Building a People Intelligence Product at HRBench - Jason Murphy & Matt Maguire - #196

[00:00:00] Hello friends of the podcast and welcome to Directionally Correct, a People Intelligence Podcast with your host Cole Knapper and today's guest Jason Murphy, Chief Product Officer and Matt Maguire, Chief Revenue Officer at HRBench. In this episode we will cover what it actually takes to build a people intelligence product Where people analytics need to move from being kind of a transactional function to more of a product based function

[00:00:26] How a product leader and a revenue leader see the same market very differently That's one of the sort of big things that gets me really excited about sort of the evolution of how product has kind of changed And what changes when you start treating your work like a portfolio that you actively manage Choosing the right projects to work on rather than just doing a job on the projects that you are undertaking And finding the right problems to solve, which

[00:00:57] Now let's get down to business Before we get into it, a quick word about HRBench, the company powering this podcast Most HR tech was built for the problems of a decade ago HRBench is built for what's next Analytics, Workforce Planning, Surveys, Data Health, Job Architecture And AI Workforce Transformation unified in one AI Native platform HRBench innovates faster than anyone in the market

[00:01:26] And scales to solve real organizational problems The future of people intelligence isn't more dashboards It's providing solutions for the business in real time Want to learn more? Book a demo at hrbench.com Slash directionally correct Find out more about the company powering this podcast And building the future of people intelligence As always, all opinions are our own And thanks for being a listener Thank you guys for joining me today

[00:01:54] Jason Murphy, the Chief Product Officer is here for the first time And then we've got returning guest in Matt McGuire, Chief Revenue Officer at HRBench Thanks you guys for joining me today Thanks for having us What do I get for the Two Timers Club? Is there a hat, a coffee mug? If we had hats or coffee mugs, I would give you one They don't exist But yeah, there does need to be like a reward for coming back for a second time

[00:02:22] What's the most that someone's come back? Do you know? It depends on how you count it I think we have had someone that's been on here for a third time But it was like kind of not in a traditional way So I think most we've ever had is like two times as just like a regular guest Nice Yeah I'm honored to be back And I know Jason is for his first time, right? Yes, I am It's like Saturday Night Live Whenever they have like jackets

[00:02:50] That's the Five Timers Club The Five Timers Club Yeah, Tom Hanks You've got a few more to go But yes, I'm very excited to be here Thank you very much, Cole, for inviting me and having me on your podcast Well, you'll be surprised that there's a lot of people out there listening that are very curious about what it means to be in the product space So there was this notion, I don't know, maybe five, six, seven years ago where people analytics need to move from being kind of a transactional function

[00:03:20] to more of a product-based function But most people that are, you know, I say most probably almost all people in people analytics or kind of this people intelligence that we've moved into None of them really have like a product background, like a traditional one And so I thought it might be really interesting to have kind of a more product-focused episode today about like what does it mean to be a product person? What skills do you have? What are some of the products that you guys have been building?

[00:03:48] And I think it just might lead to a good discussion But Jason, I'll start with you You know, what is it? Like what does it mean to be a product person? And what kind of things do you need to know to be able to build a product effectively? Yeah, that's a big question So if I think about in general, what does it mean to be a product person? It's you have a group of people, right?

[00:04:16] Your customers that you're trying to deliver business value to with some sort of a solution And it's our responsibility to work with and understand the problems that these people are having And try and come up with solutions with the technology that we're building to solve those problems To deliver value, to deliver business value to the folks that when they log into our application

[00:04:43] They know that they have, you know, one, two, three, ten different things that they have to do And our technology helps them to do them more efficiently than had they tried to do them without using our product Yeah So it's really trying to deliver the business value to folks who are using our applications

[00:05:05] Well, Matt, I want to bring you into this in terms of, you know, we talked a little bit about like what it means to be a product person But the product of people analytics is evolving into people intelligence And you have a lot of discussions in this space How have you seen those discussions evolve over the last few years? It's night and day from a few years ago So like my, we talked about this in the first time I was on my background

[00:05:34] My background is a customer success So working with a client base really to take feedback Pass that through to Jason and his engineering team to put into practice And even just like, I don't know, 18 months ago It was still somewhat the same process that Jason and I have been working on for the last 15, 20 years Put together mock-ups, you know, in a PowerPoint presentation Share that with someone, get their feedback

[00:06:02] Then go start to build a little bit Come back, share that with someone Maybe still mock-ups And then once you get to the end goal You know, the product, new feature, functionality, data, ready to release Then you push that out to the client base It's a long process And it's imperfect, you know, because there's going to be changes along the way Now, to fast forward to today

[00:06:27] It's amazing what can be done from being on a phone call with a client or prospective client Hearing something new that sounds really interesting and a problem that we can solve That maybe another vendor can't solve today Bring Jason on that call Sharing my feedback from the call And having him being able to rapidly prototype what this could look like in our environment With actual point-and-click functionality Let's walk through it with that person The next day, two days later

[00:06:57] See if this is going to work And then turn that over to our engineers And the people that are going to actually build it and put it into production Jason, I don't know, did I miss anything on that? I don't want to short change it because it's really amazing And it's my favorite part of my job now Is being able to take that sort of feedback that we're getting And then seeing it in real life Hours later Oh yeah, you didn't miss anything That's one of the sort of big things that gets me really excited about

[00:07:26] Sort of the evolution of how product has kind of changed And how, you know, we talk a lot We hear a lot about sort of the impact of AI And how it's going to impact the workforce And AI workforce transformation And this is one of those areas Where it's giving folks like myself I think a little bit more of a superpower than I've had before

[00:07:54] Where, you know, I can take the feedback that Mac gives me I can take the feedback I hear from customer success or support Or if we have something that comes through our idea portal Or we're doing some research And we're seeing that there's a lot of commonalities between these And I can work within sort of the AI coding tools that we have And build an interactive prototype that's code-based, if you will

[00:08:24] But not, you know, production ready code that, you know, that we ship to our clients But I can take the look and the feel of what our core application looks like And I can replicate and I can ideate And the feedback loop just went from, you know, weeks and maybe months To potentially hours or even days if it's a larger sort of undertaking

[00:08:50] But it's a hundred times faster to be able to get something in someone's hands That looks and feels like our application to get their feedback And the faster we understand if we're on track Or we're a little bit off of like solving that problem for those users The faster we can fix it, the faster we can get it out there And the faster we can deliver business value to the market It's been amazing I love that

[00:09:20] And you know what's so interesting about it is like You guys use this word prototype quite a few times And I think a lot about When I think about trying to stand up a product-based model Of doing this type of analytics and intelligence work I'm not sure how much people are actually prototyping what they're doing And then also another layer is like putting it in front of the right personas of customers So not just creating a one-size-fits-all model

[00:09:50] Which is, you guys know I rail against dashboards all the time A dashboard is and the intent is to try to create a one-size-fits-all model For all of your prospective customers And therefore it dissatisfies everybody And so, and we even had Kevin Trowbridge on the podcast not that long ago And he was just gushing about how much it feels like to be a customer And to be asked exactly what is it that you need And then things being prototyped and built to spec

[00:10:19] And then shipped and put into production Exactly for what he's looking for to address Not just trying to be the lowest common denominator like a dashboard might be And so, you know, one of the things that I've been talking a lot about is this When you get to this evolution into intelligence Like Matt was talking about a second ago That really it just collapses down into two layers There's an intelligence layer and a data layer And there's really not much else And so from a product standpoint

[00:10:47] How do you architect the right intelligence and the right data layer To make that intelligence possible, Jason? That's a great question I will say the way that our application is architected And the way that our CTO and software engineering team You know, brings the wealth of experience into enterprise software

[00:11:13] They've been smart enough to build the foundational layer To the point where it enables that intelligence that you're talking about It's very extensible And it's very easy to build upon To provide those solutions like Kevin was talking about We are not really sort of pigeonholed in a way In one way of doing things The everything

[00:11:42] The back end is flexible enough to build insights and intelligence That are necessary for this evolving world For sure, I love that And when I think about this is like What are some of the hot topics of conversation that you've been having lately, Matt In terms of what are people looking for in this intelligence space? And how is HR Bench addressing that?

[00:12:09] Yeah, I'd say there's a couple key areas One, it's the problem of how do we take our data from all of our disparate sources Feed it through some sort of layer like HR Bench To establish methodology, calculations, data cleanliness I don't need to minimize any of those Those are big things to put in place To then pass through MCP connectors to other AI tools Everyone's looking to understand

[00:12:38] Alright, how can we leverage AI at an enterprise level across our organization To maybe eliminate dashboards and the need and the creation of dashboards The maintenance To maintain your Power BI or Tableau dashboards for all your leaders throughout the organization How can we surface engagement insights throughout the org in an efficient manner? And I think everyone's looking at AI to solve that And I think that's, you know, the right move in many cases They're wondering how HR Bench fits into that

[00:13:08] And I think that's what I was mentioning It's like we serve as that middle layer of connecting your systems into one place Running data health checks, spotting anomalies, delivering the data methodology And the security, most importantly, to then pass that data throughout There, I mean, I can get into more layers of what people are looking for It's then once you've established that Let's take a look at all of your data today

[00:13:35] And let's take a look at jobs and do AI impact scoring By, you know, looking at your job architecture, your duties, responsibilities, tasks, and skills And seeing where there's opportunity to leverage AI at scale Those are some of the big things that, you know, I know you and I, Cole, we hear about a lot Yeah, you know, one of the things I did the other day, nobody even told me to do this I just, we had like a, like an ad about HR Bench at the beginning of the podcast Since it's powering the podcast

[00:14:03] And I just felt like it was so old and stale Because it wasn't reflective of what we were actually doing And so I changed it to having, you know, analytics, workforce planning, surveys, data health, job architecture And AI transformation all in one place in an AI native platform I was like, that seems to be a little bit more reflective of what we're doing nowadays But one of the areas that we've really been leaning into is the workforce planning and AI workforce transformation side

[00:14:31] Jason, what are some of the things that you have been building in this space? And I really feel like we're on the bleeding edge of a lot of this stuff And what like, you've had a lot of conversations with people who are trying to really tackle these problems Every organization is going through it What are the problems that you're seeing that need to be diagnosed? And then how have we been addressing it? That's a good question So some of the big problems, like you mentioned, Brad

[00:14:57] Like leadership today, the big question is like, how is AI going to impact my workforce? How is AI going to transform my workforce? And what am I going to do about it? And folks don't really have a good way to break down and understand the tasks that are responsible That people are responsible for performing within their workforce And so what we've been doing

[00:15:26] Some of the some of the stuff, some of the newer stuff we've been doing is You know, we we've been working on building out our proprietary taxonomy And we have some layers where you can actually map in your company jobs into the proprietary taxonomy That has a ton of different metadata around it And including the impact that AI might have and specific tasks for those positions

[00:15:53] And so by doing so you're building the bridge between between your actual position, the tasks that are performed within that position And you're starting to understand like where where AI might be impacting or where you may be able to utilize AI a little bit more to be efficient in your workforce Now that's sort of just one of those things Some of the other big challenges that we're hearing though is that's all well and good

[00:16:23] And we know that we want to build sort of an agent to do, you know, triage like P1 level call center issues, right? And the big problem now is what are folks going to how are folks going to handle Reskilling and redeployment? And so there are a handful of employees that, you know, may have an impact or may be impacted by these decisions

[00:16:50] And so leadership now is looking for a way to be able to understand who those folks are How they can reskill them and how they can redeploy them across the organization where they can be still valuable contributors to the company So it's really been a very interesting evolution to learn about sort of how some of these leaders are thinking about it

[00:17:18] And some actually are on the complete opposite scale where they're not even thinking about it at all Where they, you know, they might be a heavily manufacturing sort of type organization where it doesn't really feel like it's going to have an impact And others may be thinking really far on the other side of the spectrum where they've already gone ahead and implemented these measures But it's been a really interesting experience sort of to understand what the challenges are

[00:17:46] And try and work through them with, you know, trying to work through them building out the prototypes and building out the functionality to solve those problems Yeah, and Matt, I think like the conversations that you and I are having, we're talking to most advanced teams in the world, right? It's no joke What do you think it's like being kind of the secret in the marketplace? And it's almost like people don't believe how innovative we're being and how quickly it's going

[00:18:16] And, you know, I've been having discussions with people recently about how if you have kind of a blank slate Like if you're a company that started post-AI, you really can just innovate so incredibly fast But if you started pre-AI, you have a lot of technical debt and you need to, you know, kind of rip things down to the studs And then rebuild again and that really slows you down What, like, what, what, what have those conversations been like and how, how is that happening?

[00:18:41] Yeah, it's exciting. I love it that like we, we have our finger on the pulse of sort of what's going on because we're meeting with some of the best, biggest and brightest minds in the space right now And it's kind of exciting that they don't know the answer

[00:19:00] Like that's fascinating. I feel like you're, that's not a normal sort of environment to be in when you're talking to the brightest minds in workforce planning and people analytics and talent and intelligence They're all kind of trying to solve the same problems And there aren't really vendors out there that I know of that are doing all these things It's exciting for me just to sort of learn about how they're thinking about it It's exciting in a sense for HR bench because, all right, we can go and build this stuff

[00:19:27] We have the ability to do this today and to your point about companies that have started sort of post-AI and before AI We have the ability to go and build these things because we don't have the tech debt that other vendors don't, that other vendors have today Some of the things that get me really excited about, you know, this particular problem of AI workforce transformation is also the benchmark data aspect of it all Which I don't think we've even quite touched on yet We're certainly a benchmark company. We have benchmarks within the platform

[00:19:56] They're great, they're great for workforce benchmarks, comparisons on turnover and talent acquisition metrics But I think what other, the appetite that's out there for the people that we're talking to is like, what does this look like for other organizations too? Because you don't just want to carve your own path. It's risky to do that It's always helpful to know what the other big companies are doing out there today

[00:20:20] And I think that's where we also have the opportunity to present really meaningful data as people start to take action in our platform I don't know if Jason did it, did I just create another product for you to build right now, Cole? Hold on, let me go take notes and prototype that up real quick By the end of this podcast, I want to see the prototype Done I love it, that's how quickly it goes

[00:20:43] I mean, Jason, you know, brass tacks, did like these AI tools just change your life in terms of the speed and the quality of what you're doing? Yeah, absolutely. So as we were sort of, as I was thinking about preparing for the, you know, just this discussion I started to think about like in everyday life, you just, you have a challenge and you just sort of like do it And you have a work task and you just do it

[00:21:09] And as I've been able to sort of get deeper and deeper into these tools I've sort of expanded the abilities that I have And I don't necessarily take stock I don't sit back and think like, wow This is everything that has been changed for me so far But before this call I did that And I was thinking to myself like just that one conversation we had about, you know, the prototyping and closing the feedback loop from, you know, months to days

[00:21:37] That's just one thing But if we think about other parts of product management and, you know, the software delivery lifecycle, you know, documentation and QA are two really big things in that process as well

[00:21:53] And I've been able to utilize these tools in order to create skills that we can take, you know, QA sort of test documentation and the AI can go in there and it can write up these specific test cases and it'll do automated testing for me with a skill that I've gone ahead and created and it'll create a document like in a, like a, like a summary document of everything that it did for me.

[00:22:21] All the test cases that passed, what it tested, why they passed, anything that failed, what it tested, why it failed. And so then I can go in and I just have a summary of everything and it might be 20 or 30 test cases that are done within maybe 15 to 20 minutes where before that would maybe take a week or two weeks to sort of do manually. And I can manually go in and I can check and look at the ones that didn't pass and I can confirm whether or not it's accurate.

[00:22:51] And then I bring that information to the dev team and they fix it and boom, it's done, you know, within within 30 minutes or so we've collapsed down like weeks worth of effort into, you know, 30 minutes for one particular feature.

[00:23:04] And the same thing goes for documentation, right? If you think about like user guides and you think about release notes and you think about everything you want to have inside of a product for users to be able to find out right sort of self service or to be, you know, informed about the things that we've released. We can, I have a similar skill. It just goes in and it understands exactly what has been released and built.

[00:23:31] It gets screenshots of it and it builds a full on word document and with screenshots descriptions. I go in, I, you know, I modify it. We have a writing guide. It actually, I've created a writing guide because, you know, we know AI has like AI pros, right? There are certain, certain grammatical structures that you can sort of, that AI sort of like injects into its results.

[00:23:56] And so we've created our own sort of house style writing guide here for rules about how we want things to be written. And whenever we're creating documentation, it follows our writing guide. And so after the documentation is done, I go through and I make sure that everything looks good.

[00:24:13] It's factually accurate, modify it if I need to and save it and then send it to the people who need it, whether it's customer success, whether it's support and whether or not it's something that's going into our resources area of our product. Like all of those things that I just discussed would take weeks, if not months to do on their own. And now that's just compressed down to, you know, a couple of days worth of work. Can I give a, what's that?

[00:24:41] I was going to say, can I give a real life example too, of something too, that we've taken from a call. I had a call with a company that is growing by leaps and bounds in the automotive market. And they saw our people analytics solution and they liked it. They saw our org charting features and they loved those. We had got into headcount planning. They liked that as well, but the process that's broken for them.

[00:25:08] And it's now it's starting to pop up more consistently is that approval process and headcount planning. It's a spreadsheet process. It's always been a spreadsheet process. There's no centralized location to be able to administer that workflow, to have checks and balances and audit trail. And so this company is growing so fast that people are just opening up positions when they need positions and sending them off to talent acquisition. It's not going through the finance team.

[00:25:36] It's not going through FP&A, you know, to approve these as part of the budget plan for 2027. If it does, it's going through a spreadsheet that's passed around in email. That's not efficient. We didn't have anything in there for that. I took that feedback directly from that call, delivered it to Jason, and he specced and built out the workflow approval process for adding headcount to deliver basically on a platter to our engineering team to build out.

[00:26:03] So adding in the workflow and the right people because we have all the people in there. We have the salary information. We have the budgeted amounts and stuff like that. And so we can set up this routing workflow. We can put in thresholds, you know, based off of this amount of money, it needs to go to these people. There's a process that is broken across many organizations that can be fixed really easily in a short amount of time. I love that.

[00:26:27] I think it's a great example of like, you know, most of these problems that turned into spreadsheet problems were actually shortcomings of system of record problems. Right. A system of record should have done this out of the gate, but since it doesn't, we have to create workarounds.

[00:26:45] Well, the great thing about taking this kind of product based mindset is you can start to create systemic solutions to work around problems for these systems of record and just be kind of that intelligence layer that sits on top of the data that comes from these systems of records. I love that.

[00:27:02] And one of the things that I think for the people who are listening, who may even just be trying to build their own product function is between Matt, what you've said, but a lot of what Jason has said, frankly, about what it takes to, you know, validate a product, to spec it out, to, to, you know, all the skills that he's built. They don't know any of this stuff.

[00:27:24] And so just getting to hear what a true product person does would increase the quality of the products that are already out there that have been built in house too. So I appreciate you sharing some of that, that wisdom today. I guess, you know, is there anything that's, that's hot off the press that we've been working on that's coming released and in the next few months, it might even be out by the time this podcast comes out. You want to make sure people are aware of. Jason, you want me to take that one? Yes.

[00:27:55] While you, while you pause the cough, Jason is battling a cold right now. So we'll give him a little leeway there and I'll take it. And hopefully I don't overstep. And you can slap me on the wrist if I do a couple of things. I mean, even last week we're in the office for, you know, some company meetings, sitting around a table and designing what performance could look like in the system. Again, a separate solution that someone has to go to, to run different types of performance reviews.

[00:28:23] We built out stack ranking, nine box reviews. There's a couple other that are going to be in there that will be released by the time this call comes.

[00:29:01] And that's the way that we're going to go to the architecture library that Jason is referencing. That is ready to go as well so that you can match to our taxonomy and do that AI impact scoring. Those, I don't know, three, four things right there in the past, just to, you know, sort of tie this up. Those would have taken years to build in the past. I was just going to say that everything that Matt just described would have been a full years or even two years worth of a roadmap in the old world.

[00:29:26] And it's just the compression of, you know, delivery time is, it's astounding. What about tokenomics? What does that look like? It's a hot topic. People are trying to manage their agents and their workforce. They're trying to manage their spend. They're looking at things through the lens of strategic workforce planning. What are you guys seeing in that space? Well, we had a call yesterday, the day before, someone that was, their main problem was their token spend is in the middle of the world.

[00:29:56] They're trying to get millions right now. And they need to find a way to, A, reduce that, but B, to be able to track it and manage it effectively. And so this is not the first time this has come up. And so Jason and I have talked about this in the past, and we've been working on the ability to be able to capture that information both in HR Bench, but then in like an org chart view too, Jason, right?

[00:30:20] Where you're able to sort of land your AI agents throughout, manage your spend on like a payroll perspective, but then on a token perspective too, right? Yeah. So there's another part of the sort of, I'll call it the skunk works that we've been sort of working through is, you know, that taxonomy and everything I talked about with the Java architecture mapping and AI impact.

[00:30:42] Since tasks are related, tasks themselves are, have sort of a, a token quantity potentially a tie to them, right? You want the agent to do something and the agent itself spends X amount of tokens in order to do so at a specific cost, right?

[00:31:05] Different models or different providers have different models that have different, you know, input output costs, which translates to how much, you know, your tokens and how much you're actually spending. And so we can build a way, I don't want to call it a digital employee. It's not a digital employee.

[00:31:25] It's a, it's like an agent that captures the tasks that would commonly be performed in the workflows that Matt was talking about this, this, this organization was describing.

[00:31:37] And you can assign a dollar figure to that, and then you can actually slot that, that agent into the org chart that reporting directly to the person who may actually be running that agent that performs those tasks that consumes those tokens that equates to dollar figures for those tokens.

[00:31:58] And so you can put a, like an annualized dollar figure on what those tokens are going to cost for that particular agent to do those things. And it's reflected in, in, you know, the total cost and everything related into your workforce plan. I don't want to lead the witness, Your Honor, but also, Also, I do want to. Yeah, I do. Go ahead, do it.

[00:32:23] But we could theoretically then estimate the hours return, basically how much they're saving in terms of hours per task, right? If they're leveraging AI for this task. Absolutely. Yeah. It tasks themselves also have an estimated amount of time per week that they're given.

[00:32:47] And so, you know, through the, everything that's related to that agent and what it performs can actually bring that information back and show you your savings and cost related to the, to the hour savings. And that's all reflected in the workforce plan. Well, let, let me ask you the really important question, which is have either of you guys ever eaten at Wingstop? Uh-oh. Uh-oh. Uh-oh.

[00:33:17] You don't like this answer and I know that's why you're bringing it up. I have not. And it's mostly because it was a regional thing. I was never near a Wingstop when I was in Massachusetts. In July, I moved to California. There is one close by to me, but I'm in a new place where there's lots of restaurants. And every week we make dinner during the week and then on the weekend we eat out. And so I only have so many meals to choose. And there's all of these amazing places. There's Mexican restaurants. I'm outside San Diego.

[00:33:46] So the best Mexican restaurants in the U.S. I think there's burger places. There's In-N-Out Burger. There's, I don't know, endless amounts of restaurants, but I will add Wingstop to this list if I become part of the three-timers club and get invited back to the podcast. Ooh.

[00:34:06] So I've never eaten at Wingstop for similar reasons to Matt because I'm in Massachusetts and I think there maybe is one here now in Worcester. But I will say I had sort of a very close run into a Wingstop in an old job that I had. You had a close call? I had a close call. It sounds dangerous right now. I don't know where we're going with this one. No, no, no, no.

[00:34:33] It was just an adjacency thing where, you know, I used to travel to North Carolina a lot for a job and I would stay in sort of in the Concord area. And I, this, the hotel that I stayed at, I remember there being a Wingstop, like in the same like Plaza area that the hotel that I stayed at. And I do like wings. I just never really got around to getting over to the Wingstop to try it.

[00:34:58] So you, you love wings and you're actually very close to wings and hot wings and peppers and stuff like that, too. So I think you have a history there. I do enjoy a good wing. I do like spicy stuff. I used to grow my own very hot peppers, fatali peppers, red savina habanero peppers. You get pepper powder and sprinkle that on my food. And yeah, there's just something about spicy food that I really enjoy.

[00:35:27] And there's some good wing joints around here, too, that serve up some real spicy stuff. I remember B-dubs, Buffalo Wild Wings. We went in and did, this was back at a previous company, but some people that we work with here at HR Bench, we all went together and did, it's called the Blazin Challenge. And they give you, I think, 10, I think it's 10 wings of like the hottest sauce possible.

[00:35:56] And it was hilarious when they did it. They brought it out and they have this like, you know, think about like the flashing red light that goes around. They brought that out and they had like a whole, you know, procession of waiters and waitresses and they were all clapping and say yelling my name saying I was doing the hot wing challenge. And they got the whole restaurant to like come around. Did you do it? Did you complete it?

[00:36:26] I did. I did. So great. You have to do it in a specified number of minutes, too. Like you can't just like. Oh, I've done it. I've done it. I did not complete it. I did it in North Carolina after a wedding with my brother and sister and it did not go well. I made it through like two wings. Cole, have you ever tried it? Let's put it back on you. No, never tried it, but I do like all the flavors that wing stop. I'm going to bring it back home because that's that's like my favorite place.

[00:36:55] And I will say I think the three time club where you should get is some of the wing stop ranch as a gift like maybe a gallon. And then the five time club could be the blue cheese like because that's that's really where it's at. Yeah, but I think that's that's that's really the gift there. But I mean, Jason, you've actually done some pretty crazy stuff. Didn't you like used to do like video editing for like the NFL and stuff like that?

[00:37:20] Yes, it was a company, excuse me, a company I worked for that specialized in digital video editing for sports teams. And they're you know, their predominant market was football with NFL and in college football. And they were expanding out into different verticals. They had basketball as well, but they were looking for somebody who knew motorsports. And I actually sort of a secret that I am a huge NASCAR fan.

[00:37:49] And so I applied for the job and and I got the job and I got to travel around and go to NASCAR races for work. And, you know, I would be in the pits on Sundays setting up our system for our customers and I would make sure that everything ran right. And then we'll go to the shops in North Carolina during the weekday. And, you know, if I needed to, you know, update the system at all, I would do that. If anybody needed training, I would train them on how to use the software.

[00:38:19] It was it was it was a pretty wild job. It was pretty neat. But yeah, it was one funny story with the NFL is, you know, the Patriots, the New England Patriots facility isn't too far from where I live. And so they needed a big upgrade on on their software. And so a group of us went to the facility. This was during the off season and we were in there and, you know, updating everything. And we're in the locker rooms and all that stuff.

[00:38:48] And and we had the lights on. And all of a sudden I turned the corner and I run right into Robert Kraft. And he was like. He shook my hand. He was really nice. He introduced himself. I introduced myself and he was just like, what are you guys doing here? And I was like, oh, you need to talk to so and so we're up here. We're updating your software.

[00:39:13] But that was just a you know, that was my five seconds of, you know, fame and meeting Robert Kraft. But it was a pretty it was a pretty wild and fun job. And he's the owner of the Patriots, right? Yeah. Sorry. He's the owner of the Patriots. Yep. Yeah. So and I guess Kraft mac and cheese fame or whatever. No, no, different company. Different company. It's a paper company. Oh, oops. My bad. Yeah, that's OK. Yeah. Well, now that I messed up, do you guys want to join me in Kohl's Corner?

[00:39:43] Let's jump to Kohl's Corner. I love it. Welcome to Kohl's Corner. All right. All right. Well, Matt, you've been on the podcast before. So you've been asked some of these rapid fire questions. So I'll kind of gear these toward Jason. But let's do some rapid fire questions. Jason, if you weren't doing product work, what would you have done with your career? Oh, my gosh.

[00:40:09] Uh, probably brew beer. Brew beer. I think I'd be a brewmaster. I'd love to open a brewery and brew beer. What kind? He actually brews really good beer. I've had it before in the past. So I really enjoy an IPA. I like, I do like a logger, but the process is a little bit more involved and I don't have, at least we're up in my house, I don't have sort of the, the equipment to do loggering.

[00:40:37] So IPAs are sort of a pretty standard kind of formula. So I, I would start with IPAs and as they became more successful, get the loggering equipment. Nice. Well, just one thing, Kohl. So Jason likes, uh, beer, NASCAR and wings. We got an American beer. Yeah. And I live in Massachusetts. There's like one of these things doesn't belong. I agree. Wow.

[00:41:06] Well, Jason, what's the place you've never been to that you most like to go and why? Oh, um, I think Japan would be pretty interesting. Uh, you see a lot of stuff on, you know, at least for me, I've been to a company on TV about sort of technological advancements of Japan. And I always sort of think of it as like a, a futuristic place to be. Uh, the, I hear the cuisine is really good.

[00:41:33] Uh, I think it would just be really interesting to go to the complete other side of the world and experience that. I hear the NASCAR is good over there too. Um, yeah. Funny story about that. They actually, uh, uh, back in, I think in 1998, they ran a road course in Suzuka, Japan. And my favorite driver, Rusty Wallace ended up winning the race there. Well, a little bit of NASCAR history for you. There we go.

[00:42:03] Well, last one for you, Jason, if you were a character in any book, TV show or movie, who would you be and why? Uh, Homer Simpson. I feel like it's probably a little bit like Homer Simpson at this point with, uh, you know, the trials and tribulations of a sixth grader, a junior and a senior in school.

[00:42:31] And, uh, but that's, have you ever had a guest that didn't have an answer to one of your questions? Well, I'll wait them out. I've had them, I've had them there. They were like, I don't have an answer. I'm like, I'll sit here for the rest of the day. I don't care. What about a TV show? Anything there? Yeah. I was thinking George Clooney, but you know. Which, in what role? Am I, like, TV show, like early George Clooney in ER?

[00:43:00] I was thinking like up in the air because you used to travel a lot for work. There you go. Oh my gosh. I don't, I honestly don't watch a lot of movies in any of the TV shows that we typically watch or. Silicon Valley. You, you did watch that show. Who are you there? Are you Guilfoyle? No, I kind of think of like some of the other guys on the team is maybe Guilfoyle. Yeah. I think you're Jared. Wow. That's rough. Yeah. Burn. That is kind of rough.

[00:43:30] Uh, geez. I don't know that this is, this is a, I I'm like usually not speechless, but this is one's kind of stumped at me right now. I'm Jin Yan in Silicon Valley. I like that. I guess. Should I be like, you're early. Yeah. My big head. Yeah. Maybe. There you go. Your big head. That's good. All right. Let's do a one big question. And I want to bring both of you back in on this one.

[00:43:57] And, um, I think this will be an interesting answer, frankly, because I don't think we've ever talked about something like this before, but a lot of things we've been talking on the podcast is like how the innovation, innovation cycles are speeding up and how something that may have taken six months or a year or weeks, you know, is taking, you know, hours or days to do what, what does a product in the intelligence space look like a year from now? Right.

[00:44:24] Because a year is like five years or 10 years in the past. What does it actually look like a year from now? Um, if, if this is the, if we stay on this kind of pace of change, what I think it looks like a year from now is the, the barriers of like workflows inside of a application are broken down and

[00:44:54] you're either using a vendors interface to get everything that you need to have done, done, or you're using a different interface. Like Matt mentioned that we have an MCP server and that you can connect with a different like cloud cloud, cloud cowork for instance. Right. Where we're the base layer of a specified number or a specified set of data for a specified set of

[00:45:24] outcomes that are just a puzzle piece into like the mosaic of what an organization wants to do. It's like they're the barriers are basically gone. And it's, it's, it's a box and people can get whatever they want from a box. I think even just to take it from like, what is the space look like a year from now? Maybe a year, maybe we'll push this out a year or two years from now.

[00:45:51] I think those vendors that are hyper focused on a specific value that they provide like a specific feature and they're specialized in that are going to be few and far between. I think the need for those types of vendors are going to go away.

[00:46:06] I think to Jason's point from a user perspective, I'd say the large, in a year from now, the largest, most technically advanced companies will maybe use something like an HR bench as that middle layer is that intelligence layer to connect to all their other tools. And then everything's going to be conversational for their key stakeholders, their leaders or managers, and they'll just be able to ask what they need for.

[00:46:32] I still think there's going to be a value to the deep level analysis that, you know, maybe a people analytics team is going to be doing within a platform. So I don't think that completely goes away. But on the other side of the coin, I think your small to midsize organizations in a year from now, they're not going to be as advanced. They're not, they're still going to need a people intelligence solution.

[00:46:54] Yeah, I think I like, I like both of your answers. The thing I would say is, you know, I've been saying that people intelligence is just combination of everything collapsing down from what it used to be into a singular intelligence layer and a data layer. The one thing that's left out of that equation is like the system of record platforms. Yeah. And the reality is, so system of record platforms are gobbling up entire different things to try to like make their platforms better.

[00:47:21] But again, they're going to have to tear all those things down to the studs and rebuild them back up again because they weren't built natively to be applications. And so until you know you get a bunch of truly AI native platforms, you're still going to have to rip out all your data, which is going to create that data layer. And you're going to have to have the intelligence on top of that. One of the things that Matt, you and I have been talking about a lot. It's like what, what used to be called products are now just features because everything is collapsing.

[00:47:50] And and so when I look a year out into the future, I just see, you know, we're still on our journey to get to like a true level of intelligence about everything. I think a year from now we have that intelligence layer on everything. And then you start to look at like, ooh, what are the cool adjacencies? Right. What are the areas that, you know, historically data didn't even play a role in? Maybe we're getting more embedded with finance, maybe more with operations, more with sales. Maybe you're getting more into that platform layer.

[00:48:17] Maybe you're showing you're starting to become a workflow company on top of all the intelligence that's being done to kind of reshape the work itself, not just analyze how the work is being reshaped. I think you can go a lot of really interesting directions if you start to play out these trends and just extrapolate, you know, into the future as if the trends going to continue. So I think it's going to be a lot of fun to be on this journey. Absolutely. All right. Let's let's do some. What am I reading? I'm going to pull up the first one here.

[00:48:45] Did you guys did you guys read through these beforehand or not? Or am I going to carry the carry the water like last time? Like last time? Wait, I thought I read them the last time. Jason, did you give me a look? No, you're going to have to carry my water. All right. So you're carrying some water. Just like Matt's last time. I need I need you to carry it for me. All right. So the first one we got, it's called Design for Value Portfolio Management and People Analytics.

[00:49:14] And this is written by a guy named Jack Costell, who looks like he's at General Mills. And so he's saying if you're trying to do build a portfolio of people analytics, you need to start with the value chain. And good people analytics is not just better analytics. It's creating reliable value creation from people data. And so he goes through and shows how there's a difference between portfolio management versus operational excellence. And we tend to historically focused on operational excellence, sometimes to the detriment of the portfolio.

[00:49:42] So he creates this two by two chart of high versus low execution quality and high versus low portfolio quality. And you're really wanting to get in that top right quadrant where what he calls a value engine of strong bets delivered well. And then he goes through and he gives a few more examples of the potential upside of taking this portfolio approach of the magnitude of value.

[00:50:04] If it's successful, the strategic relevance, the scale, the population affected the risk avoidance and then times the realizability of all these things. And then last, if you're looking at things like a portfolio across the entire world, you can kind of look at it from a heat mapping perspective of saying, you know, where do you have low or weak value? Where is the value medium? Where is it high? He does this geographically speaking across the world.

[00:50:29] But you could really look at it in a variety of ways across the portfolio of how you're trying to support a different business. I'll stop there. What did you guys think about this concept of portfolio management rather than teams just focusing on operational excellence? Jason? Pause 20 seconds. Back to you, Cole.

[00:50:53] Well, the thing I think I when I read that, I did read the article and I went through it and it was talking more about choosing the right projects to work on. Rather than just doing a job on the projects that you are undertaking and finding the right problems to be solving, which I think is, you know, a great way of approaching it. If you're giving people analytics teams autonomy to work on the projects that they feel are the right projects rather than just being hand delivered. Here's what you should be working on.

[00:51:22] I don't know. I thought that that was an interesting insight to pull from the article. Yeah, for sure. And I think that's what he kind of means by portfolio is like, what are the types of like the suite of projects and kind of deliverables that your team is doing? I think there's a little bit of corporate jargon involved there around like value creation is like, what does that actually mean? I'm more of, you know, direct, plain spoken type of person when it comes to a lot of these things.

[00:51:49] And so what I think he means is he's trying to say, you know, should you try to optimize for just creating like I keep saying that like a better mousetrap? Yes. Are you trying to spread across the different aspects of what your product and the things that you're building might do to support your business customers? And the reality is, it's a lot of stuff we've been talking about today. Frankly, that's why I chose the article.

[00:52:14] It's choosing the right personas, not trying to create a one size fits all model. Dashboards are stupid. You know, let's get into what can we do to make sure we're meeting our customers where we're at. We're designing things with the user in mind. A lot of that type of stuff I think is very relevant. What he calls a portfolio, I might call a product based approach. So I thought this was a really good one.

[00:52:36] But let's go ahead and move on to the next one since this is going to be a lot of fun explaining these hopefully in a way that Jason can weigh in. So the next one, friend of the podcast, Ludic Stella, previous guest, wrote an article which I saw was excellent. It's called Company Culture as a Forward Signal of Financial Performance. And he is the consummate scientist that he is. He is very clear that this is not causal, not causal work.

[00:53:06] So I'll say that again, it's not causal. So what he did is he took a bunch of data about I don't know exactly how many companies. I think it's over five. Oh, yeah, it's the whole set S&P 500. So 500 firms and analyze their financial performance compared to some data that they pulled out of this tool called Culture X that looks at glass door reviews posted between January 2023 and January and April 2025.

[00:53:33] And showing the results of the relationships between, you know, these kind of what you might think of is like like how people are feeling about their work, the culture, their sentiment about their organization. And does that relate to the organization's financial performance?

[00:53:53] And so he looked at things like the ROA, the return on assets of a firm and the relationship between things like transparency, leadership, openness to feedback, toxic culture, supportive culture, strategy, innovation, agility and work life balance. And there seems to be a pretty clear relationship in terms of things like the more transparent is actually the most predictive of return on assets and having low work life balance is a negative predictor. So pretty interesting stuff here.

[00:54:22] And then he splits it out across not only return on assets, but operating margin revenue growth. And I can't read it. It's a little bit small. So there's a fourth category. The relationships aren't quite as distinct if you look at operating margin revenue growth in the fourth category. But if you're looking at things like return on assets, there seems to be a very, very clear pattern.

[00:54:43] Again, this is not causal. So it could be that companies with great return on assets tend to have, you know, higher culture X type of ratings, not the other way around. But I thought this was a fascinating one. Matt, since you actually got a chance to read it, what did you think about this one? And Jason, I'd love to bring you in afterwards. Yeah, I actually love this article. I love the analysis that he was doing there. I had a couple of thoughts about it. One, they're really similar to that.

[00:55:10] There's a couple of vendors out there, companies out there that are doing some some of this type of analysis and investment. Well, there's the good job score, which is a survey that we actually administer in HR bench that was put together by Six Sigma FOW partners in conjunction with PwC, which is a survey to run to tie job outcomes to business performance.

[00:55:33] And the idea is that if people feel better about the role, the quality of their job, that there's a future at the organization that they're going to be higher performing. And I'm a firm believer in that, not just the survey, but that notion in general. Also, I'm trying to think of the company that does this, but there is is it Energage?

[00:55:51] No, no, it's not a survey vendor necessarily. It's a it's an actual fund that you can invest based off of the scores that these companies do you know this one? I've seen I've actually seen them present before. I can't remember their name, but it's it's something like future of work investors. I know that's not who it is, because that's an investor that you were just talking about a second ago.

[00:56:14] But it's a name similar to that. And I have seen them present before when they say that it's not 100% of their portfolio, but it's a pretty substantial part is invested in companies that have higher culture scores on some of these kind of open source metrics. And they're really also kind of from like an activist perspective, trying to push other investment firms to invest in these companies as well.

[00:56:37] Definitely. And just showing how if you invest in your employees that will actually make your company have higher valuations, which ultimately would lead executives to want to make those types of investments. Is with that study, am I correct in that it was Glassdoor primarily that he was using for the. Yeah, it's culture X analyzing Glassdoor data. Yeah. And that's where I like, you know, I have a problem with some of the Glassdoor data because it's not giving a full representation of the current employees that are there.

[00:57:06] There's tricks and trades of Glassdoor companies also like getting people to go and post positively about the organization. And then there's also the notion of like the people that are most unhappy and leaving an organization are going to be the ones that are most vocal and posting stuff and adding reviews of that company. I think it's a good data point. I think there's a lot more that could go into it, including internal surveys that you're running.

[00:57:30] And that's what like I would love to be able to pull those external data points, which we should be able to do with HR bench like Glassdoor and other indicators into your analysis of company health and employee health. Looking at external measures like that, looking at customer sort of scores, too, because there's another side to this is a different study that maybe we could do is like.

[00:57:53] NPS scores for your customers and how that impacts employee net promoter score, which then how that impacts company performance. NPS scores for your customers. Any thoughts on this one, Jason?

[00:58:05] Yeah, no, honest, like taking all like the scientific method out of it, as I'm listening to Matt and I'm listening to you describe it and I'm looking at it now, like, I'll just say like, like, at a 35,000 foot view, every company that I've ever worked at has always either been really good or really bad because of the things that he is actually pointing out right now.

[00:58:31] Now, the companies that have been most transparent, the leadership has been great. People open to feedback. All the above have been very well run and very high performing companies. And so as I all the science aside, right, it just passes the smell test. And it's very fascinating that he's put it all together with sort of the use of data to sort of prove that out.

[00:59:00] Nice. I love it. Well, let's round this out with the last one. This is a little bit more dense, but frankly, Jason, I thought you would like it. It's just a shame we didn't get a chance to read it yet. But this is from a friend of the podcast, Dan Putka and the team at Humro. I did a webinar with Dan last year. He's really fantastic. I remember it's a white paper they put out called indexing the impact of AI within the O-Net system review of methods and development and recommendations.

[00:59:28] This is through the National Center of O-Net development, which was prepared under RTI International, which is going in the way back machine. I worked there as my first job out of grad school. And we actually went to the center that actually collected a lot of the O-Net data, which is fine. But I want to skim down here because essentially what they're trying to do is to show how you would use AI or how you could start to use how they would collect data with O-Net to put AI impacts into O-Net.

[00:59:56] Looking at things like the AI conceptualization, the AI impact definitions and scaling, drawing inferences about AI's impact on specific elements of work, drawing inferences about AI's impact on occupations. And then they drew up this model here, which I thought was fairly interesting about the AI impact literature review framework, where they're looking at the first step is looking at AI capabilities and AI usage.

[01:00:20] The second step is showing that relationship between those two things with task statements, KSAOs and the knowledge skills, abilities and other characteristics, and then bringing in job vacancy elements as well. And then showing how that there's AI impacts on the task level. So, task level AI impact scores, as well as the KSAO level AI impact scores, ultimately leading, if you aggregate those up, to an occupation level AI impact score.

[01:00:49] So, again, very, very technical article. I think it's like 40 pages long. I'm suspecting that nobody read the whole thing. But I thought this is a really interesting framework, especially within the confines of the conversation we were already having today about AI's impact on the future work and on the workforce. What do you guys think about this one? Do you know what the job vacancy elements were in there?

[01:01:11] Yeah, I think what that's really talking about is essentially if you're thinking about not every role is filled at a particular time. Right. And so even though if you're looking at things from like a scientific perspective, you've got to minimize error in your model. And one of the ways of producing error is like, let's say you've got 100 roles at your company, but actually five of them are vacant at any particular time.

[01:01:40] You still need to have the data associated with those five or otherwise your model isn't encompassing of 100% of what you're trying to analyze. So it's like trying to be very scientific rather than kind of satisfying and just going with what's easy to choose. But I wouldn't overly index to that part of it. I think it's just to make sure that they're accounting for 100% of roles. Yeah, that was just curiosity on my part because I saw that in there. It was interesting of what could be captured in there.

[01:02:05] My initial thought is, all right, can we get on the phone with them, Jason, to talk through the study and then pull that into our AI impact scoring? Because the level of depth in there and the analysis that they're doing is impressive. Yeah. As I'm looking at this sort of, you know, this little architecture for this technical schematic, I think it's amazing, first of all.

[01:02:26] But as I've been thinking about this whole thing over the, you know, this past year, if you will, like the very first column, like AI capabilities and AI usage, I think are the two hardest things to actually nail down. I know, you know, larger sort of, you know, companies like Anthropic and OpenAI produce papers for the impact of sort of AI on work.

[01:02:54] And, but I think they're sort of like backward looking. They're relatively, like we talked about the pace of improvement and the pace of sort of innovation and AI and what it can do. By the time a paper is produced, it's already sort of stale, if you will, right? The way that these models come out now and the way that, you know, they're able to do different things than they were even, you know, two months ago.

[01:03:21] So like that, that part of the whole sort of framework, I feel like is the hardest to nail down. But everything else about it, like part two, part three, I really like the idea of the job vacancies and bringing those elements in to be able to make sure you have sort of 100% coverage.

[01:03:40] But I fully agree, like the whole, talking about task statements, that's basically, you know, we're talking about the taxonomy a little bit earlier with sort of our job architecture. And that mapping work is, you know, everything's distilled down to a task. But if you think about, like, this may be overgeneralizing this, but if you think about work, it's just, you know, it's a, it's like a box.

[01:04:06] It's like a job title and there are duties and responsibilities and that's going to be in a specific location. And somebody who has the skills in order to perform the tasks that are, you know, required for that to be successful and to accomplish the goals of that job. It's like all distills down to tasks.

[01:04:27] And when you think about agents and agents in general, you're just, you're putting a framework around sort of this automation of, I want this thing to do X, Y, and Z. And here is the rules about what it can do. Here is what the outcome should look like. And here's what happens if it doesn't do what you want it to do, right? Like in Eric, like what happens if there are mistakes? Like how do we recover from those?

[01:04:52] And so distilling this stuff down to the task statements related to all the particular jobs and the vacancies and all the skills and all that stuff. And then sort of bringing that back up into sort of their occupation level and how they have their, their sort of, you know, their taxonomy structured. I think this is amazing. I think this is pretty awesome stuff. I love it. Well, I'm really glad I got to bring you guys on. I think this is a fantastic episode.

[01:05:22] Before we wrap up, is there anything you want to chat with me about or any other topics you want to do before we close up? We'll just start planning for the third episode. We can, we can cover that offline. Really appreciate you bringing us on. I know Cole, you talk to me all day long. So stomaching another hour with me is greatly appreciated. We're big fans of you and the podcast. And so speak for Jason saying we're honored to be on the, on the pod. 100%. This is my first ever podcast.

[01:05:51] So I hope I didn't like sully your reputation at all, but this has been an amazing experience and I really do appreciate the invite. And I thank you very much. Yeah. Thank you guys. That's really kind. I'm really appreciative of having you guys on it, Matt. I'll make sure I get the gallon of ranch ready for your third time gift. Perfect. I'll drink it. If you haven't tried Wingstop by then, I'll be very disappointed, but we'll, we'll make that happen. Maybe I'll send it as a, as like an Uber Eats or something to both of you.

[01:06:20] I might have to go do this Wingstop. I might have to drive to Worcester and go get, go get some wings. Is this podcast sponsored by Wingstop? If not, it might be soon enough. Yeah. I'm trying to get them to sponsor, you know, free wings for life. You don't have to give me any money. Just give me free wings, Wingstop. If you're out there listening. So you've been listening to Directionly Correct at People Intelligence Podcast with your host, Cole Knapper. And today's guests, Matt McGuire and Jason Murphy. Thanks for joining me guys. Thanks, Cole. Thanks, Cole.