Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Kevin Plunkett: Why Your Compensation Strategy Is Only as Good as the Data Underneath It
7 Plays11 days ago
In this episode of MustardHub Voices: Behind the Build, Curtis Forbes sits down with Kevin Plunkett, Vice President of Partnerships at Salary.com, for a conversation about the infrastructure underneath compensation that most organizations have never thought carefully enough about.
Kevin came to this world from an unlikely direction, spending years as a talent agent in Hollywood before his brother founded Salary.com and pulled him into building out what became one of the most widely used compensation data platforms in the country. That origin story matters, because Kevin sees compensation data the way someone who has negotiated on behalf of talent sees it: as something with real stakes for real people, not just a number in a range.
The conversation covers why COVID permanently moved compensation from the back room to the boardroom, what happens inside an organization when salary compression goes unaddressed, and why the promise of integrated talent management has consistently failed to deliver because nobody bothered to get the job architecture right in the first place. Kevin also makes the case that AI is not a threat to compensation data companies so much as a demand signal for better data, since a prediction engine is only as good as what you feed it.
If you are building an HR platform, running a staffing firm, or trying to make workforce decisions with data you are not sure you can trust, this one is worth your time.
About the Guest
Kevin Plunkett is Vice President of Partnerships at Salary.com, where he has spent the better part of 26 years at the forefront of compensation data. Before finding his way into the world of comp, Kevin had a different kind of career entirely, working as a talent agent in Hollywood. Today, his focus is the wholesale side of the business, bringing Salary.com’s compensation data to the staffing firms, HR vendors, and platforms that build it into their own products. He works at the intersection of data, technology, and the fast-changing economics of pay, helping partners understand not just what the numbers say, but how markets are trending and why it matters. With more than two decades watching compensation move from the back room to the boardroom, Kevin has a front-row seat to how data is reshaping the way companies attract, retain, and pay their people.
Recommended
Transcript
Introduction and Career Overview
00:00:05
Speaker
Welcome back to another installment of Mustard Hub Voices Behind the Build. In these episodes, I sit down with the people building, backing, and running better workplaces. I'm your host, Curtis Forbes, and my guest today is Kevin Plunkett.
00:00:18
Speaker
Kevin's Vice President of Partnerships at Salary.com, where he spent the better part of 26 years at the forefront of compensation data. Before finding his way into the world of comp, Kevin had a different kind of career entirely, working as a talent agent in Hollywood. Today's focus...
00:00:34
Speaker
is the wholesale side of the business, bringing salary.com's compensation data to staffing firms, ah HR vendors, and platforms that build it into their own products. He works at the intersection of data, technology, and the fast-changing economics of pay, helping partners understand not just what the numbers say, but how markets are trending and why it matters.
00:00:54
Speaker
With more than two decades watching compensation move from the backroom to the boardroom, Kevin has a front row seat to how data is reshaping the way companies attract, retain, and pay their people.
Hollywood Career and Transition
00:01:06
Speaker
Welcome to Behind the Build. Thanks for joining me, Kevin.
00:01:08
Speaker
Thank you, Curtis. appreciate it. been ah Been looking forward to this one a lot, um you know, for for a lot of reasons. Number one, I'm super interested to hear about your background at Hollywood. But ah even though that's that's kind of secondary, um you know, salary.com has really built this infrastructure that a lot of organizations run on. And um and that's a super interesting conversation to have, especially, you know um you know, given the ultimate outcomes, right, that we're looking for, attracting, retaining, sure and paying their people. So, um you know, before we kind of get into that, i have to ask about this career arc, right? It's not the one most people in this space have. You spent time as a talent agent before landering landing at salary.
00:01:54
Speaker
yep what What got you started there and then what Pulled you from that world. yeah Fair enough. um So initially, you know, upon getting out of college, i i went actually into an an acting. I was I went to go pursue an act ah MFA in acting. Right. OK.
00:02:16
Speaker
At a conservatory program. I spent about a year in that program and I realized, you know, the actor's life was not going to be for me. So I thought, well, let's get into producing because, you know, that was interesting. So I wanted to do television production.
00:02:30
Speaker
And started locally, worked at, um you know, local newscast for about a year and a half and then quickly found out at that at that time, this is back in, you know, late 80s now, ah everything was running out of Hollywood, right? There wasn't much coming out of New York. There wasn't much in Canada. Like you didn't have all of these outlets and places and things. It was all, everything was centralized in in in Hollywood.
00:03:00
Speaker
So I quickly realized, you know, if I want to make a splash in this business, I got to be out there. So packed up the car, drove across country.
Building Salary.com and Salary Wizard
00:03:09
Speaker
um As luck would have it, I wound up landing a job at you know, the premier talent agency in Hollywood at the time, Creative Artist Agency.
00:03:20
Speaker
um Happened to have a family friend that worked there and he helped kind of grease the skids and yeah got in as a temp. And within a week, I had a permanent placement. So um started off in in in what they call TV lit, which was writers and directors for television.
00:03:38
Speaker
And did that for probably about six months. And then a desk opened up with um in feature motion picture lit. um, that, you know, again, represents, uh, writers and directors. And these are, and and the second desk is the one that I spent, you know, a good chunk of my time there. And then, um, it was a big, you know, there was this whole kind of, um, uh, uh, takeover. Well, if you will, like the, the, the head guys that were running the show, this guy, uh, uh, Michael Lobitz left. And then, um,
00:04:15
Speaker
the sort of baton got passed to this group, what they call the Young Turks, of which my boss was one. So all of a sudden, my stock kind of rose with his, right? And, you know, now I'm all of a sudden working on, you know, one of the premier desks in in all of Hollywood. So that was kind of exciting. And yes, we worked with lots of stars and people you would know, directors, writers, actors, all of it So I did that for quite a while. And then eventually I left there and um went and became an agent at a smaller agency, you know, where I could really kind of dig in and cut my teeth. Did that for about two years. And then I just got burned out on the whole thing.
00:04:56
Speaker
And, ah you know, and the Internet was just kind of getting started. This is around 2000 now. And um my thought was, well, maybe I'll move up to San Francisco. I had a bunch of buddies that were out there.
00:05:10
Speaker
and you know see what I can find out there. At the same time, as I'm sort of figuring all this out, my brother um started, ah he he actually started salary.com and had pulled it together initially. It was, um the initial piece was more like um what Yahoo, ah almost like a directory of compensation.
00:05:37
Speaker
And so sort of my first job was helping him build out the website so he could go out and get some funding. And what ah what when I was you know required to do is go out and find salary data information that was posted you know on the internet. Remember, this is back in 2000 when the internet wasn't what it is today. There was no Google. There was no super easy way to find stuff. You had to really dig.
00:06:02
Speaker
So we would you know the goal was to find all this stuff, organize it in a way we basically... had these categori categories and we had all this salary and compensation information tucked in and it was really like a directory. You'd click on the link and would take you to a page.
00:06:19
Speaker
Eventually we wound up building out what we call the salary wizard. It's still in existence today. It's the thing that really launched the company as a whole. And it was it was a product that people could go in put in a job title, put in a geography and industry and get a salary range back.
00:06:41
Speaker
And at the time that was pretty, you know, that was pretty um ah cutting edge. yeah um And back then my job was ah to run partnerships for that product where we basically syndicated when syndication was a big deal, we would syndicate content and we syndicated that to every career site you could think of every newspaper career section, every magazine career section, um you know, anything and anything that had to do with careers, we, we syndicated this salary wizard at the time. think at its peak, I think we had like close to 800 partnerships where we had syndicated this, this
Evolution of Compensation Data Solutions
00:07:25
Speaker
content. So wow that's what really got salary.com as a brand launched was through that syndication network.
00:07:32
Speaker
And then as that grew, um we started getting more eyeballs and so we needed to monetize them. And so we started doing advertising. So oh I did both of those things, led both of those groups for quite a while until we got enough traction within the ah HR side of the house um with our with our HR specific products, which is really what the company is known for now is the um the HR products that we provide.
00:08:02
Speaker
And then I started doing partnerships um with that suite of products. So I can go in more depth there if you want, but that gives you at least an idea of kind of the basics. That's ah no, that's a that is a great. It's a really interesting story and it's a really interesting kind of career trajectory for folks who don't kind of live in this world. When you say.
00:08:24
Speaker
You sell compensation data like wholesale. What what what does that actually mean? Who shows up to buy it? What are they trying to do with it? Let me take a step back and and just give a quick overview of the of the company as a whole. And that'll help frame how this little section that I've been working on and working to grow and develop it came to be.
00:08:45
Speaker
So... When you think of salary.com, you sort of think of us as as sort of your one-stop shop for any and all compensation needs, right? We've got data, we have software, and we have services.
00:08:57
Speaker
And we tend to meet the customer where they're at in their journey and help them to be as self-sufficient as they possible as they want to be um within this you know journey of compensation, right? Typically ties into philosophy and growth and a whole bunch of different, um, other macro kind of, um, goals that the organization is trying to meet and compensation obviously is, it, you know, supports all of those.
00:09:23
Speaker
So how do, you know, how do we help that company get there through, through, through the lens of compensation as part of that? Um, you know, we created probably one of, but probably one of the the best and, and, and most sort of broad based, um,
00:09:42
Speaker
interpolated data sets available, right? it's It is all HR supported data. We don't have any, you know, there's no, um you know, user supported data in there or or user inputted data. And it's all, it's all HR coming coming directly from HR organizations, typically from um salary surveys is really at its base and it's hundreds and hundreds of publicly available salary surveys.
00:10:07
Speaker
So we've built out this whole big, broad taxonomy of jobs, um with, you know, it's like 20,000 different industries across 224 different sub industries, you know, company sizes, you know, here in the US, we were able to draw all the way down to the zip code level. So a really robust data set.
00:10:28
Speaker
And we as an organization got to be known for having this really high quality data set, right, that that is really broad, and really touches pretty much all jobs across across the US, Canada, and and now we're in 30 different countries.
00:10:43
Speaker
So we only sold that data set through our our software products and directly to end customers. We would get asked by other HCM vendors, payroll companies, staffing firms, what have you, if they could get access. And we for a while said, no, no, no, no. no Then we started giving people access through the product But eventually what happened was people wanted to get access to the data through some kind of a feed, an API or ah a data file.
00:11:14
Speaker
And they wanted to take that data, put it into their offering, either through through a data lake and merge it with other data sets, or um they wanted to tap into an API and be able to pull it right into the the software product. we didn't We for a long time didn't want to go down that route.
00:11:33
Speaker
That changed um about a year, year and a half ago, maybe two years ago um when we started to dip our toe in that. And that's what I'm doing now is I'm sort of heading up the sales in that area where we're looking at effectively wholesaling the data sets that we have.
AI and Compensation Data Trends
00:11:50
Speaker
And we've certainly grown. We have more data sets than than just the the comp analyst data set now. um But it's the it's the wholesaling of all of those data sets.
00:12:00
Speaker
and and the delivery of either through an API or through cube, a data cube, right? ah Like a CSV file. um That's really where, um what's really been driving this. And AI has been a big supporter. The advent of AI has been a big help in pushing that because people want more and more data into their systems.
00:12:24
Speaker
when when a When a staffing firm or HR vendor, you know, when when they come to you, i'm curious, what are they what are they really after? Is it the raw data itself or is there a story they're trying to tell their own ah customers, their own people, their own? Yeah, so it will depend, right? The use cases vary.
00:12:48
Speaker
um In some cases, they're trying to support an existing methodology that they have, like vis-a-vis a staffing firm. um you know Staffing firms are, the you know one of the bigger use cases is they need to sometimes need a third-party validation to help educate all parties about what you know a particular job is in a particular region. So where where staffing firms get into you know, some challenges is they go in, you know, they've got a customer that's kind of got, you know, ah you know, champagne taste and beer budgets, right? When it comes to providing compensation and the staffing firm is trying to say, hey, look guys, we can't really go recruit for that because you're paying well below market.
00:13:37
Speaker
And, you know, the customers go, well, this is what I see the market at, you know, that's where we can really help and say, look, here's a, here's a valid, quality, you know, um a third party that can validate what the market is for that given position.
00:13:53
Speaker
And so, you know, there's there's a data set there that does that. um And that helps a lot. you mentioned You mentioned these different ways that you're that you're delivering this now, right? The API, the bulk, you know, feed the the in in into customers' you know own data lakes.
00:14:14
Speaker
um Unpack that for me a little bit. Like what what kind of customer reaches for which and why does the delivery method matter as much as the data itself? Again, sometimes this is use case based other and and other times it's more um just sort of philosophical.
00:14:32
Speaker
So where an API works is you've got some volume running through your system um we've got this sort of ai matching functionality, right? So you just put in, i mean, bare minimum, you could put a job title and you can get a salary result. The more data you put in, the better the return of the salary range will be.
00:14:55
Speaker
So you'll get a much tighter range, a better range. um So as an example, um a company may be looking to provide guidance around pay transparency on, let's say, a staffing or um or through an ATS system.
00:15:11
Speaker
um and they want to be able to, you know, give advice to the posting, the company that's doing the posting, hey, look, this is, you know, more or less the range that you're, that, you know, um the market is suggesting, you know, you're you're suggesting a salary range, you know, either really high or really low, you might want to rethink that, right? That's one application.
00:15:34
Speaker
Another application is um You've got, let's say, a compensation planning module within a larger HCM suite, thing a payroll company, and they want to provide benchmarking to help provide some external analysis around jobs so so their companies can understand you know what's the benchmark for this position and this in this area.
00:16:00
Speaker
API is great for pulling that in. where ah where the data queue with the queue might be valuable is some folks may not want to deal with an API. They may not want the, you know, sometimes APIs can slow performance.
00:16:15
Speaker
Ours is pretty fast, so we haven't had that problem, but there's a perception that, you know, folks want to deal with everything in-house. They they want they want to deal with all of the data. They want to dump it into a data lake.
00:16:28
Speaker
Maybe they're commingling it with other data sets or other um taxonomies or what have you, or they want to handle and control all the matching. Right. They want to go through and handle and do all that matching because they've got their own taxonomy that they like and they use and they don't want to be reliant on our AI matching technology to work.
00:16:49
Speaker
So, you know, ah control, having a certain element of control of data um will help differentiate which, which mechanism works better yeah for one versus the other.
00:17:03
Speaker
You, you, you made an interesting point, I think in our last comment, we, we've spoken a couple of times, made an interesting point that that for a long time comp lived, you know, in the back room, I think is how you described it, you know, and and I'm curious,
00:17:21
Speaker
How that's changed. Like when did compensation start moving to the forefront of the ah HR conversation? So comp's always been this like, you know, um little, you know, section within HR, right?
00:17:37
Speaker
Yeah. it It is always, you know, comp is certainly responsible prop for probably your largest expenditure across an organization. and it was always kind of – you know, run by some folks in the comp area that are really good with numbers, you know, really good with data, really good with spreadsheet, like really good analytic type people.
00:17:59
Speaker
All of HR doesn't exist in that realm, right? You look at recruiters, you look at um people that are doing um learning and development. You look at people that are doing, you know, handling sort of just general HR.
00:18:12
Speaker
That is not the skill set that generally is found within ah HR. It's that's, the the The COMP sort of skill set is probably more akin to like something you might find in accounting, right? It's heavy data driven, heavy numbers driven. um So the, you know, so the, so just by its nature, it did just kind of, the the skill sets didn't necessarily mesh together.
Pandemic's Impact on Compensation Strategies
00:18:39
Speaker
um COMP has always been important, obviously, it but COMP kind of tended to sort of operate on this own little island almost. And, um It typically was not the place that got the most budget, right?
00:18:52
Speaker
Recruiting is always very sexy and everybody thinks, oh, we got to recruit. We got to recruit. You know, money gets thrown there. Money gets thrown at payroll. Money gets thrown in a whole bunch of areas. Comp has always been one area that didn't really get lots of budget, right?
00:19:05
Speaker
It's important. It's valuable. But you always had to fight for it, right? It wasn't like- It was a really big turning point for- It was a huge turning point because what happened was it changed the dynamic because now all of a sudden the companies were no longer really in charge of driving comp in the way that they had always historically been.
00:19:29
Speaker
Now you had individuals that, um you know, maybe were headquartered in, you know, the the company maybe was headquartered in New York City, but you had somebody living out in, you know, Wichita, Kansas, how are you going pay that person?
00:19:44
Speaker
You're going to pay them based on how the headquarters operates or where the person lives through a big wrench into things. Now, all of a sudden you could start to recruit people. People didn't have to be within a 20, 30, 40 mile radius of the office. You could get that, you know,
00:20:02
Speaker
that, that analytics guy that lives in, you know, Sioux Falls, Idaho, and, you know, who was a rock star, but, you know, he didn't have a ton of opportunities. Now of a sudden, you know, you can recruit him in, he's working for a top rated firm in, in New York city, you know, again, it it, it just really threw a wrench. And then, and now you've got all these, all this, um,
00:20:28
Speaker
you know, all these people kind of moving, right? Changing jobs. There lot of job change going on during COVID. A lot of people coming in, a lot of people leaving. You had retention issues. You had all of this stuff. And then what was the response? The response was, oh, we need to pay competitively. And now all of a sudden the market is nationwide. It's not localized.
00:20:47
Speaker
So that put another input. And so now what you, and then there were signing bonuses. So now all of a sudden you had people coming in at much higher rates um You know, ah you had somebody coming in with a two years of experience getting paid more than someone there that was at seven years. spent Now, all of a sudden, you have compression issues. I was going to ask about that. All at the same time, and it just magnified, it was just a confluence of issues that really drove the the value of having a good, strong compensation strategy.
00:21:23
Speaker
That's what organizations figured out is if I don't have a good, solid comp strategy, Any one of these kinds of influences is going to screw up my whole comp plan. what happens to an organization What happens to an organization when that compression goes unaddressed?
00:21:41
Speaker
Well, you i mean, you know, you're going to have some very unhappy people in your organization. um I mean, you know. You've got guys that are there, you know, senior level experience guys, you know, with five, 10 years of experience at the company who are are all of a sudden getting paid at the same or maybe less than, you know, people are that are rolling in fairly new. i mean, that's not gonna, that's not, you know, you're gonna, you're gonna have some major retention issues. Yeah. ah Good for culture.
00:22:11
Speaker
You know, and you know, then it creates, then it creates problems, right? You got, uh, You already have enough tension when it comes to levels of experience within the same job.
00:22:24
Speaker
But now all a sudden, now you've got people getting paid at different rates. That just adds to that tension, right? And just makes work harder. Yeah, that's, ah it's definitely not,
00:22:36
Speaker
Not ideal for culture, I think, within your organization. um i do want to shift to something I mentioned when we just got started, right? This this angle I kind of find fascinating, the the bigger shift happening where the infrastructure, I think, you know, that the the unseen layer in the stack powering the the platforms, it it becomes...
00:23:01
Speaker
the defensible moat, right? Rather than sitting, you know, on top of it. So I'm curious, like how you see salary.com fitting into that, into that picture.
00:23:15
Speaker
So, you know, I think through COVID, right. Comp had got it, certainly got a boost right in its, um, you know, stronger voice at the table.
00:23:27
Speaker
um ah more attention to ah coming up with a managing and handling and keeping a pay philosophy intact, checking on it, making sure it was, you know, you don't just build, you don't just create the pay philosophy and throw it in the drawer, right? And let, you know, managers hire whoever they want. And all of a sudden, you know, all help breaks loose. um So People are definitely looking at having a comp philosophy now, in in not that they didn't, um but it's it's clear that having a strong comp philosophy and ah and and one that you stick with and one you keep you continue to work with um a helps avoid some of these pitfalls that can happen.
Aligning Compensation and Recruiting Functions
00:24:17
Speaker
So I'm not really answering your question, but um like Like in recruiting, right? if If you give a lot of power to your manager to go out and recruit, they're going to come back and say, oh, what historically in in in in compensation, compensation folks have always tended to want to pay by the job.
00:24:39
Speaker
Recruiters and hiring managers want to pay by the person. Look at this person. Look at their resume. Look at where they went to school. Look at all these great skills. We got to pay this guy top dollar. He deserves it.
00:24:52
Speaker
Yeah, but, you know, he doesn't because he, sure, school is fine, that's great, but it's not really that impactful in this particular job, nor is the fact that he's a PhD and we only require college education.
00:25:07
Speaker
Like, there are things that get people all hyped up and fired up and excited about that may or may not truly impact, um you know, the compensation or you may may have a less of ah a compensable input um into a job than let's say a recruiter or a hiring manager or even the individual themselves, right?
00:25:29
Speaker
And now, um so those things kind of create this tension and put tension against that pay philosophy on an ongoing basis, right? And there's been always this battle between recruiting and compensation, right?
00:25:44
Speaker
The comp folks generally handle the budget. They generally set the ranges. And then the recruiting guys go, oh, those are too low. We can't get anybody. We can't get anybody. And there's just been this constant tension back and forth.
00:25:57
Speaker
It's getting better. And because of the software and because of the tools, there's more communication going on at that level. There's more feedback. you know Recruiting can sort of comment and say, hey, look, this is what we're seeing in the market.
00:26:10
Speaker
Um, you know, this is where we have our range set. Can we take a look at that? Can we make adjustments? So there's, there's more of a two way street going on now than there ever has been. I don't think that really answered your question, but. Well, you, at the end here, you definitely alluded to, to, to some of it. Like, so when you, when you're, when your data is embedded inside someone else's platform, that end customer never really sees salary.com.
00:26:35
Speaker
Right. And, um, but Well, they can. I mean, a lot of times because of the nature of who we are, um you know, we offer this as ah as a as a white label or co-branded. But, you know, I haven't yet to have anybody tell me they don't want our brand on it.
00:26:52
Speaker
Sure. um You know, you mentioned... You mentioned a while ago when we spoke this this promise of integrated talent management, I think, was kind of the way that you referred to it. And and um you know it's interesting, right, because a dozen years ago, right everything was supposed to connect and they did it in different ways. and i'm um you know Nowadays, are we finally seeing platforms move in that more holistic direction? i think so, yeah. I mean, you're seeing some skills, right? So as as part of comp, the other area that's
00:27:24
Speaker
big is is the job description. And, you know, often, to again, this is another back and forth between recruiting and and comp. um Recruiters typically will build out the job rack, but it's usually based off of some kind of a job description that comes out of the comp team. um And, you know, they've set tiers and they they have there are certain competencies that are associated, proficiency levels that are needed. Right. When you do a true comp analysis, you're you're also doing going through what's considered a job architecture process where you're tying in all of your jobs. So they all make sense.
00:27:59
Speaker
Right. The levels are all the same. The competencies that need to be spread across the organization at certain levels are all the same. So that when you get to the pay part, right, you are paying equally, even though the jobs may be very different in nature.
00:28:15
Speaker
you're able to categorize them in a way that, you know, um a manager of, um you know, linemen in an electric company is the same as a manager of the sales team that's outputting and that there's there's um there's consistency among those levels, right?
00:28:37
Speaker
So, As part of that, as part of going through that job architecture piece, you're looking at skills, you're looking at competencies, you're building all of those out. Those are all the building blocks for integrated talent management when you get down to it.
00:28:52
Speaker
And we're seeing a bigger, or bigger push into recruiting for skill sets, paying by skill sets. um or certain skills, right? There's been a big, there's been more of a push in these areas, but harder to do the pay by skills piece than it is the recruit by skills. Yeah. Still, they're both valuable and important. And because there is more data flowing in underneath that foundational level around competencies and skills, the the promise of that integrated talent management, again, makes sense because in effect, you're able to
00:29:29
Speaker
Do that architecture so that you know that a manager is a manager is a manager is a manager, regardless of what department they're in.
HR Data Challenges and Governance
00:29:36
Speaker
You know, a directors director is a director is a director. Again, regardless of what what department you're in. Now you can start to draw parallels ah between and amongst those jobs. And you're able to also then...
00:29:49
Speaker
build those skill sets and the proficiency levels that are necessary for those jobs. And now you can really take something that maybe was a little more objective and make it far more subjective.
00:30:03
Speaker
I want to talk more about this sort of this foundational, this the architecture at this foundational level. Like I, i hear a lot that ah HR data analysis is, is, harder than it should be. A lot of it comes back to job data and architecture not being unified.
00:30:22
Speaker
um I'm curious, and in from you're in your perspective, right what what is broken at that foundational level? And when the job data isn't really aligned or structured consistently, consistently what does that actually cost to people trying to make decisions with it?
00:30:40
Speaker
Well, I mean, that's it, right? And you know you'll see this time and time again. and anybody that has done any kind of implementation work, um pulling data across with some of these systems, let's say you're switching payroll providers, or ah let's say you're trying to utilize, you know, one of the workforce analytics platforms where you're pulling data from multiple different sources. The challenge is it data is data, right? And if you don't have good data governance and you don't have a good process for managing that data effectively,
00:31:13
Speaker
you're never going to be able to draw the um the parallels that are necessary for the analytics you want. And that's the that's where the challenge becomes. Now, the the promise around the integrated talent management piece was, oh, it's all going to be on one platform, therefore all the data will work.
00:31:28
Speaker
That's true if it was done right. but what we find But what we find is a lot of shortcuts were taken. yeah Like ask any payroll provider, they're going to tell you like at a basic level,
00:31:41
Speaker
A job description is not required. A job architecture is not required. And therefore, you have you have all these payroll companies with you know millions and millions of records in there.
00:31:53
Speaker
And in a lot of cases, they don't even have jobscript they don't even have job titles, let alone job descriptions. Like, you know, job descriptions, maybe 18, 20% on any payroll platform will have an actual job description in it.
00:32:07
Speaker
I mean, that's crazy. how can you How can you possibly start to look at parallels yeah and draw comparisons when you're not even even looking at the basic, you're not even starting at a basic level with data. I mean, you're you're you're asking people to you know run a marathon and people showing up in cars and bikes and all sorts of different contraptions to run the marathon and there's no rule.
00:32:31
Speaker
Nobody cares.
AI's Role in HR Analytics
00:32:32
Speaker
Right. Right. I mean, that's really an interesting way way to look at it. Like I never really considered it, we you know, with that analogy, but I i mean, that that makes a lot of sense. And I think that, you know, just to kind of even take it a step further, I know that that AI obviously changes a lot of the the data conversations.
00:32:52
Speaker
um It can. It doesn't necessarily, it can. Talk to me about that. to Talk to me about where where it changes the game. Well, so so data, I mean, AI in its of itself is is a prediction engine. let's' You know, right. It takes data and it looks at it and gives you a prediction as to what you're an answer that you think, you know, that it thinks is accurate.
00:33:14
Speaker
Right. You know, there are hallucinations. Right. We've all heard about those within within um AI. A lot of those I'm not going to say all, but some of those hallucinations are based on the fact that data that was inputted isn't accurate and right.
00:33:29
Speaker
or it's off, or it's mismatched, or that's what creates some of these challenges. So if you have an AI platform that's pulling data that's accurate and strong, great, awesome.
00:33:42
Speaker
It should work well, right? It should be able to predict effectively. But if the data that's going in it or the data that's supporting in it isn't unified, isn't, you know,
00:33:57
Speaker
And i'm i'm goingnna I'm going to change what I just said in a second. But, you know, it's it's the it's the it's the data in, data out concept. If the data going in is garbage, so is the output.
00:34:09
Speaker
Now, AI can help clean that up. Sure. AI can help do a lot of those things and make it better. But at the end of the day, AI needs to be told what the baseline is. AI can't, AI doesn't know what the baseline It's guessing what the baseline is, but it doesn't actually know necessarily. So it needs help and understanding.
00:34:27
Speaker
So it has to be educating. That's why you have these LLMs, right? ah Language models. They help educate and and and and and and help um educate the AI and help it get stronger.
00:34:41
Speaker
I think that, you know, I think a lot of people see it as ah as a threat to data companies. Others see it as an opportunity, you know, you raised a point, um, that I think about sometimes, you know, I think that, um, and this is, you know, when, when, when we spoke before, but I think a lot of customers are perfectly happy with a free answer they get from like GPT or Claude or whatever. And, and others need to know that, you know, how those numbers, you know, how they got,
00:35:16
Speaker
got there and and how they were arrived at. I mean, I'm, I'm kind of curious, like a give me the pros and cons to that. And then be like, who's who's the one that needs the deeper answer? Why is it so hard to get it from those tools?
00:35:33
Speaker
So the short answer is I don't know a hundred percent. I can hypothesize some of it and some of it, I can give you a fairly decent answer, but I don't a hundred percent know the correct answer.
00:35:48
Speaker
But here's what I do know. So, yes, there are folks that will throw a number or ask a question into ChatGPT Claude or what have you and get a response back and go, OK, great. I got my answer.
00:36:04
Speaker
Now, is that the right answer? Maybe, maybe not. Again, data in, data out. So the question is, what are those languages, know, what are those um different you know applications, where are they getting their data?
00:36:20
Speaker
For the most part, it's going to be publicly available data. It's data they scraped, data did they pulled. Now, if anybody's worked with posting data, which we actually have a posting data set as well, you're going to find re it's it's not so easy to kind of pull and aggregate that data because it's it can be all over the map.
00:36:41
Speaker
yeah And so as an AI tool, how's the AI tool going to know what's the right one and what isn't? How are they going to know that you know um you know this is an accurate answer and this is not an accurate answer?
00:36:55
Speaker
It doesn't really know that. I would argue that our salary wizard is probably the the seed for you know a fair number of the answers you might get through AI because it's probably available and and was easy to scrape.
00:37:13
Speaker
I would never use that as ah as ah as a ah product to base my compensation strategy around. Sure. It was created for you know consumers to give them an idea of the basic you know idea of you know a basic salary range. was better than nothing. They had nothing.
00:37:30
Speaker
This was better than nothing. But I would say... you know, in some cases you can then drill down further with AI to understand what is the underlying data. And sometimes you'll get the answer to that. And sometimes you won't.
00:37:48
Speaker
So is, is, is that is AI going to be a problem for us? Sure. You know, there will be customers that go, I don't want to pay you thousands of dollars to access your data and software. I'm going to go get it for free from clock. Okay.
00:38:01
Speaker
How about it? Yeah. Yeah. But you know, How do you base in your pay decisions on something you don't really know? What is the underlying data for? Right. I'm curious how you see the role of comp data changing over the next few years. If we're going to like look ahead for a little bit, you know especially as AI keeps sort of reshaping what's possible and you start seeing a lot more of what you're what you're talking about right now, how do you see the role of the data maybe be changing over the next, or or is it going to change over the next you know a couple of years? and and um
00:38:36
Speaker
I don't know enough to know how the data is going to change. I do know that AI certainly will impact compensation. um I mean, we're doing it already, right? All of our products are are all AI enabled. So instead of walking through workflows like you do in software, you can now have a series of agents do that work for you just by asking the right questions.
00:39:04
Speaker
And so we've built, you know, this, an AI interface on top of maybe not almost all, I don't think it's not quite on everyone, but almost all of our software tools now.
00:39:16
Speaker
So, and we're getting much smarter about AI as a whole, because at the end of the day, you know, if AI can help make a process faster and more efficient, that's a win for everybody. Yeah.
00:39:33
Speaker
Right. yeah And today, you know, when you're using software, you as the user are asking the questions, you as the user are acting like the agent right that that that gets created by ai so if ai can do a lot of that stuff quickly you can move on to something more strategic and even better than you know what was hanging you up so as an example when we created um comp analyst market data the process for you know market pricing a traditional market pricing a job you had to look at you know
00:40:08
Speaker
X number of surveys, you had to you had to go and match all the jobs. You had to like doing that analysis by hand probably took a really good person, maybe like 15 minutes and, you know, maybe on average about half an hour per job.
00:40:23
Speaker
When we rolled around with with Comp Analyst, all that work was done for you. So all you had to do was just find the right job and match it and then you got a salary result. So we took what was, you know, taking people a half hour and changed it into seconds. Did that make that other stuff obsolete? Absolutely not.
00:40:42
Speaker
All it did was freed up time for that compensation analyst instead of spending two weeks doing a compensation analysis. Now they can do that whole analysis within maybe an hour or two in in a day, and then they can move on to the next piece.
00:40:57
Speaker
And they can, you know, there's always stuff that doesn't get done. yeah There's always that next thing. And, you know, hopefully AI will allow us to be more efficient. and get us there faster yeah so that we can do the more strategic right work that we've always dreamed of.
00:41:15
Speaker
I love that. And i I tend to, you know, kind of hold the same perspective on those things. So, you know I can certainly appreciate that on just about every level. um I want to wrap us up with, with um you know, question i got I love to ask everyone. All right, Harry, before you do that, sorry.
00:41:32
Speaker
Yeah. But as it relates to AI, okay, let's let's look at past history. Okay. right You know, 1900s, most people got their information from newspapers, right? The radio came in around what 1920, 1930, 1920s, I think, right? Or maybe even earlier.
00:41:52
Speaker
it was, oh, my God, that's going to blow away newspapers. Yeah. it did No, the two, so the two found a way to coexist. And then television came out. Television was going to knock out both radio and newspapers. Did that happen? Nope. Of course not. right They all evolved.
00:42:09
Speaker
Same thing happened with, um you know, the internet, same concept. I mean, newspapers, they they've They've had how many different technology changes and they're still around? They're still they're still around. Albeit theyre they're they're a far cry from what they were, but they're still relevant and they're still here and
Importance of Data Foundation and Conclusion
00:42:30
Speaker
they're still new. So, ah you know, do I think AI is going to wipe out all this software and wipe out everybody? No, people are going to adjust and they'll make changes and they'll they'll they'll learn to
00:42:41
Speaker
coexist with these new technologies and hopefully utilize those new technologies in a way to make themselves and their products better. I like that. um I like that. And I think you're absolutely right. You know, the same the same thing happened when the internet... um you know, came around and, and everybody thought that it was going to, you know, just, to just about destroy everything else. And it didn't, they all, everybody found a way to coexist.
00:43:07
Speaker
You know, we still have radio, we still have newspapers, we still have all those things. I'll be at, you know, like you said, it's, it's, it's evolved a little bit. It's not the same as it was before, but it doesn't mean it's going to go away.
00:43:19
Speaker
Correct. Um, Kevin, if so if someone were to come to you, a business owner, a leader, an HR exec, or even ah even somebody you know running ah partnerships for for an h some HEM platform or vendor, and they really want to get compensation right and build it on a solid data foundation, right? Yeah.
00:43:42
Speaker
you know, a lot of them, I think, are going be tempted to chase that that shiny, ah you know, AI answer. what Sure. what's What's the single most important piece of advice you give them if you just have a moment? Maybe you're on an elevator right before you get to the top floor and the door is open and they walk right out of there. What is it that you tell them?
00:43:59
Speaker
There are no shortcuts to getting it right. You've got to put the work in. And a lot of times that starts with, you know, doing a cop analysis and doing a job architecture process.
00:44:12
Speaker
project where you're aligning all of your jobs, if you can get all of your jobs aligned and you can get the data, you know, uh, um, you know, synchronized throughout the, you know, all the all the different levels. Yeah.
00:44:27
Speaker
You can then start to really build. If you build the foundation correctly, you can build a major house, but if the foundation is crumbly or it's shifty or it's on sand,
00:44:38
Speaker
It doesn't matter how shiny or how cool or how wonderful your building is going to look like. It's still going to shift and move and probably collapse. yeah So you've got to get the foundation right. And the foundation of all this is good, strong data.
00:44:52
Speaker
And it's it's job data, skills data, um you know going through that architecture. process It's a pain in the butt to do it, but it really pays benefit and dividends when you get it right.
00:45:06
Speaker
And then obviously having good, strong, trusted compensation sources. yeah um You know, if you get if you do all of those things, you're bound to get it right.
00:45:17
Speaker
I like that. Thank you, Kevin. Appreciate you joining me today. Yeah, no, this is great, Curtis. You know, ill listen, I can talk for another two hours, so. We'll have to save it for another. We'll have to have you back. Well, I'll tell you what, if you really want to hear some of the Hollywood stories, we're going to have to have a separate podcast for that one.
00:45:34
Speaker
That's right. Maybe with a glass of whiskey. Yeah, but for sure. Big thanks to ah to everyone watching and listening to Mustard Hub Voices Behind the Build. Be sure to subscribe so you don't miss the next episode. Visit mustardhub.com to learn about Mustard Hub, and AI predictive foresight, how it powers behavioral workforce intelligence for the global workforce. Until next time.