases of this where sometimes I just have to turn off and I don't want to see one more post about AI.

You're getting left behind or you're not catching up. I know.

I work in tech and I get overwhelmed, right? So, so prepare yourself emotionally by playing around with it now so that you are prepared to use it professionally when it's no longer an option whether or not to do so.

That's really good advice. There's the emotional barriers that we don't think about.

Yes.

So I mean, literally just like pick one thing and, and just play with it and see what it—

Yeah, just make fun images or whatever.

Yeah. Right. Yeah. One thing. Yeah. And then start with that. Yeah. Yeah.

Just, just sign up for ChatGPT and start messing around with it and, and See what it can do.

Okay.

Yeah.

Cool. Uh, so I'm going to put the, uh, systemadmininsights.com is the community up here and also the podcast, which I had to shorten a little bit. So this is the, you can listen to this on Spotify, Spotify 3YQUPGP for listeners. So that's the spotify.fi/3YQUPGP. And of course, if you want to connect with Alex, here's Alex's LinkedIn. Thanks a lot for being a part of the show. Appreciate it. And definitely would love to have you back, especially as these things evolve and mature. Next up, folks, if you want to come back, we have Dart Lindsey. Dart was in the industry many, many years ago and has left TA. He was referred by Jeff Hunter, another character from the mean days of— you think things are changing now? They definitely were changing a lot back then when technology was just starting to take root. And Jeff referred Dart into the show because the work that he's doing now in his own way, he's talking about the crisis of what he calls the working dead, employees that have emotionally checked out. So he's not in recruiting anymore, but he's definitely a product architect in our space. He's going to be on the show next week.

Sounds great. Well, Shelly, thank you. I really enjoyed this and I'm looking forward to coming back.

Yeah, thanks a lot. All right, everybody, be kind to yourselves and we'll see you here next week. All right.

Take care.

Bye-bye. Bye.