Adam Treitler Live on Shally's Alley

Hello and happy New Year 2024! And a big F you to 2023. Man, that sucked. That was just, that was an insult of a year. Um, I mean, happy at, happy at work, but just so much other stuff, just garbage going on. But anyways, 2024 is definitely going to be better. I already can sense it, and I did not say that at the beginning of 2023. So it's definitely good. My crystal ball says that '24 is going to be awesome. And for our very first Shelly's Alley episode, we have a new guest, someone that I have never met before. I think this might even be the first time— I, I know everyone else has been on the show, and some of them have been longtime, uh, friends and collaborators or co-conspirators, depending on which way you look at it. So this is my very first completely alien guest that I've never met before. I was introduced by a mutual friend. You might know Jim Schneider from a company called Recruiting Advisors. Jim has a team of consultants, and one of Jim's senior folks and myself were, you know, out and about talking business, and they were telling me about this guy, Adam. And the name Adam kept getting dropped. By the way, it wasn't just me. They were just dropping Adam bombs all over the place, right? And I'm like, who's this guy Adam? They're like, oh, you know, he does this and that, he's really good at this. And I'm like, really? Let me see. And then eventually I just said, what if, what if I had him on the show? And they thought, oh, you guys would get along really well. So that's how you ended up here, in case you were wondering. But for those of you that are yourselves wondering, uh, you know, who, who is Adam, first of all Here is his, uh, I guess network stuff. There he is. And I'm gonna move our honored guest up to the top of the screen so you can reach out to him and connect there. Um, borrowing from your LinkedIn, like, headline thing, you say you are challenging assumptions and reshaping policy, process, and products in the world of work. Damn, that's a big ask. It is.

Someone's got to do it, right, Shannon?

Someone's got to do it. You know, on my, my, uh, profile it says, um, I, I hate when people say that they're thought leaders. Like, how dare you say you're a thought leader? First of all, you may very well be, and great, but yeah, not to me, right? So I don't like saying that I'm a thought leader. It just irks me. If you want to do that to yourself, go right ahead, but I like to say that I'm a thought provoker. And your bio totally, like, resonates with that because you're, like, challenging assumptions. That very much translates into thought provoker to me, you know? 100%.

I'm always, I'm always asking questions.

So that's me. That's, that's literally all I do, just ask questions. And then the people are like, wow, that's so wise. Like, no, I'm just dumb. But, uh, yeah, you know how that goes. So anyway, um, they tell me that you got the smarts in, uh design principles and automation and things like that. So that's kind of what I want to talk to you about a little bit there, folks. Adam came from ViacomCBS, which you might know as Paramount now. Is that the actual official name? And it keeps changing, right? Yeah, for now. Today it will be like Apple, Warner Bros., Paramount Co., Apple, Taco Bell, Pizza Hut, Paramount, HBO. Twitter after that. So ViacomCBS Paramount Twitter, now Recruiting Advisors with our buddy Jim Schneider. And at Recruiting Advisors, he works on digital transformation and consulting. That sounds very technical. That's because it is, which is what I wanted to talk about today. I— you seem to be very passionate about, uh, design design principles and, uh, smart technology, you know, AI type generative technology, which is really interesting to me that you're interested in generative technology considering that you're an actual published poet. Don't know a lot of those.

Yeah, what's that like?

What's it like to be— I'm a publisher.

I definitely— it's, it's very, it's very interesting because so I, I went to school for, for poetry, uh, my original intention from a degree perspective studied under Marie Howe, who was the state poet laureate of New York. I actually was published as early as high school. I was a big fan of Dylan Thomas, Charles Simic, if people are familiar with his poetry from World War II. Big fan of telling compelling stories with as few words as possible, and that was kind of what I wanted to do from that poetry perspective, was, you know, creative storytelling without having to write— I don't— I didn't want to write a novel. That sounded like too much work. Let me just write like these compelling things with little bits of words and frameworks. And I was also someone who defied frameworks back then too, right? My professor would be like, write a villanelle. And I'm like, no, I'm just going to write something freeform. So we'd get in trouble all the way back then. But what is interesting to your point about that pivot is I still write. I do not write anywhere near as much as I did when I was in college. I pivoted. On that college journey, shall we, from poetry to film, I thought that would be more profitable. I specifically was interested in screenwriting and discovered that was about as profitable as being a professional poet. Um, so that was the introduction to the entertainment industry, which is what led me to my early, early journey at Viacom, uh, through experience, uh, doing talent and casting internships. That was kind of that gateway into recruiting. We can get into that in one second. But from that—

different kind of talent sourcing.

Yeah, very different kind of talent sourcing. I got to— one of the first people I was involved in taking interview notes on was, I think, Joey or something like that from the Jersey Shore. One of the— Joey, Vinny, one of those Jersey boy names. Definitely different than supporting coordination work for, you know, an executive assistant role or even a VP of marketing. But some of the same personalities. But yeah, so I do have that unique perspective. I think from a poetry perspective, I have a lot of friends who are creatives, people who run art galleries, things like that, people who create their own content. And I think we do have some mismatched opinions on AI, right? Because I personally love AI. I use it to write some poetry now. I use it to help me write better poetry, to help me write more effective me creatively and professionally. Um, and I definitely have a number of friends who, you know, are intimidated by it. A number of friends who are like, if they ever use my artwork, I'm going to sue, you know, the life out of ChatGPT. And, and, uh, right, right, right.

But you, but you're using it for what, like, you know, smart people do, which is as a tool to work faster and smarter. And, and yeah, right, exactly. You're not going to it to, to have, to have it provide the answers. Right, so, um, that makes a lot of sense. So, um, you know, being a poet, does that mean that like when you talk to candidates you always rhyme?

Not anymore, no. You stopped in the early days? It kind of caught them off guard, Shelly.

You stopped doing that when you realize that might, might be getting annoying?

Exactly, exactly. I actually had a brief period of time when I was at Viacom, uh, where I used to— I became responsible. So in my recruiting journey, and again, the theme, I think, that transitioned from poetry was just it really helped me convey big ideas in as few words as possible. My challenge early on in my career is I didn't actually understand the concepts. I didn't understand how business worked. So my profound statements often fell on confused ears because I didn't know the ins and outs yet. Uh, but as I learned recruiting, I think another way we align is second I got in as a recruiting assistant and I had zero experience with recruiting at this large media and entertainment conglomerate, I just started poking around the systems. I was like, how does this work? Why does this work this way?

Asking questions, right? Like, there you go.

Exactly.

So you're like me, you ask a lot of questions and then you synthesize all of that and summarize it to like a bullet point that people can understand. And that's, you know, that, that's basically you're the human ChatGPT.

To some degree. I remember back then it was really funny because I had 3 different managers over my time who were always like, if we could clone Adam. I'm like, well, the closest thing you're going to get now is maybe like an Adam GPT. Right.

Yeah. Read these 27 research papers and write me 3 bullet points.

Exactly. But yeah, at the time it was kind of funny because being in that mindset, Shelly, I asked a lot of questions and I could tell I was generating some friction with the manager of talent acquisition that I was reporting into. We had a solid personal relationship, but she was like, why are you asking all the questions? Just do what you're told. Just schedule these meetings. Just get these people in here. Take these interview notes, you know, like, um, get the feedback to who it needs to, uh, all this fun stuff. Uh, and I'm like, well, why do we do it this way? Why do we do it that way?

You know what, you asked all those questions, but I think the reason they got irritated is because they didn't have the answers.

In hindsight, 100%. At the time, I did not realize that at all, but it was exactly that. It was 100% that they did not They couldn't answer it. They were like, this is just the way we do things. I don't know why we click these buttons. I did not make the decision.

Just do it. Stop asking. Just do it. Get it done. Exactly.

They were not the decision maker. They were not the policymaker. And so what's interesting is that journey actually led me in the first 3 to 4 months of my time there into the office of this new department that was forming, this talent acquisition operations department. And I was thrown into this manager's office and I thought I was getting fired. I thought I was like coming in there because like I thought she was going to tell me they had recently shifted the operation. So the coordinator supported specific recruiters. So I had like my manager who I was supporting, but on paper I reported into this operations person. And so when I came into her office, I was like, oh, I'm in trouble. Like, I don't come into her office unless like, you know, something big is going on. She just signs my paychecks. And she was like, "I heard you've been asking a lot of questions lately." That was the whole situation. Shall we like, "Oh no!" Should I say yes? Or should I be like, "No, I don't ask questions. Never, never, never me." So, you know, she's like, "I heard you've been asking a lot of questions about the ATS and you've been asking questions about data and reporting." And I'm like, "Oh God, is that, you know, should I stop doing that?" And she's like, "No, no, no. We have no one else in the entire talent acquisition department who's asking these questions. Would you want to?" split your work. So maybe 20, 30 hours a week are recruiting, recruiting coordination, and the other 10 to 20 are supporting this operations work and helping us build out a reporting function. I was like, let's do it.

Try to figure out how to get answers to these questions.

Exactly, exactly. So finally a sponsor, someone who's actually going to help me get answers to these questions. Uh, and that kind of just led— that, that was the springboard for my, my career trajectory. So like, just to give someone a quick shout out, uh, coming— Lee was that manager, which She was my manager for quite a bit of time at Viacom. She really kind of laid the foundation for, you can do recruiting, you can do the recruiting stuff, and you can actually think tactically, strategically. You can figure out how to use tools, right? You can figure out how to use systems more efficiently. It's not just about the recruiting aspect and it's definitely not just about siloing it too, right? It's not just about attracting talent and then leaving them. Yeah.

And then abandoning them.

Really that kind of—

Attract, abandon, attract, abandon. Hey, welcome. Bye.

Exactly.

I mean, that's— so how did that lead to, to design principles?

Sure. So one of the things that I just noticed, right, because I got really hands-on at the time we were using iSims and a few other tools, but iSims was the core application. I became an administrator, certified administrator, 6 months into my time at Viacom. A kid who knew nothing about recruiting, I became a certified administrator of an applicant tracking system about 6 months in. And I just kept tinkering. I kept trying to figure out what's the art of the possible. Well, we have 25 offer letters for the United States in the system. Why? Why are there 25 different forms? Why are people moving around? And I just started asking questions and trying to figure out answers, right? Shelly. So I was like, all right, well, if you consolidate this variable here, if you make this dynamic, if you add some logic here, we could have 2 offer letters, one for interns and then one for staff that dynamically populate. Instead of having a non-exempt and an exempt letter, we could use logic to say if the role is non-exempt, populate this language in the same letter as if the role is exempt.

Yep.

No one had thought about that before. No one was thinking in this way.

Yeah, no, that, that's, that's, that's, uh, that's smart because otherwise you have to like choose the particular template and there's 27 templates. But something I ran into that I think you would appreciate as I'm trying to, uh, automate the offer letter using what you described as like, you know, select the choice and then insert the text. There are 27 different states with different rules that are so widely different that it's not just a sentence or two, it's like add 7 paragraphs and don't have this other paragraph that's not valid in these states. And it gets so complicated to do that.

I mean, hey, if you want to connect offline, Shelly, I've built that in SAP SuccessFactors and ISIMS. We've, we've, I've, I've, I've, I've architected those exact solutions where we've had—

Yeah, but you know, so you know what I'm talking about. It's like so many, you know, you got to select the state and that kicks off adding and removing all this text from— So the UI part of it, or the, uh, sorry, not UI, but, um, it's, it's UI/UX.

Like UX design of it, really.

The user experience of it goes from, oh my God, this is overwhelming, I need a— I need to take a course on it, to like, ah, I can just do this.

Exactly, exactly. Like, my passion became making things as easy as possible. And again, thank you to— I'm a big gracious person, grateful person. I'm really thankful for the, you know, support that I had both at ViacomCBS and at Twitter. Matt's here, so I have to pretend I like him. But, you know, I learned a lot through their sponsorship of learning, and I went in and I studied machine learning. I studied data science before before these were buzzwords, 'cause I just thought it was a cool way of processing information. I was like, what are these frameworks? How do we do this? And through that I was like, all right, well now I have a strong sense of analytics. I have a strong sense, and I think you even talked about this on your last podcast, plug for that. Like I have a strong sense of analytics, but I don't have an understanding of how to convey this information effectively and how to get people to take action. 'Cause everyone was saying that. We brought Vizier into play, like we brought all these cool vendors into play and we were doing all this amazing stuff. And no one was going to the data. Like I'm skipping entire segments of my career, but like, right when we got into this people analytics phase, when we started to look at human capital information, we started to look at recruiting information. People A were still in that stage of find me facts that back my opinions, right? Give me a stat that backs my opinion, not find me problems, not, you know, use the data to inform strategy. It was, you know, I want to know, a stat that makes me look good as a leader in this space. And it's because they didn't know how to use that information. But that's kind of that thread that I started to tug on, and that's what led me down this UX road, right? It was, okay, what if we make things so easy for people to use that they just kind of come to these conclusions naturally on their own? They navigate through this process where it becomes obvious where the roadblocks are. It's no longer a question of, is the roadblock the ATS is really clunky and hard to use, or is the roadblock my messaging to candidates suck.

Like, you know, like, isn't it amazing how very quickly and easily these, these things in TA get overcomplicated? Like, it starts out as, we should have an application. What are we going to ask on the application? Well, let's ask for their first name, last name, phone number, and email, and then they can upload their resume. Okay, great. And then it's like, uh, we also have to ask about Diversity, equity, and inclusion. We also have to ask about disability. Oh, now we have to ask if they have a non-compete. Oh, now we have to ask if there's any, you know, if they're military, or now we have to ask about not just the gender you were assigned at birth, but the gender that you are identifying with now. And so before you know it, first name, last name, phone, and email is now 57 questions. And we did that.

Yeah, no, no, it's been a continuous evolution of, I think, you know, not asking the right questions and kind of getting sometimes too many cooks in the kitchen or sometimes not enough, right? It can be both sides of that coin where you have one person who's like, well, I did this at this company, so let's just ask these questions. Let's see if it'll work at that company.

Why? I don't know.

That's just the way we do it. One thing I can tell you, and I know a lot of people in the TA space, you know, I love employment lawyers. They're brilliant people. But when you bring too many of them in and you don't give them enough context, they're just gonna end up constraining your entire process, right? Their immediate mindset is going to be risk aversion. Yeah. And so if you cannot explain why you do things, they're just gonna say don't do it. And that's a big problem in TA 'cause there's a lot of, I don't really know why we do this. So the second you walk into a room and you're talking to a lawyer and you're like, you know, I wanna put this on the website and they're like, why? I'm like, 'cause it looks cool. And they're gonna be like, don't do it.

Let's not do it, right?

But when you have data to back it up, like, well, we've seen competitors do it. If we put this content here, it's gonna drive more traction, I've looked at the local laws and things like that, and my understanding is, you know, here are some articles I've read on it. Like, you know, I want your legal expertise to come in and confirm it, but now you're giving them some materials. They're not just like flying by the seat of their pants, treating you like a quick client where they need to get to something else. They now have some of that context. And I just say all this because, again, this is what kind of gravitated me towards those UX principles, right? Like, you have sort of these really 4 high-level principles of human-computer interaction, or design, which are linguistics, physical heuristics, navigation, and conceptual heuristics. And a lot of people don't think about this. One of the things that gets heavily overlooked is concept. But when we talk about, you know, step 1 is just linguistics. Are we using the right language on our pages, in our systems?

Oh my God.

For the end user to understand. We have an ATS, for example.

Terminologies are so all over the place and definitions of terminologies. Like, is it a rec, a job, or an opening or a position, rec job opening position.

Is it exactly— and are people using that language consistently? Because what's— what a position is to, you know, your administrator in Workday, and what an opening is to a recruiter, those are not positions. Exactly. Like, they literally are different objects, which makes managing the data even more complicated. So 100%, I mean, you have these entire data architectures where people think that they're cool because— or hip because they came up with a new way to call something. And it's like, no, like one of the things that Matt and I recently settled on is like one of the concepts in kind of core programming is parent-child. If you're going to have these relationships between jobs where one is over another one, we're just going to use the term parent and child, right? Because trying to create this, like, it's an evergreen that feeds into a hiring rec that feeds into, you know, it just gets too complicated and people forget and people start thinking that, you know, an evergreen rec is something that you, you put on your dashboard. Like, um, they get lost. So like the language is fundamental. That is the core.

So the 4 basic principles of applying basically the design, the 4 principles of design applied to recruiting, right? Number 1, language. So pick the— don't define it, pick the language that is most commonly defined that way and try to use the simplest possible name.

Speak the language of the end user.

Speak plainly. And in the language of the end user. Okay? Don't reinvent it. Don't try to get fancy.

If it's an ATS, if it's a career site, make sure that they know where they're clicking on the career site. Make sure that they know what the buttons do. If it says something on it that no one is gonna understand what that means or where it's gonna take them, you know, and I feel like so many people overthink this in the TA space. They don't think about it. They're just like, we want this carousel here. We want this over here. And that kind of leads into, into one of the next ones. We'll go with the physical heuristics. Physical heuristics are how people interact with the world. And in the case of an application, it's just how people are used to clicking things, right? So physical heuristics in the context of like Zoom, for example, would be the location of the camera button. If I create a new app and I put the camera button in the upper right-hand corner, it's in a completely different space. I'm gonna have no idea from a physical, from my experiences, where to go to click that. And all these ATSs try to do different things and they expect people to understand, well, why did this one have me upload my resume here?

Like, you're just used to always having the search box in the top middle.

Exactly.

Now the search box is in the bottom right and nobody finds it.

Exactly. So physical, when it comes to like an application, is like the placement of things on the screen, um, or like, you know, the way—

wait, time out. Who decided that? Who, who decided what the principles are? Like, who decided that the buttons should go on the bottom? Like, where did the common language of, uh, physical design originate? Because let me put it this way, there's no way the ATS has invented this stuff, right?

It's technology. Not at all.

Where did it come from?

So that's a big thing that I, that I, that I write a bit about and that I talk about, is that there are all— there's all this research and there's all this rich information and all these studies that literally look at things like Fitts' Law, like where people look on the screen, where people most commonly click. These design principles are what inform a lot of B2C technologies. Why does Netflix work the way it does? It's not perfect. We all have our own personal gripes with it because we all have our own different understanding of the world.

They just changed everything and moved it all to the top, by the way.

Exactly. But most people can figure out how to use it, right? They can figure it out intuitively. ATSs, things like that, people don't always figure out as intuitively. For a variety of reasons, something I'm continuing to tease out and have conversations with people about. But in terms of why does the button go here, why does it look the way it is, it's a combination of psychological, sociological, and, you know, technical research. There's research—

what you're saying is it's basically kind of hardwired into most people of some kind of common culture.

Exactly, exactly. And so if we continue to kind of weave the narrative, shall we, right? So Going from the physical, and we can come back to it, to the navigational. Navigational is literally on a website navigate. Okay, I did step 1, now I'm at step 2. Does that make sense? Right? This is to your point of all those, those questions. You see lots of different ATSs that, that have different steps in the application process. Does it make sense that I'm filling out this here and then on the next page I'm filling out something completely unrelated? Can there be breaks in my navigation journey? A lot of the times with hiring managers, I see this done really poorly. Right. They have no idea what they're supposed to do after they do something. They fill out interview feedback and they send it and there's no communication. There's what, what am I supposed to do next? Like, why haven't we hired Paul? He was great. You know, like they don't know what to do next. So that design—

Yeah. Like seriously, I've had this example where they complete the scorecard, they finish the scorecard and save it, but they don't submit it. So then they get a reminder.

It's a design problem.

And yeah, and then they say they get a reminder from the system, hey, you have not submitted your scorecard. They're like, yeah, I did, I filled it out yesterday. And then they get another reminder, hey, you have not completed— they're like, what do you mean I didn't submit? It's like, you got to go back in there and open the one that you saved and click submit.

And that's a great example of both a physical heuristic and a navigational heuristic. It's also an example of not a great design within the system, which oftentimes is because we're applying to these constraints, right?

So you have this challenge, there's some sort of legal reason why you have to save it. Before you submit it, or you can't edit it after you submit it, whatever.

Yeah, exactly. Whether it's a legal reason or a technical reason, sometimes it's a combination of both. But for one, if the root— or if the user doesn't understand it, that's a design problem. Um, that could be a physical problem of where the button's located. It could be a navigational problem of it's not clear what they're supposed to do. They click this and then on the next page, what am I supposed to do? Um, then from the beyond— and navigational also applies to processes, right, which is a whole thing to go down in the ATS space. But if we think of workflows, right, does the way a workflow goes from when someone first applies to when you hire them, does that make sense? Something that Recruiting Advisor specializes in is making sure that those workflows make sense, that there is a strong navigational flow, that you don't get stuck in a step, that a recruiter doesn't leave a candidate somewhere, right? Um, then the big one, and it's funny because like these first 3 would be good enough to apply to recruiting because enough people, like enough people don't think about them. Conceptual heuristics. This goes to what you said earlier, and I think there were some comments on this already too. Worldview, experience, concepts. What do people just naturally understand, right? What are the metaphors that we live our lives by, right? If I see a certain word, it's going to evoke a certain reaction in me, right? Do we understand those conceptual heuristics, those cultural things, those elements of just kind of how people interact with the world on a daily basis? Is that feeding our linguistic navigational and physical heuristics, right? Is that— now we've built a beautiful website. You get from page to page. Do people know why they're doing it? Do people know why they're applying to a job, right? Do people know why I have to complete this application? That's the conceptual piece. It's the why. It's like, why do I have to go through all these steps to finish this job application?

Well, in the case of a job application, I know recruiting is more than just a job application, and you're not talking about just ATSs. I mean, you know, ATS, CRM, HRIS, and advertising and all that, even LinkedIn, which I think has a slew of design flaws. What you're talking about here is the I almost want to say the semiotics, right? The kind of the base language. If you took language out, would you be able to figure out what it does by the shape and look of things even if you couldn't read or couldn't read that language, right? And is that really common across— I mean, how is that— Take for example, there was a comment on this here. Here, I'll put it back up on the screen here. Top to bottom versus left to right. And yeah, so that's part of it, right? But beyond that, things like, for example, color, okay? In the study of semiotics, which is a sort of passion or hobby of mine, this is one of the examples that they use commonly to illustrate the differences in language versus symbolism versus semiotics. You know, language is written or spoken. Symbolism is more visual, metaphorical. Anyways, so red, the color red. When you use red in certain cultures, Western cultures for example, it signifies attention or danger. Stop signs are red to warn you and grab, you know, first grab your attention, hey, it's red, and then warn you, right? But in other cultures, in some Asian cultures, it is, uh, it means something completely different. It means fortune or luck. A red sign might be a, "Come here," not a, "Stay away." How do you reconcile that?

Sure. There's two angles to this. One, I love this because it may not sound like it, this is a great transition into AI. Actually a great transition into AI because with machine learning models and things like that, applications and platforms can offer more dynamic experiences. Not only that, we can also address the gaps in our knowledge. So to your question, say I have a job, a career site that advertises to people in China and also advertises to people in the US. And I don't wanna build different portals for these things. And I just want kind of this one space. You could have dynamic user experiences that recognizes, are you in China?

Are you in US?

If you're in China, use this color scheme to evoke a better understanding, use these words, right? That's actually another huge problem. I love John's question because another huge problem is a lot of companies will outsource their translation. I saw this at Viacom when they acquired Telefe. Luckily, we had a lot of native Spanish speakers, but there was this whole effort to translate lots of elements of SAP, which is a huge system, translate these things into Spanish. And they would get stuck on things. They would— translating things like time to fill into Spanish, as Matt said in the comment section as well, time to fill is already a nebulous concept. People get stuck on it. Now try explaining that to someone in Spanish. It may not directly translate. Maybe there's even different cultures within the people analytics space and the understanding of statistics. So how do you personalize that? And I think AI starts to give us a leg up both from the question of what is the more colloquial way of approaching this? What is the basic concept? What would someone in Argentina understand with an average education and an average background? What would their experience look like?

Sure, but now I get that, but now you're talking about language localization.

Yeah, because I think that's an extension of the color and the placement on the screen, etc. Like, I think all of these pieces are, are interrelated. There's, there's various degrees of effort required to address them. Um, but if we talk about where the button should be on the screen, there are some—

that's more like— that's what I was asking. That's— is there a universal language for where the button should go? And if so, where did that come from? I mean, was it like the first website had the button there? Therefore, every website after that.

So there's a really interesting, uh, to turn people onto this because it's, it's a neglected space of research. There's a really cool, uh, area of research called conceptual metaphor theory. Um, I think I want to say they're from Stanford, but I could be completely misrepresenting that. But, um, Stanford, Harvard, one of those big universities, they did this study to look at, um, how do people interact with the world, right? People of different cultures, people with different backgrounds. And they discovered that almost no matter what your background was, for example, uh, on a screen, if you put something at the top, people think it's more important than if you put it at the bottom, right?

Um, if no matter what your language is— okay, yeah, all right, yeah, because there's— there are left and right and top to bottom, but still top to bottom, still the heading would be on the top and the left. Yeah, I got you.

There are certain universal concepts that just humans—

such as like the title goes here, right?

Okay, exactly, exactly. We think that whatever is directly in front of us, what's at the top, is more important. They did this whole study where they would take your picture. Let's just use Recruiting Advisors, for example. They put me at the top and they put Jim at the bottom and they say Jim's title is, you know, founder of Recruiting Advisors, CEO, bigwig. Adam is little guy. I'm at the top. People are going to think I'm more important. If you quiz people afterwards, they will think that Adam was the more important person, even though you had very clear language. You, you, you specified, you know, Adam was the little guy, right?

You did the reverse order of like junior to senior. And very clearly explained it, the progression, they thought the other—

okay, people will still think that I was about size.

If you're picking— yes, yes, they did the same, they did some research and they actually found a dimension on top and you're in the bottom, but yours is like page wide.

And, and there's like interesting thresholds of almost like annoyance. It's like if something's too large, too big. Okay, so, so, so there is, there is research in the space. To your point, I think that, uh, it's a space that I want to learn more about You know, I'm not going to pretend that I'm a deep expert in the multicultural aspect of—

So like, okay, so there's wrong place, too big or too small, it doesn't mean anything to me, and hold on, so I'm doing what you do, right? So I don't get the language, I don't understand, right? I don't understand. Um, it's not in the right place, it doesn't make sense in the flow, and I don't understand why I have to do it, right? Those are the 4 things.

Yes.

Why do I need to do this? I don't get the language. Where's— why is this button here? And why is this the next step that I need to take? Okay.

Exactly. And I think also like the why really just falls in the conceptual piece. 'Cause like if we talk about one of the things I just shared, the $1 trillion—

Right, that was all 4, sorry.

UX problem. Yeah. The $1 trillion UX problem, right? That Amazon released statistics on recently. Companies lose 35% of their customer base because they have poor user experiences on their websites. No one's done the research to see what that looks like internally at companies. You know how many recruiters are doing things other ways? 'Cause they don't know where to click in the ATS and they get so frustrated.

Yeah.

They don't report it.

Most.

Exactly.

They don't even think they can change it.

I was gonna say, they don't even ask why. Like when it comes to the linguistic conceptual, not the conceptual, the linguistic, the navigational and the physical, they don't ask why. They just say, this is annoying. I don't understand it. It's not working. I have a job to do. They go straight to the conceptual. My objective is to put butts in seats. I am not going to follow the instructions you gave me. I am not going to read your training documentation because I can still do this probably in a non-compliant way. I can still do this without your tools, without the process flow that you've given me, because it's too hard. I don't get it. I don't want to do it. And that's a big problem for organizations to solve. And that's where kind of like, A, having recruiters that understand these concepts. So even if something is frustrating, you understand. You know, like you have to do it. That's one way of doing it. But is there a way of improving the experiences by applying design thinking and systems thinking? Is there a way of improving access to information? Is there a way to configure systems more intuitively? And I think another challenge and like the direction I see you leaning is, and I fall into this camp a lot too, Shelley, the answer does not have to be perfect, right? It does not have to be perfect.

Right. Yeah.

It has to be less— It has to be an increment and improve.

Or better than, yeah. Gotcha. Right. How do you solve for the tension between the powerful forces that, that, you know, you mentioned legal, and somebody on the comment here— here, I'm going to put this— where did it go? Here we go. It was John who said the lawyer mindset doesn't work in HR. They're partners, not drivers. So I understand that. I understand that the lawyers are solving a problem. But there are, you know, there's the chief, counsel, there's the CFO, there's this marketing. So, you know, marketing wants to drive brand, finance wants to drive clarity of, you know, codification so that they can have common financial structures, and legal wants to drive risk reduction, right? So how do you reconcile all these forces? Like, for example, asking for a friend, There's an organization that has— you might need a pen to write this down— there's an organization that has a requisition code, an opening ID, a position code, and a job code. So you got rec, opening, position, job, different numbers.

I think, I think the hardest challenge—

are you having an aneurysm right now?

To some degree, but I've experienced it before, so it's like, oh my God, no, I've— exactly, no, I've 100% worked with clients like this. I think it's a huge challenge in the RPO space, oftentimes massive challenge in the RPO space, right?

But this is one entity, I'm just saying. And the reason for the numbers are, you know, recruiting needs these, marketing needs that, finance needs that, and legal needs that.

So I think, again, this is, this is one of the harder, harder questions to answer because it's like You and I, I think in particular, and I—

I'm sorry, did you think this was going to be easy being on Shelly's Alley?

Easy Shelly. No, I think the hard part is reframing the mindset, right? It's, it's you and I would love to be able to solve these problems on our own. And I think one of the hardest challenges is kind of acknowledging we can't. It's not just us. Like, this does become a team effort. And I do think to some degree it's figuring out the levers of influence, right? It's who influences who. right? Who is going to effectively persuade this different party? If it's not gonna be me, right? Like step one kind of in that space is do I have the knowledge to persuade this person? Like, do I know what they need to hear to persuade them? If I don't, do I know someone who can persuade?

Or is it too late now? It's kind of baked in. Like, yeah, go back and fix it.

Exactly. And I think, I mean, one of the, one of the things is finding those levers, right? Like one of the, one of the best levers for better or for worse. And one of the reasons I like working at Recruiting Advisors is we're usually coming in when companies are doing implementations. This is a fresh opportunity where a company is implementing a new system. And while I think there may be better ways to work in a lot of these organizations and better ways to kind of order the operations from when they make these decisions and how they decide to implement their tools versus their processes and their policies, because they don't always do them in the same order. A lot of organizations will do tool selection Okay, we picked our HR system. And then now what should our process be?

Right.

Also, we need some policies. Oops, I forgot about that. Or we need to update our policies now that we're— we moved to the cloud. Like, so to your point, doing things in the right order, asking the right people, getting them in the room early to be key decision makers, I think is really, really helpful because otherwise legal is just acting from that protective space. If they're brought in earlier, if things are explained to them, and oftentimes if they're shown to them, This was something, again, to kind of loosely pivot to like the AI space. When I did some of the early automation work that I did at Viacom, when I first pitched the use of Zapier at Viacom, and I first pitched the use of some really basic machine learning models, it was immediately shut down. Why? People didn't know what this stuff was. They couldn't understand it. It was exactly control, mitigate risk, stuff's working fine. I had weird use cases that I didn't have clear objectives. I didn't have clear measurable outcomes. Took me 2 years to then finally put together a deck where I showed how using Zapier to automate conversion hires from—

Vendor to full-time.

Exactly, exactly. Which was a very high-volume operation. There were oftentimes thousands in a year at Viacom. I showed them how much time it could save us. I showed them how it could be built. I built a proof of concept. Boom. Sold, implemented within 3 months. Like, took me 2 years to figure that out, but they would have done it 2 years prior probably had I understood that, had I been able to demonstrate, make the case, okay, make a case literally.

So I'm sorry, but I'm gonna have to revise your former statement that you have— there's 4 design principles: linguistics, physical, navigational, and conceptual. I'm gonna have to add a 5th one to that. Because without this, the other 4 aren't possible. So this may be an antecedent to those 4, which is architecture, like purpose for this. What is this supposed— what is this whole thing supposed to do? Which what it seems like you're saying, and maybe this is not what you're saying, so confirm. What it seems like you're saying is sometimes people choose the tool and then create a purpose around it.

Yeah, yeah, yeah.

And again, like, it's like, well, we invented this thing, we don't really know what it is, it goes around in circles, uh, maybe we can put a car on it and call it a wheel. No, it's like they invented the wheel to make the car run faster because the purpose was to make the car, or whatever, the, the stone slab run faster. It sounds like what you're saying is people are just going around inventing wheels after the fact.

It's, it's the classic, it's the classic, uh, uh, fake Henry Ford quote of, you know, If Henry Ford hadn't invented the car, people would have just asked for faster horses, right? They would have just, you know, they literally would have just been injecting their horses with ketamine or whatever. Like, that would have been how they got to their destination faster.

Well, we did breed faster horses, so that kind of happened.

And that was how they tried to solve that problem. And I think there are a lot of spaces in the B2B space, even outside of HR, but it's definitely very present in talent acquisition and HR, where people are trying to make faster horses. They're not thinking about Can we build a car? It's one of the bigger challenges I think that we're going to face with ChatGPT and all these large language models and integrating them with other AI platforms. People are not thinking about what is our actual problem, to your point. What are we trying to solve for? What is that foundational challenge? And what are the available tools? What are the available people? What are the available skills? Like, we didn't touch on that at all, Shelley. What are the available skills If we don't have that availability, organizations really need to shift from that sourcing mindset. Sourcers will find you purple unicorns, but there are some purple unicorns that do not exist. There are companies who want someone with 10 years of experience with OpenAI. It hasn't been around that long. You have these challenges of how do you address these gaps? I think to me, one of the biggest things is upskilling is critical. There's so many ways to address these broader User experience problems that we talked about. That's another space where one of the things I'm playing with, right, is in the GPT space with Recruiting Advisors Labs. I'm building things right now in our own proprietary instance because there's always unease, as we talked about, and bureaucracy in organizations to adopting these things. Using natural language to query an ATS. What if I could say, pull up, you know, the 5 most recent candidates for Job A. And all I had to do was type that. And I pulled up the 5 candidates that I wanted.

But yeah, I get that bothers me. You know why? It takes too long. I mean, I understand what you're saying. But what are the top 5 candidates in product? You know what, what you could just as easily write top 5 candidates. I mean, like, the rest of that, what are the, is fluff. We use that when we communicate verbally. Because there's that— sometimes there's that need for conversation, for politeness, for due process, for protocol, and so on. But if you're talking to a computer, you don't need to be polite. I mean, sure, you can be, but you don't have to say, "Excuse me, do you mind if I bother you for a moment? What are the top 5 candidates?" You can say, "Top 5 candidates." You know, like, it's a really, really great critique.

And I think actually, like, in hindsight, I almost violated my own principles, and I don't even think it's a great example. Let's, let's, let's, let's. Okay, rewind.

Let's rewind.

Let's pull it back. And honestly, no, it's a great point. And this is why I think we need conversations. There's lots of. You can do this. And I build these things going on. Yeah, do it. But to your point, should we do it? Sometimes I get ahead of myself with the technology. One of the use cases that we built that I think might be more interesting is, you know, okay, you find a candidate in your ats, right? You want to know more information about them, and this is what tools like Intello, et cetera, have tried to do. What if you could quickly update the information on this candidate? And what if you could quickly get a summary of their experiences that you can tailor as an expert still to share with a hiring manager, but you don't have to spend all the time combing through their LinkedIn, et cetera, combing through their different sources just to create a compelling narrative for this person. You're given a narrative that you can restructure. That's an example of a way you can use ChatGPT. You find a candidate through the course of your natural work as a sourcer, and you want a summary of them. You wanna figure out what are the best ways, and you could start measuring some of this too in the future, right? What are the best ways to reach out to different hiring managers, uh, to share candidates with them? They all have their different preferences. Some of them want a massive resume Bible. Some of them, you know, just want the top 3 candidates. Some of them need to be persuaded that they should hire people who didn't go to Wharton. Um, you know, you have these challenges and, and I think AI can help with developing some of those compelling narratives. Another space is, again, there are areas where I think it's gonna take time to figure out what it looks like, but where natural language can come into play. For example, oftentimes when kicking off a job, people have conversations back and forth, right? There's, there's a kickoff call, something like that. What if the items that you talked about could be structured into the bones of a job requisition in your system? Would not be the final job requisition at all. Like, the stuff that you guys talk about probably is not what the final form is gonna look like. But now you don't have to distill all this information, write down disparate notes across the HR business partner, the hiring manager, and the recruiter. You now have those notes in a data structure in the bare bones of this requisition. And now all you have to do is kind of compile it. Oh, this field is wrong. This is not actually the functional area, right? This is, this is a great starting point for a job description. Let's revamp this, right? You get that, you get that, architecture to some degree in that experience foundationally set so you don't have to do all this work to rediscover information that was already shared, to hope that your interpretation of things was correct.

The universality, if I may invent a new word, of these models, these large language models, Actually, as you were talking, this occurred to me, maybe you can use this question. Models like ChatGPT and Bard and Claude and all that, are they only in English?

This is a big thing. I think John Kelly's here and I can't remember if it was him or another person who presented this information. They're not only in English, but they were heavily trained in English, which means that the way they present other language is more like Google Translate. It's not always culturally sensitive. That is a big problem. Bias is a huge problem with these models. It's a huge risk. That's why I think a lot of organizations are at least exploring how to build their own proprietary models to feed out the bias.

Yeah. Well, I didn't want to go into bias in a negative sense, but more like my question was, I'll revise, the universality of these models, means that you're actually using a more common sense approach. In other words, one of the things that you talked about was simplicity, plain and simple language. The fact that ChatGPT uses the aggregates of all this knowledge probably helps to make it more plain and simple because it's using it literally catering to lowest common denominator. I don't mean that in an offensive way, but it's the minimum level of understanding, right? So maybe the universality of it is actually a benefit. Now you're talking about bias, which is not, but that's kind of where I'm at.

Yeah, yeah, yeah. No, and again, it's a whole new world, right? But no, 100%, if we don't go down that rabbit hole in terms of what it can deliver right now, I think that's a great example, right? It does, it makes things simple, right? We had Matt say, why waste time say lot word when few word do trick? That was somewhat of an ode to the notion of being able to say less or do less and still get to the creation of a job and move that forward. But it also applies to the same thing of if you don't know how to talk to a particular stakeholder, if you're struggling to get them to react to you, or you have to write a lot of copy in your day-to-day job, these are great tools, right? If you're struggling to find research on a particular topic, there are tools like Perplexity that specialize in research. That can help you find information, find relevant statistics, may shift your own bias about something.

Okay. Processing, processing. All right. I got you. Okay. So I still stand with my original revision of you need 5 principles in recruiting, but I don't really know what to call that first one because it's not architecture then, it's purpose. Did that, uh, goal? I don't know, I have to think about that one. But all right, so once you know what the heck it is that you— first of all, what are you— what do you want to do, right? Okay, then use plain and simple language. Linguistics component of it is make it— make sure that the user recognizes it. The physical component is make sure that the user knows where to expect things: buttons, sections, search bar, you know. Title goes here, things like that, which you would think is common sense, but apparently not. Then you have the navigation, which is what is supposed to do, you know, when I click here, what's supposed to happen next? And that's where you run into a lot of complaints that recruiters have about their systems is way too many clicks to get anything done. And I don't know if you were one of these, you know, complainers about iSims, but iSims notoriously, I heard from a lot of folks, oh, it takes 27 clicks to get one thing done. And so maybe the language is simple and the physicality of it, everything's in the right place, but you still have to click too many things, right?

Exactly, exactly. And I think there's a lot of overlap with the heuristics. And I really think that like if you kind of peel back the onion a little bit too, the conceptual heuristic does bleed into this foundation. Like we framed it as just the why, but it really is kind of that whole, it is the concept of the person understanding the person. Purpose. And again, when I talk about these heuristics too, these heuristics are from the perspective of the end user. There are other heuristics you can apply to actually building something. But the foundation of the model should consider the intent potentially of the end user, but what you foundationally establish and then what you build on top of may be distinct from that. Which is why there are these different layers.

Hold on a second. Are you telling me that the principles of design are not centered around the user? In other words, are you saying that—

I'm saying these are human-centered principles. I'm saying that the—

The end user.

Yes, correct. The person who's actually going to be using the product, that is what you want to consider when you think about the language, um, but dude, I mean, the candidates are the ones that are going to be using the ATS the most because they're going to be more people applying for jobs.

However, the recruiters are going to be the ones using it the most in regards of time. So like, a single person is only going to use the ATS for that 1 hour that they're a job candidate, but the recruiter is going to be in there for 8 hours. So it's kind of like, you know, Is it total time in use or number of individuals?

This is what you need to be targeted. It's a huge design challenge. It's one of the reasons I think you see some of the other research I've shared recently is about feature bloat. I think Gartner published a study about feature bloat in SAP and Oracle. They just keep churning out features because more things for people to buy.

I've been a consultant for most of my career, 20— I don't even want to— more than, let's just say 20 plus because it's getting to 30. And I can tell you that scope creep is literally the number one killer of consultants. Most consultants pass away due to heart failure and stress because of scope creep. So you're telling me that, you know, feature bloat or whatever, it's literally the same thing. Oh, that's cool, but can you do this? Yes, fine. Okay. Oh wow, that's really nice. Now can I do this?

Like, ah, and this is, this is exactly why you need your customers, uh, you know, from a consulting perspective, but also just from like, if you're an internal HR practitioner, sourcer, recruiter, leader, you know, you need to think about what, what does the purpose, which I think you laid out, that is, that is number one, right? This intent, this purpose, this foundation. And you need to not be constrained by your tools. That's why what we talked about earlier is the first thing you pick is your tool. The question should be, what is going to deliver the best candidate experience? It should not be, how will I use this tool I already paid for? Money is always going to be kind of the biggest thing, but it has to be kind of that question of What is going to deliver the best candidate experience? And then you need to weigh it. I think again, back to your, your previous podcast, right? Kind of that best case, worst case, uh, most likely case kind of scenario and presentation. That's kind of one of the ways you need to think about it. Okay. Most likely case, we're going to use the existing tool to cater to candidates. That's the most likely. What can we do to make that the best?

What are the most, what are the least friction-inducing simplest things that we can do to create the most fluidity? In other words, What, what few things can we do to make this much, much better?

Exactly.

Because it's never going to be perfect and you're not going to satisfy everyone. Okay, I gotcha.

Right. And again, if we could, if we could bring these principles and these thought processes up front, we wouldn't have to back into it. We wouldn't have to say, okay, we already have Workday, you know, how do we make Workday meet all of these requirements? We could say, here are all of our actual requirements. Because I don't know if you've seen technical requirement checklists and—

oh yeah, um, for some of these companies supervised them, authored them, filled them out, both sides of the table, and even the intermediate. I'm gonna really piss you off now. It's really, really gonna piss you off. So if you are, um, if you are, let me put it this way, at what point user be damned? You know what I mean?

I mean, again, it really depends context.

I'm not saying this. I'm not going to de-advocate, again, making that up. I'm not gonna, I'm not gonna reverse the advocation towards better candidate experience. I get that. But still, candidate is still candidate. So at what point do you like, all right, this is enough. They still need to apply for the job. I mean, you know, at some point they're going to have to apply for the job. So if you make it— and here the reason I'm saying this is Easy Apply, LinkedIn's Easy Apply button. When I turned that off, we went from 14,000 applications per month, half of which, 50% of which were duplicates, to less than 2,000 applications per month. So we dramatically decreased the number of applications. Now, for those of you that don't know, Easy Apply is literally on LinkedIn. You click a button and you apply. Like, there's nothing. You just click a button and like, boom. So dramatically reduced the number, the volume of applications, but the pass-through rates went up and the number of qualified candidates went up. So we got a lot fewer applicants and a lot better applicants by just removing Easy Apply. So there is a point at which it's too easy to apply and too many people are applying many times, 10, 20, 30 times, where when we made it just a little bit harder— it's not hard, it's still an easy apply, but you have to go to our website and fill out 3 things— things got way better. So what—

that's, that's exactly how you apply design, and that's exactly how you apply design thinking. That is, that is not user-be-damned. You just changed who your user, who your intended target is. You changed who your user target is.

In other words, from just anybody that can click apply to people that actually want the job.

Okay, exactly. And that's, that's one of the— like, you know, we, we didn't really get into this, but The whole notion is your user base or your target users may not be everyone. To your point exactly, if you want certain people not to interact and you have something that's targeted out there, you want to limit the scope of who's coming in as opposed to increase the scope or broaden it.

Who is the intended user?

Exactly. Same principles apply. What language can I use that will turn off somebody? These words are too big, I don't want to work there.

Maybe this is a job for people who like big words. I got you. That makes sense now.

So you can—

dude, we are so like out of time. I'm gonna have to bring you back. I, it's like this is the most fun I've had with a complete stranger. Appreciate it, Charlie, in, in a long time. So let's do this again, promise?

Yeah.

All right, but I can't let you go without asking my two questions that I always ask. The first of those is, what do you want folks to think about over the weekend?

Sure. Just think about how they solve problems at work.

Thought-provoking.

Yeah. Just think, think about how they solve problems at work on a daily or weekly basis, right? If you're a recruiter, you're a sourcer, you're in HR, you're a practitioner, how are you going about solving problems? How much thought are you putting into it? Because again, that can apply to the systems. Are there things that you experience that are negative? Is there a dedicated resource at your company, like an operations team that will take your feedback? Can you give them that feedback? Have you been holding it back? Have you been working around things because something sucks? Is there someone you can go to, right, and have that conversation? Because maybe that will change things. Uh, maybe it's just that no one has raised—

nobody's asked the question.

What—

how about like just go up a level, right? So here's, here's the thing that you're having problems with. Try to go— try to track that backwards to the next step before that and see if there's something you can do there because you may not have to go all the way to the top to fix the problem, right? It could literally be, um, switching from, I don't know, a dropdown to a button or whatever, right? Okay.

There's, there's a great book, like quick plug too, there's a great book, I believe it's by Dan Heath called Upstream that talks exactly about that, how people oftentimes try to solve the problem where they're at when the problem is actually further up the process.

I like that. Upstream. Okay. Solve problems. And the last one for this show, because you're coming back, you promised, uh, who should be our next guest?

Sure, I would highly recommend Megan Whittington.

I don't know, you were prepared for this one.

But yeah, again, I've been listening to your podcast, Shelley, but I would highly recommend her just because I think, again, from that sourcing space, that knowledge base— she's at NVIDIA, you know. I talked a little bit about AI, I didn't get too into it, but she's at a company that's literally bringing in people who are building the GPUs that are going to power this future. I think her perspective would be really, really cool. I'd love to do it.

All right, I'll reach out. Man, thanks a lot for being on the show today. This was really informative, entertaining, eye-opening, and one of my favorites, very lively. So better than a cup of coffee to end my week. Thanks a lot, man. Appreciate it. Happy New Year to everybody. It is now 2024. Do something about it and be kind to yourself.

Amen.

Bye-bye.