Totally Rewarding Chats | Ep. 15: Defining and Selecting Quality Data Sources in Today's Marketplace

Enhancing compensation benchmarking with quality data

Belinda Roberts, the data and insights leader for the US and Canada for Mercer, discusses the importance of quality data in compensation benchmarking. She emphasizes the need for a robust job architecture and coverage of jobs across the market. She also highlights the value of understanding and participating in industry and local market surveys. Sean and Belinda Roberts discuss the importance of data curation and validation to ensure accurate and reliable data. She also discusses the potential of real-time data, the challenges of implementing it, the hurdles that companies face in obtaining quality data.

In this episode

• Host, Sean Luitjens, General Manager of Compensation Benchmarks, Visier

• Guest, Belinda Roberts, North America Survey Products Leader, Mercer

Episode transcript

Sean Luitjens
All right, here we are, Totally Rewarding Chats. I have Belinda Roberts, how are you?

Belinda Roberts
Great, thanks Sean, nice to be here.

Sean Luitjens
Great, all right. So Belinda, for those who don't know, is the data and insights leader for US and Canada for Mercer. So, which sounds very important and very nerdy.

Belinda Roberts
tremendously important. It is kind of nerdy, I've got to be honest with you. I'm a real data nerd and also a tech nerd, so it's right up my alley. Compensation data and tech.

Sean Luitjens
Okay, well that's... So that's your current job. So give us the elevator pitch and you can take however many floors you need in your elevator to tell us how you got where you are, which obviously is a very American accent, by the way.

Belinda Roberts
Absolutely. Grew up in New Zealand, did my what they call the big OE, which is the overseas experience in my kind of early 20s. And so in the late 90s, I kind of fell into HR on accident. I got like a temporary job working in personnel at an investment bank. But I was working in the HR department and so started doing compensation and benefits work matching to surveys, things like that and so in kind of the early 2000s I moved to Australia and started working with MRSA.

And running their financial suite of products in the Australian market. So I spent about seven years there and then I moved over to the US in 2007 to work on some of our proprietary compensation technology, old e -prism from back in the day. Some of the people have been around a while will remember that. And I came out to help on a couple of big projects that we needed help staffing for six

Sean Luitjens (01:43.384)
of the day.

Belinda Roberts (01:51.52)
So six months goes by, a year goes by, two years goes by, then now, 17 years later, I'm still here. But about five years ago, we moved out of that business.

Compensation tech business and an opportunity came up and I moved over to lead the data and insights group. So my team make the surveys and the data products for the US and Canada. So all the compensation and policies and practices and all those products fall to my team. So some really great stuff there.

Sean Luitjens (02:29.516)
Awesome. noticed, I think I only know like two or three people and I've been around a while who actually didn't fall into HR. So it's kind of funny. And you kind of are dating yourself when you say personnel department. That's the other way.

Belinda Roberts (02:41.364)
I know. Well, it had been the Edmund department not long before that, so.

Sean Luitjens (02:47.074)
Okay, so just to prove you're human, what do do for fun outside of work now that you are not in New Zealand and, you know, here?

Belinda Roberts (02:56.302)
Yeah. I have a 13 -year -old, so not a lot. But when I do have available time, I like to oil paint.

I paint mainly portraiture, so that's kind of fun. And then we try to spend quality family time while he's still open to talking to his parents. And we have a vintage 1971 Airstream. So it's my little piece of Americana, know, the shiny, you know, airplane -like bubbles. Yeah, it's really fun.

Sean Luitjens (03:27.98)
Yeah, yeah. We're big trailer family, so I did not know that. So that's the other reason I'm doing this all the years. was that, like, I don't know how long I've known you, 20 years. And like, I did not know that, because actually that's...

Belinda Roberts (03:42.434)
Yeah. Yeah, we've towed it all the way out to like Utah and up to Maine and we've had some great memories made.

Sean Luitjens (03:52.55)
Well, we'll have to talk about working remotely at some point in time because I've done a bunch of work in there. yeah. Yeah, there are plenty of people camping. Now there's less. Thank goodness. So the reason I wanted to chat with you and kind of go through this

Belinda Roberts (03:55.958)
Yeah, we did that during the pandemic. Yeah. Right.

Sean Luitjens (04:13.782)
We talk a lot about getting data, at least in our software, in our tools, and where we're familiar, everyone incorporates and puts data in. So which begs the question.

What makes a quality data source and how do you find quality data to actually start the benchmarking process, which then predicates the whole comp and comp planning and comp management system, but it's kind of underpins, in my opinion, if you don't have quality data, then the rest of it all just sits on kind of, to be honest, it's on a bed of lies for the rest of it. So how do you, where do you start and what makes a quality data source?

Belinda Roberts (04:51.352)
You know, I think foundationally, the job architecture is extremely important. Coverage of jobs across the market that actually reflect the market that you are working and employing in.

Just that robustness of the job architecture and making sure that's curated and maintained. We at Mercer have a fairly substantial team that is dedicated to maintaining our job architecture and that's supported. It's global architecture. So that is super important to keep it really relevant. I think when you start talking about the data itself.

Making sure you have like your peers are in the organization, maybe even your competitors. So understanding and participating in companies I think is really important too.

Competitors in your industry, yes, but also maybe competitors in like a local market. So those people are competing for talent, not just your direct competitors in your industry. So those things are important. think there's two things there, you know, understand what your peer group is, but also understanding if you need to know locally, if you've got hourly employees, for example, who that those those local competitors are who are going to attract away your employees and who you need retain your employees against.

Good local coverage, think that's important. Rural and urban, you know, as much as you can. Some jobs it doesn't matter so much, but where there are jobs, often times in the hourly space, having good local coverage, location coverage is really important and being able to get down to those areas where people are coming into a plant or a manufacturing facility and things like that.

Data curation, validation, that's pretty important. Just like we spend a lot of time looking after our job architecture, we spend a lot of time validating with clients and curating the data to make sure it makes sense year over year and there's not just sample spikes that are changing things wildly.

That's something we really pride ourselves on, is trying to make sure that we're talking to clients. Does this data set make sense? Is this the right LTI statistic you gave us? And do you have the right revenue there so you can do proper cuts? So as much as we can in the time we have, we try to do as much validation and curation as possible. Honestly, I could talk for hours about the data quality, but there's lots of other things that go into it as well.

Sean Luitjens (07:41.686)
How does a company check though? I mean, so there's the interesting thing, right? Because as you look at every vendor that's out there collecting data and offers data, nobody says they have really bad data. Nobody even says they have marginal data. So how do you, you know, how do you go if you imagine someone who hasn't been around as long, like how do I figure out the best sources of data and then to pile on questions to you?

Belinda Roberts (07:51.746)
Yeah.

Sean Luitjens (08:07.66)
You know, you've got survey data and the difference there between there's now a lot of other providers out there that are collecting data from other sources, payroll, scraping, just a myriad of places. How do you weight and rank all those things as you look at it?

Belinda Roberts (08:15.874)
Yeah. Yeah. Yeah. I think a good question to ask is what's your methodology? Where did the data come from? Does it have a robust and defendable methodology for how the data set has been created or amalgamated? If there has been manipulation to understand what's happened there? I mean, obviously people aren't going to give away their keys or their secret source, but understanding in general what has happened to the data and where it originally came from. I think that's super important.

I think people who are going out there are so many data sets now. Just taking a kind of dispassionate lens. I'm not sure that price is the issue anymore from data perspective, but really just knowing where that data set comes from and understanding how you need to use it within your organization. If you're going to be called to ask to provide a peer group analysis for your executives, you might need to have a specific data set that you can do that for. But if you're looking to do something else and you need data

A different data set might be fine. So understanding what your use case is. And I think we're going to talk probably at some point about real -time data. And that has great use cases for some things, but it's not always helpful for everything if you're doing your basic comp analysis. So...

I think those things like methodology are super important. I also think that understanding the scope of the data, so from a total rewards perspective, does it go all the way through to long -term incentives if you need that total direct? I mean, obviously, base is kind of, base salary is table stakes, but how...

Encompassing is the data, how much data is going to, how are you going to get variable components and things like that if that's what you need to do measurements on variable. So I think those are all good questions for for compensation people to look at when they're looking at data.

Sean Luitjens (10:29.304)
So have you talked to companies about real time data? mean, to me it's interesting, right? I mean, my perspective, which people can take or leave, I think the payroll data is basically base pay, arguably could be better than survey data if you get it in the right place. There's some wonkiness when you talk

Belinda Roberts (10:33.782)
Yeah. Yeah.

Sean Luitjens (10:56.736)
one year total cash, is it in bonus or is it actual? Cause it all gets treated as ordinary income out of payroll sometimes depending on how it goes. So then you get to your point of validation, like what do you want to do? And then I think it gets a little bit tougher as you start talking about benefits, allowances and LTIP. I think it gets really important when you get out to LTIP, you know, cause, cause how people cash when they cash, like it's all different. How long they held, nobody's going to know that out of a payroll system.

Belinda Roberts (11:01.87)
you. Yep. right.

Sean Luitjens (11:24.49)
So I don't know how you weight and rank those knowing you're slightly biased as you go through, but you see a lot of clients.

Belinda Roberts (11:30.402)
Yeah, you know, as a yeah, as

As an organization that data products are a huge part of what we do, we would like to be more nimble in collecting and reporting back data. I don't think we'll ever get to real time. That's not something that we feel comfortable with. There's a lot of laws in place around anti -collusion and anti -trust, around age of data, not forward -looking. We want to stay on the right side of the Department of Justice, and we want to make sure we keep clients safe as well so yeah

Sean Luitjens (12:08.334)
Well, let me be clear when I mean real time, let me, I guess, let me be clear. So if you want to look at the 90 day window, whatever it is, but kind of a rolling ever, maybe evergreen is a better term than, than real time. So real time probably does have that negative like payroll happened yesterday and now I'm going to upload my numbers and they're going to change three days across. No, to me, it's more of an evergreen piece. I guess real time being in that near time, maybe I should, maybe it should be termed near time.

Belinda Roberts (12:16.834)
Yeah. Yeah. Right time we like to say.

Sean Luitjens (12:38.325)
okay, I'll take that.

Belinda Roberts (12:40.162)
Yeah, so RightTime gives us the opportunity to have more of a subscribing type of a model and still stay safe within the bounds of the restrictions. So that is actually something we're looking to do. We've got some products that are already based on that rolling collection and they publish quarterly. Our Comptrix product, we would ideally like to move a lot more of our database towards that.

Without making our clients give us data more often, because that's something we don't want to ask and they don't want to do, which totally makes sense. It's a ton of work to participate in a survey. So those are all areas we see a lot of value in, is getting more regular data to our clients. And it's a lot of where we're spending time focused on our CAPEX funding and spend is to try to solve for that problem and get more regular data.

I do think that there is, I think it's valuable, the information you could get from a payroll provider. I think we just have to be a little cautious about the ability of a comp department and comp structures to be able to handle real -time data or right -time data. You know, it's hard enough to do a once -a -year structure review.

I think there's a lot of value with hiring and other hot jobs to understand how things are moving really quickly. But in general, I think it's a more limited use case than that. I still think the majority of comp teams can only really handle reviewing their data once a year en masse, except for those anomalies with the hiring and the hot jobs.

I absolutely want to go that way. I think we need to go that way. But I think there's some caution around how much you can actually use real -time or real -time data.

Sean Luitjens (14:38.03)
So you mentioned a couple of those, but with companies you've seen struggling to get quality data, right? Because I think, you know, I always kind of term it, know, blue skies, rainbows and unicorns version of we've got the right data and all the right jobs and all the right things and everything lines up and I can submit on my data and my architecture lines up to your architecture thing. But if you had to name a couple of the biggest hurdles of where companies struggle, you know, kind of getting and maximizing quality data, what would those be?

Belinda Roberts (14:47.842)
Hahaha. Probably knowledge of the data world, know, small to mid -sized businesses, they've often not heard of MRSA and our traditional competitors. And there is a lot of other options out there. I'm not saying any of the options are bad. I think all data is good. It's total data democracy. And I think for some of those data sets, it's good enough for a lot of organizations. I think as...

As organizations get bigger and bigger, they need a more specificity in what they are needing. They get more specialized roles and oftentimes that's when they start to need to get into a peer group analysis or really hone in on an industry specific job. So I think it's almost related to size of the organization. As they come into the market, they start to understand that they need data in general and maybe don't have any comp

Team don't have a lot of comp specialty and so they're using it in a fairly limited way in the beginning and as they kind of move through and they start using compensation data more and more then the appetite for a more what I would call highly curated data set which would be kind of like a survey from a Mercer that's when that starts to become a lot more useful to them but sometimes that's overwhelming for a smaller company too, right? So our job architecture is over 30 ,000 jobs. And so that's a lot.

Sean Luitjens (16:39.598)
Well, that's, that's what I was going to say, whether you're at Mercer's IPE or Haypoints or something like that, just reading the instructions when you've got, you know, 500 people, you're like, there's this is, I'm not going to do this, but do you think.

Belinda Roberts (16:44.14)
Yeah. Yeah. Yeah.

Sean Luitjens (16:54.176)
Okay. Yeah. Cause to me, I think the hardest place for people is I think in the hundreds, again, by size, I agree with you by size and the hundreds good enough. you know, they're really doing a gut check. They don't have structure. got one or two people. don't have to worry about stuff in the thousands. They've got comp teams that can kind of really dig in. But how do these companies that are big enough to have one comp person maybe, or maybe a comp and Ben, and, know, I'm an American, so I'll tell you benefits then probably is, you know, three quarters of your time or more.

Belinda Roberts (17:04.162)
Yeah. Yeah. Yeah. Yep.

Sean Luitjens (17:24.494)
because we just make it difficult in the US for benefits people.

Belinda Roberts (17:28.236)
Yep.

Sean Luitjens (17:30.222)
So it becomes, think how, how tech can and how we can overcome the hurdle, whether it's tech or manual or instructions to allow them to match. like do I have an accountant one or a four? I have an accountant, so I can't even rank order my people. I've got an accountant. don't, can't tell if they're one or a four or three, which is actually an accountant 24 over at Hey Group and an, and a something else somewhere else. And how do I compare?

Belinda Roberts (17:36.952)
Yeah. Yeah, you mentioned two things. think technology is a real savior for those sort of small to mid -sized clients who aren't big enough to have a large complex comp team who are going to deal with a big compensation management tool and load their couple of hundred global surveys in there. Having good data baked into those is really helpful. It takes away all of that need to be managing a survey library and trying to get it into your tool and all of the things that you have to do to get to a place where you can use the data. But also the job matching piece of it.

And we are spending a lot of time right now on building our AI to help us and our clients match to our taxonomy, our architecture, but also others. we really hope that that can sort of save a lot of that very transactional and foundational work when matching to a data product or a data survey. Because for us, that's really a lot, it's a huge time suck.

It's just massive. And really, they spend so much time matching that they don't have time to do analytics. So those are the kind of things that I think recommending people get their data via a technology product, if they're smaller especially. mean, obviously, the bigger they get, they need a technology product. But yeah, that's really helpful. And then I think we can come to market with some kind of AI -driven job matching that would be fantastic.

Sean Luitjens (19:31.502)
I think that's gonna be the first and foremost place that, not to give Amazon new props, but anything else, but companies like you matched like this for this role. And so you're like, okay, that kind of makes sense. Or it's more important to me, I'll move it up or down, but we kind of in comp planning when we give recommendations call it an anchor point. But to me, it's kind of like, of all 30 ,000 jobs I can match, give me five.

And this is the one we think is right to start. so, you know, that's not nearly as overwhelming as go find a, you know, job function family and then grade a job, you know. I think that's gonna be, I think that's gonna be huge. I think the other thing personally with tech is gonna be super cool. You guys have all the history over the years. So even people with near term.

Belinda Roberts (20:00.642)
Yeah. Yeah. Yeah.

Sean Luitjens (20:20.91)
or right time data or recent data, real time data, it's fairly recent with that piece. so being able to catch hiccups and kind of be able to map out what might be a hiccup and what traditionally does happen actually. So as you're hiring up the curve, if you hire at the top, it's going to take you four years to get out from under that when supply and demand levels out. If you hire at 75%, those people stay. I think that stuff, because you have the the survey providers also have the,

Belinda Roberts (20:26.882)
Mm -hmm. Yeah.

Sean Luitjens (20:50.914)
incumbent over year over year, which is super cool data.

Belinda Roberts (20:53.23)
Yeah, yeah, and actually that's exactly what happened over the last couple of years with the hot labor market is people just hired in what they needed to hire at and now they have these huge compression issues that they're still trying to dig out of where they've got you know people getting paid more than their manager and things like that so

Technology hopefully will get more sophisticated to be able to allow for those types of analytics and right now a lot of the time it's more just benchmarking and you know we need to make benchmarking more automated so we can move on to more analytical and insights generation.

Sean Luitjens (21:30.734)
Well, I think it's the data insights, but I still think it's the data submission. You you put the data in and then, you know, and actually where you match and then recommended peer cuts to your point, you know, an hourly person at, you know, working at a big box store is just as likely to go to a warehouse or a restaurant for a dollar an

Belinda Roberts (21:36.685)
Yep.

Sean Luitjens (21:50.048)
as anywhere else, right? And you can start to do that matching and watch across and let them match and suggest peer groups as well, you know, and where they should and should not have peer groups.

Belinda Roberts (21:58.156)
Yeah. Yeah, and actually that's a really great point, that whole concept of moving with the hourly world. have a data product that allows you to kind of look at all of the hourly data we have, which is substantial, it's in the 8 to 9 million records range, and say, I want to price a labor pool versus a job. So we had companies saying to us, I've got to price a job in my laundry, I'm a hotel on one side of the street, but I'm competing for talent with the hospital across the road for laundry

And really it's not laundry workers, it's someone who's willing to do that job at that location in a not overly skilled way. You learn everything on the job, you want to work in a warm environment with air conditioning and that's the job profile. So our tool sort of lets you say, give me all the data for all of the jobs with this skill set who work in this market and it can be hyper local.

From like a zip code perspective and say, tell me how much they get paid and not tell me how much a laundry worker gets at hotels in the city. So it's really a different way to look at it. It's kind of like taking a step back and saying the hourly pay strategy is a little different to the salaried pay strategy a lot of the time.

Sean Luitjens (23:20.686)
I take that one step further. So I use the example of accountants all the time, right? And, and company size, because a lot of companies will say, well, I have to benchmark against my company. I'm, you know, I'm a hundred, 150 people. Okay, fine. Then you think you're, paying it, you know, 70 % comp ratio, right? And, but you're losing everybody. So you think the problem is something to do with your environment, but when you benchmark against your whole area for an accountant, somebody who's portable until they get to a

Certain level and, you know, create dollar, you know, demand dollars for certain skills in that industry. You're like, actually in the market, I'm paying 40. And so I'm not as cool as I thought. This is why I'm losing problem because then they can't work the right problem without if they, if they choose the wrong data set and look at that data the wrong way, they start to functionally try to solve the wrong thing.

Belinda Roberts (24:01.741)
Right. right. Yeah, yeah, and that raises a really good point. As we move on through the next three to five years, I really think the data landscape is going to become a lot more complex. We're creating products that have base through LTI total direct, but.

We're gonna need to understand demand data, supply data, and how that correlates to compensation. How much do gig workers get paid in that job, maybe in that location? There's so many other lenses that we need to apply to the compensation data now to say, what's a real competitive index? To your point, you might be paying at the market ratio of 100 or one or whatever, but if there's high demand for that job in that market or that specific job,

And you're having a really hard time filling, maybe you're not at 100%, right? Like maybe you're 70 or 80 and you need to pay a premium to offset the demand side of things. So we're trying to think through. Mm -hmm.

Sean Luitjens (25:12.674)
That's what they're near time data is interesting, right? Because you've got this long window of data that you have that's collected and very highly curated. And then you've got this data that you can, I mean, we term it interpolation, but you you can start to figure out actually, you know, that's a good buoy number, but I've got a little bit more of a supply and demand lens on what's happening right now. Doesn't mean I need to move, you know, I just need to adjust the other number. Don't get married to one.

Belinda Roberts (25:21.528)
year. Yeah, yeah, yeah, yeah. I think that's kind of where we see the data landscape hitting is ever -growing complexity of data sources and data inputs. Hopefully the technology will keep up with that.

Sean Luitjens (25:56.172)
And I also thank consumers, whatever you want to call them, employees.

Job seekers, you everyone's getting smarter or so because the data is more readily available. And so things like, you know, not, you know, what are my total rewards for the year? They're starting to, I think, get better, especially where the employment market is balance -ish, you know, you argue, you know, the numbers, but it balance -ish that, you know, I might take less money here because the benefits are a lot more than what is that. And so the total looking at understanding your total rewards package. And obviously I come from the

Belinda Roberts (26:19.362)
Yeah. Yeah.

Sean Luitjens (26:30.064)
background so I'm always a little biased. But I think consumers, you want to call them that, people job seekers, employees have gotten significantly smarter about the value of what they're getting and their benefits and so companies have to start looking at that. You can't be good enough and be like check the box anymore. They're doing the math now.

Belinda Roberts (26:51.02)
Yeah, yeah. And even so far as to say the personalisation of total awards, I think, is going to happen whether companies want to do it or not. I think they should try to do it. I think it would be very valuable for their attraction and retention. Yeah. Yeah.

Sean Luitjens (27:08.074)
equity too. It'll be important. What will make that really hard I think is you know the TPAs, the third -party administrators, and all the people on the Ben side.

Belinda Roberts (27:18.99)
Mm -hmm.

Sean Luitjens (27:19.054)
to be able to pick and choose and move stuff around or whatever there's going to have to be. And I think tech's moving so fast, I agree with you. And once there's a little bit of a critical mass, everyone will kind of jump that way. Because I think from an equity standpoint, how much cooler does it get? You know, you get $80 ,000, everyone in this role gets $80 ,000. How you spend your $80 ,000 is a you problem. Whether you want to use it in benefits or cash or, you know, whatever it is, that's actually a pretty cool premise. You guys will be able to watch how that kind of

Belinda Roberts (27:21.976)
Yeah. Yeah. Yeah. Yep.

Sean Luitjens (27:49.269)
because the numbers will bounce around for all the incumbents, but the jobs at that company will all pay the same. That would be really cool to see.

Belinda Roberts (27:52.673)
Yeah. Yeah, and it's super interesting to watch the pay transparency laws roll out too and watch organizations kind of get on that whole underlying communication around their pay strategy bandwagon, right? So not only employees can see how much the pay ranges for their jobs that's being posted on the internet, but job applicants, candidates can too. And so there's still a huge disconnect between what they see posted online and the reality of what makes up that range and how does that apply to them. you know.

Sean Luitjens (28:33.25)
Yeah, there's no guidelines there yet on it, which I think is really interesting. You've got companies who are like, I'm just going to put it out there because I have to. got, you know, and then you got others who are like, actually think about it. If I'm in a high demand region, I'll do my total comp package. I might have to explain that to some angry employees, but actually I just need butts in chairs. And so I'm going to add it together and I'll deal with that issue of that's not base pay when they show up. Cause I can't interview them if they don't show up. Versus the other one, do I make it actually lower?

Belinda Roberts (28:51.532)
Yeah. Yeah. Right. right.

Sean Luitjens (29:03.524)
and try to hire someone more junior and make everyone feel good about themselves that in their current role, that they're like, crap, actually the market outside is this way. It's an interesting thing from a, crap, we have to do this to how do I position it? What can I do with it? How do I communicate it thing? And you know, it didn't take long for people to start thinking strategically about what that is.

Belinda Roberts (29:08.546)
Yeah. Yeah, I agree.

Sean Luitjens (29:27.322)
it's always it's comp like, cause I came from talent acquisition comp, a lot of things you're like, how hard can copy and how often bend data be, you know, when I got into that space and it was like, how hard can it be? Right. I get paid. Like, I understand a paycheck. And then you realize like, holy crap, it's complicated. And then of course it's made more complicated because you get out of all the different humans that are involved in it and our organization.

Belinda Roberts (29:36.311)
Hahaha

Sean Luitjens (29:56.)
And then you add the external labor market that's putting pressure and you don't have as much control on that. it's, mean, like you, I fell into HR or as this guy, Paul Ronan at Novo Insights is who I heard from last, but he was like, you you don't choose comp, comp chooses you.

Belinda Roberts (30:14.414)
I think so.

Sean Luitjens (30:17.516)
Yeah, and you know, very few people basically get out of high school and go, you know what I want to do? I want to be a comp professional or comp and bed professional. But somehow they get into it. And then I think people just get jammed hooked on it like you. You're like, you know, this is, this is cool stuff.

Belinda Roberts (30:31.852)
Yeah, no, mean, it wasn't even a thing actually when I joined into the HR department, personnel I guess. You know, when I moved out to Australia in the early 2000s, the whole concept of being a compensation and benefits analyst was relatively new. And so, you know, I'm really dating myself here, but that is evolved very steadily over the years and I feel like we've just hit like a huge ramp up like it's going to get a lot more sophisticated a lot more quickly complex and sophisticated but somehow it also needs to be simpler.

Sean Luitjens (31:10.924)
Well, but the term I use a lot and people for me say, you know, believe humans are highly statistically predictable, but individually amazingly unique. so I don't think like comp teams are going to have to be more strategic and you can get more tech and get a lot more science behind it. But the art of how they use that to drive their business and align to their business is going to get more complex as, you know, I call them consumers, but the end users, you know, get more data as, you know, the senior people in board get more informed like comp becomes so I think it's tech is not going to keep up with the demands of it it's more than half of the spend at most companies like you can't hide from

Belinda Roberts (31:42.039)
Yep. Oh, for sure, yeah. And we have good analytics about the malalignment between budgeting for merit and actually what gets spent. And that causes a lot of damage to organizations because it's a lot of money if you miss by 1 .5%, which has been trending the last year or two.

Sean Luitjens (32:06.004)
Yes. Yeah, especially in low margin businesses, they just can't afford, know, when you've got a 5 % margin business and you miss by a point, that's a big deal.

Belinda Roberts (32:15.18)
Yeah. And another reason why you need good comp data, because you need to know how much you need to pay. Because if you don't, then you're going to have a retraction and retention problem, and you're going to have that overspendage at the end of the year. yeah. Right. Exactly.

Sean Luitjens (32:19.704)
That's true. Pennywise and pound foolish, right? The question I always ask near the end, it doesn't have to be data, but if you could automagically, a highly technical term, automagically solve one thing in HR tomorrow, what would it be?

Belinda Roberts (32:44.59)
Well, I'm going to stick in the comp world because that's my jam, but also technology. It's nothing new than what I haven't already talked about. I'd take the friction out of the process, whether it be creating the data products, like the ingesting data from clients or systems. I'd make that friction free. I'd also make it friction free for them to consume the data, direct feeds into whatever systems.

Honestly, if anyone wants to come and talk to me from a comp tech company about how I would design some of the systems to use the data, I think that's where I would add the improvements to take that effort and foundational functional work off of the comp people so they can do all the things you talked about, that they do more sophisticated analytics and spend more time actually analyzing the results versus...

Preparing the system to do benchmarks, right? So that's what I, if I automatically could change something, that would be the three kind of key areas I'd focus on, know, reducing friction in and out, and then making the systems do most of the work for the client, the hard, the heavy lifting. I know.

Sean Luitjens (33:59.799)
That's so nerdily practical, Belinda, but that's awesome. And I agree, obviously, we think analytics is where it's going to be in being able to basically maximize spend, right? Which they use analytics everywhere else and they create time for that. And so I totally agree. Like, can we get more time for people to do it? And I think analytics, even with machine learning and AI, every company is different. So you're going to have to be able to figure out.

Belinda Roberts (34:04.748)
That and world peace.

Sean Luitjens (34:27.232)
Use tech to get to some general point, but how do I tweak those for my business to work in my area, for my industry, for my everything is gonna be super unique.

Belinda Roberts (34:38.722)
Yeah, no, I couldn't agree more. Yeah!

Sean Luitjens (34:41.144)
So this has been great. So I really appreciate you coming on. We will have when we get this out, Melinda's contact information will be around. We'll tag her when it's posted. You heard it. She is happy to talk with people about data. She's easy if you say you want to talk about data. Yeah, yeah, that might be. You might want to aim at the call line, which is awesome though.

Belinda Roberts (35:01.454)
You can't stop me.

Sean Luitjens (35:09.816)
Talking to providers of data, people who are just so excited about it, you just know, you know, there's, there's, you know, solid data back there because it's, it's keeps you up at night, which is super cool.

Belinda Roberts (35:21.462)
Yeah, I think we're heading into an exciting time in the next five years. So I'm excited.

Sean Luitjens (35:26.294)
Awesome, well thanks so much and for those others, we'll see you the next time.

Belinda Roberts (35:30.242)
Thanks for having me.