Why skills-based pay is finally within reach

[00:00:00] Join us on a journey where we unravel the latest trends, tackle your burning questions and explore innovative strategies that are shaping the future of compensation, all with a coffee in hand. Hello everybody and welcome to another episode of Comp and Coffee. We are back from World at Work, Top of Awards. If we saw you there, it was great to connect with you.

[00:00:24] I had many people coming up to me over the three or four days that we were there saying how much they enjoyed the podcast. So thank you so much for taking the time to connect and let us know. We are ready to start delivering some more great content for you through the podcast and start recording some more great episodes for you. And today, my guest, one of my favourite guests, am I allowed to say that, is joining me. She'll introduce herself in a minute. You'll recognise her.

[00:00:54] We are finally going to get to do an episode on one of our favourite topics of debate, which is skills-based pay. Now there was a little bit of, maybe a couple of sessions talking about skills-based pay at World at Work, but it wasn't one of the headline topics, I would say, in terms of what was being discussed. But, you know, skills-based pay has been a long-standing ambition in compensation strategy.

[00:01:18] The idea that organisations should pay people based on their capabilities rather than the responsibilities of their jobs. We've never really seen it take scale. It's not a wholesale adopted practice. There are pockets of it. I heard it described the other day as one of the reward unicorns, so the things that, like, we just never get to. And we're going to dig into why shortly, so my guest and I will discuss that.

[00:01:43] But we are at an interesting moment because we are seeing a surge in demand for new types of skills, particularly related to AI. Recent research from PayScale in the Compensation Best Practice Report showed that AI skills are appearing more frequently in job postings, but they're not always commanding a consistent pay premium, and we'll talk a bit more about that as well. So skills are becoming more visible and more important, but we still see that pay practices are trying to catch up with that visibility.

[00:02:12] So today we're going to talk about why skills-based pay has been so difficult to implement, what the data says about AI skills and pay premiums, the disconnect between skill demand and compensation, and how AI is changing some of that, and what needs to happen for skills-based pay to scale. And drumroll, my guest today is Sarah Hillenmayer. This is not your first radio. You've been on this podcast a few times. Please say hello to the audience and tell them a little bit about yourself.

[00:02:41] Glad to be back. Thanks for having me today, Ruth. I'm excited to talk about skills and skills-based pay with you today. My name's Sarah. I lead the AI and data science teams at PayScale. My background is very technical on the AI and data side, but I've been partnering with Ruth and the other pay scalers for the last several years to take all of that technology and apply it to compensation. And this is, skills has been, yeah, that reward unicorn for our domain,

[00:03:09] but for me personally, almost a technical white whale that we've been trying to solve in various ways. So I'm excited to talk about the progress and the breakthroughs in the last few years that have given us a little bit of purchase on that problem. So let's go back in time and level set for people who may not be familiar with what the concept of skills-based pay was. Now, I've been around for a long time. I wasn't probably around when the first kind of skills-based pay started to take shape,

[00:03:37] but that was in the 1970s, very much focused in manufacturing and paying manufacturing workers for the number of machines they could operate. So that was one of the first instances we saw of skills-based pay. I was around in the 1980s and 90s, or probably early 90s, I remember it, early on in my HR career, where we started to implement competency-based structures. So we started to think about what went into a person's job, what were the skills that they took in.

[00:04:07] So this was more focused on white collar workers or non-manual workers, and thinking about the behaviours, knowledge and skills that it took to do a job, rather than just hard skills and what were the tasks that they were doing. So I remember building competency structures, and then thinking about how we would use them. We didn't actually build pay structures around them, but we used competency assessments as part of someone's performance review, and then that kind of went into the performance-based pay equation.

[00:04:34] So we know, you know, organisations then more recently have been trying to move away from rigid job architectures towards more flexible skills-driven models for years. Thinking about the gig economy, you know, that's been around probably for a decade now, but people were starting to tune into the concept that we would have a gig economy inside an organisation, and people would be deploying their skills to tasks or projects, rather than kind of sitting in the concept of a job.

[00:05:04] Again, that hasn't happened wholesale. I know it's happened in some organisations in terms of how they've structured their work. So, you know, we still haven't got to skills-based pay being something that we all adopt within our organisations. So let's consider, like, why has it been so difficult to implement in practice? Do you have a view on this, Sarah? Yeah, I do.

[00:05:27] And the first part is that it's been tough to, as a society, define skills in a way that's really consistent across organisations, and even across roles within a single organisation. What does leadership mean? Or what does being a good writer mean? Or what does being mathematically oriented mean? Or being analytical mean?

[00:05:51] From a data perspective, that creates the same kind of challenge that you could think of with job matching. One company says this role is a Coca-Cola technician, and one company says this role is a Pepsi-Cola technician. In order to get meaningful data about how much a soda technician makes, we have to normalise that and understand that that job is the same across companies. The same thing is true with skills. So we have to figure out how to do that, that kind of normalisation.

[00:06:20] And it's a little bit more challenging than jobs, because there's this idea of proficiency, of like, sure, you might have a skill of leadership, but are you like Barack Obama leadership? Or are you like Sarah Hillenmayer leadership? Like, those are different scales and different levels of proficiency. Unlike certifications, which is wrapped into this whole question as well, where there's a really clear indicator of, yes, you have this certification, or no, you don't.

[00:06:46] A lot of the skills that we think about as being really the things that drive an employee to be a very impactful worker and drive really good outcomes are tough to measure, and then also tough to normalise across organisations. So if you can't measure it, then it's hard to, like, put a price tag on it, basically. And that's been the problem we've seen. I mean, we've seen what probably, like, the whole skills hype was probably, like, four years, four or five years ago,

[00:07:15] where we know a lot of large, very large organisations were starting to think about building skills taxonomies. But that hasn't really trickled down. Again, it's not something that we're seeing as wholesale practice. We did see a shift last couple of years, people, like, hiring for skills. So people were thinking about, OK, in my job posting, I'm going to be more specific about the skills that I'm going to put in there.

[00:07:39] But I think until we see, like, the definition of skills as it relates to, like, let's stay with the job concept for now, like, to everybody's job, then you can't really, you know, then start to think about putting a price on that. So that's been one of the biggest problems, I think. What else has caused problems? Is it a data problem? Is it a technology problem, do you think? I do.

[00:08:03] I think part of the slowness to build this sort of technology or even collect this data has roots in the same challenges behind the rest of the compensation technology industry. We've traditionally been using very slow types of data collection methods. You run a survey, get people to participate. You normalize the data. You send it out six, eight months later.

[00:08:32] And then the act of using that data and pricing jobs has also been historically pretty challenging and takes a lot of time to do. And so because all of that is hard, basically, or has been historically hard, I think the capacity to do it faster and to add more complexity to it just really hasn't been there.

[00:08:57] And as we get to, as the technology in that space is exploding and we're able to collect data in more real time and able to normalize it more close to real time and able to make some of those compensation decisions in more real time. That's a big unlock for something like adding complexity. And in the case of skills, it's even more critical because those skills are a little bit ephemeral.

[00:09:23] They're moving faster than a grade for a job family, for instance. Pay for a specific job even, or if you think more broadly, for a role at a specific level. Doesn't move quite as fast as this hot new skill that you might be hiring for right now that's emerging.

[00:09:43] The speed of the tech and the speed of the work hasn't really been fast enough and efficient enough to allow for skills to be a part of that story at scale. And I think we're starting to see that change, which is really exciting. Yeah, I mean, if I think about sort of the data sources that are available today now, you know, vendors who have these large skill data sets that they're scraping from everywhere.

[00:10:12] Like Cast is one example I think of, you know, that's a game changer in terms of being able to extract, as you say, skills information at scale more than skills information specific to an organization. Because as you say, it's those universal descriptions of skills that we need in order to be able to like price, you know, jobs across the board, really. Absolutely. Yeah. The scraping technology.

[00:10:38] And as you noted, I mean, go back a little bit further, like posting jobs on the Internet, right, in a place where they can be scraped makes a big difference. The fact that we're seeing a cultural shift in how we describe jobs and job postings to include more of those skills, or sometimes the names of the tools that you either that you expect a candidate to come in having used before or that you know that are going to be part of their day job and naming those by name.

[00:11:06] And then the ability to use AI to extract that information at scale from scrapes job postings has really changed the game on what we can learn about which skills are important for which roles at which organizations and normalize those. So we have data that we've never had before. Then, of course, as the AI like impact here, you know, we're seeing a surge in demand for AI related skills across the labor market. But some of our data shows that those skills don't always come with a consistent pay premium.

[00:11:36] Can you tell us what we saw in the compensation for practice data? Yeah, sure. And one of the things that, well, what we're seeing in that particular survey. So we surveyed about 3000 HR and comp professionals. Many of our listeners filled out that survey and we asked if they were paying for AI skills and how they were doing it. And as Ruth noted, we saw a few organizations that are, but a lot of organizations that aren't.

[00:12:03] And I think the aren't category isn't doing it because they don't know how. And so they're a little bit just hamstrung that this feels important for many organizations, not just the ones that are paying for it. But they're not sure how to incorporate that into their compensation strategy and philosophy, whether that means uprooting their job architectures and their job definitions, or it's kind of an add on, whether that goes into someone's base pay or their bonus.

[00:12:33] How to even make the decision about how much those particular skills are worth. So there's lots of challenges there and folks haven't uniformly adopted paying for those extra skills, even if they're hiring for them, even if they're really looking for those skills, because they haven't really figured out how to do it. Yeah. And I also think, I mean, you know, the whole impact of AI is influencing the world of work at pace.

[00:13:01] I see at the moment, I talked about world of work, total rewards at the beginning, you know, the difference in how we were talking about the deployment of AI and how it will restructure the work that we do had moved on significantly from 12 months prior when we were probably more sceptible about AI. And it was more this kind of like magical thing that was going to happen.

[00:13:26] And we weren't quite sure what it was, you know, whereas, you know, we now know that many of you listening, many CHROs, many like total reward folks are thinking about how do you restructure the workforce around one that, you know, you're co-working with AI and how does that change the work you're doing? And what are the like the new skills that we have to start thinking about? I heard an example the other day, customer talking about how they're doing kind of like a skills audit.

[00:13:54] So they're thinking about the work that people do today. And then they're thinking about what they know of, you know, how that work will change with or with AI, the impact that AI will have on that. And then what are the skills they think they will need in the future? And then kind of like doing that AI audit.

[00:14:12] So I think long story, but generally I do think, you know, as we think more consciously about how work is shifting, then we start to it's making us think more intensely about the skills that we will need in the future. And then that helps with this whole discussion of like the definition of skills and then having those definitions that will fuel how we pay for those skills.

[00:14:35] One of the things I've noticed in speaking with some of our customers, which aligns with what you're saying here and gives me some comfort and enthusiasm for the world as we're headed towards it, is let's say a year ago, year and a half ago, a lot of organizations were hiring for AI transformer type people.

[00:14:56] So can you come into my organization and pick the right tools and get my team going and change the way we do work because we're scared that we're missing out. And so they were paying a really significant premium for these folks thinking that it would upskill their whole team and transform their workforce. And they'll be able to generate 10 times as much product and sell it for 10 times as much money and make bank.

[00:15:23] And what you're describing now is of this audit and thinking about each role and the skills that it needs and what skills that workforce has and how to upskill them. And that's becoming more of the story that I'm hearing now, in part because the supply and demand of AI transformer type people isn't what it needs to be in order to keep up with the demand.

[00:15:50] And so folks that are experienced in change management or helping an organization learn new skills, those people are still being snapped up like delicious little minnows in a barrel to come in and help think through what it means to upskill their workforce with these new techniques and do this audit.

[00:16:08] And I'm hearing more organizations thinking about not just how do I hire these new people with these skills that I desperately need, but how do I upskill my employees, incentivize my employees to learn these new skills? And how do I make sure that they're getting value out of that, the technology and the skills that they're developing?

[00:16:30] So it's more growth oriented, I would say, in the whole right now compared to where we were a year ago where it was like, oh, somebody just fix this for me and I'll pay whatever it is. Yeah, tell me what I need to do. And we are therefore seeing new job categories emerging. So instead of just an engineer, we're now needing an AI engineer or like if I think about within the marketing function, it's not just a content creator, it's an AI content creator. Are you actually seeing that in the data sets? Absolutely, absolutely.

[00:17:00] My holistic take is that the world of jobs is moving faster now than I've ever seen it move. The definitions of what a role is are moving really quickly. So in some ways, skills helps us accommodate that really rapid shifting world where, sure, it's a marketing manager and now it's an AI marketing manager. But in order to ground that in your systems, you might think about that as a more standard marketing manager who has this AI skill for a while.

[00:17:29] And as we're in flux in this transition of more and more parts of the white collar world, but also the physical world, leaning on AI tools, I expect that those skills ultimately will get absorbed into those roles. A marketing manager will just come with AI skills. It won't be an extra thing that they're learning or that they're bringing to the table.

[00:17:53] But right now it's changing so fast that thinking about those extra skills beyond what is typically encapsulated in the role definition gives organizations a chance to keep up with that rapid change and find data in a way that they haven't been able to for emerging roles right now in the same way. Yeah, because you come back to that. I'm seeing tons of emerging roles. Yeah. I don't know which ones are going to stick around, but they're...

[00:18:21] Yeah, I mean, it's just the pace because it comes back to that problem of annual salary surveys. You know, if you're using annual salaries, if I think about how much, you know, we've even changed the way we work internally at pay scale, like co-working with AI in that period of time, that won't show up in a survey that was done like last year, basically.

[00:18:43] And then even if you capture it this year, in six months time, as you say, like the level of everybody's kind of base level of AI competency should be higher. And that's kind of just becomes baked into the job. And so like, how do you keep track? Like if you're hiring someone today, how do you even like try and manage to price that? Yeah, I think it's just too slow to keep up with the changing world.

[00:19:10] We think of annual salary surveys of maybe, okay, let's say a job came out in January and you're collecting data in March and then processing and data is released in August, September, October, somewhere in that range. That's a, you know, seven, eight month delay, but it's, it doesn't actually work that quickly because the vendors have to add the job to the taxonomy.

[00:19:35] And then you've got the whole participation cycle where employers are matching to that job in order to populate the data. And then all of that other stuff happens. So it really does seem to take a couple of years for most survey vendors to capture those emerging jobs en masse, unless they're very niche and focused on a, on a specific area. And so comp pros that we talk to are floundering. They're like, well, what do I do?

[00:20:01] What do I do with this job that didn't exist in my survey? And how do I handle that? So what do we need to do like to get from jobs to skills? What needs to change? If an organization wants to move towards skills-based pay, we've already talked about it needs better data. And it also obviously requires a shift in how compensation is structured. So any recommendations on what an organization needs to do to move from job-based to skills-based pay?

[00:20:27] I think the first thing I would do is, is what the client that you talked to last week did. You'd think of, do a skills audit. Think about the skills that are actually aligned to your business strategy. And if you're starting this transformation, you can start with the ones that feel like they're making the biggest impact. So where, where are the places where that talent is really key and existential to your organization's survival?

[00:20:54] And the, the part that's challenging there is thinking about, we're used to measuring the value of a, of an employee on the impact that they're, that they're making and the output that they can produce. And really the skills-based framework and the challenge that folks are having is that they're trying to tie together the, the presence of this skill with the impact that they think it will make later. Right.

[00:21:21] And sometimes that's obvious, but sometimes it's less obvious. And so the, there's, there's some heuristics and things you can do, but I would say for those skills that do feel existential, really trying to get as crystal clear as possible on what you think the impact of the business will be. If you can hire or train those skills within your organization.

[00:21:44] And then the next step there is to, to, and I'll have you talk about this actually, I think is to think about how that fits into the notion of a job and a home. Yeah. I mean, that's interesting. The point you say about like the input versus the output, because that was really one of the major criticisms labeled at the whole competency-based framework. It was like to find what went, yeah, went into a job, but then we never really related it to the, to the output. And there was an awful lot of work. I mean, I love competencies.

[00:22:11] You know, they just gave a performance language that you were able to like have with, you know, with an employee that, you know, when you start to talk about non-technical skills, it gave you like layers of competency that you could talk to someone. It was really great for coaching people, but it took a lot of work to like evolve those competency frameworks. Yes. I forgot what the question was you put to me. How do you, once you've got the skills, let's say you've got your skills defined or a framework around that.

[00:22:38] How do you intersect that with your roles or your job architecture? Yeah. Well, I think, I mean, I, I don't think most people, I don't think many people are moving away from the job, the concept of a job. I mean, when I started to think about what is skills going to be based on pay? I, you and I talked about this probably when you joined. Four years ago. Five years ago. Yeah. It's like, I was like the whole concept of a job is going to go away, Sarah. And you kind of looked at me skeptically.

[00:23:06] And I was like, everyone will be just like a package of skills and we will price that person because they're a package of skills. Well, that hasn't happened. And that's a lot harder to kind of notionally move away from the concept of a job. So I think what we have to achieve now, as you say, rapidly is how, you know, most job descriptions do infer skills. But it's, it's like taking a job that you have. I mean, you can use your co-work tool to do this. Like ask it to take all your job descriptions and pull out the skills that are in that.

[00:23:36] And then we produce back to you the job description, but with the skills called out. And then you can start to map it to some of these other taxonomies. So I think it's just, you know, getting really starting to think about like that difference between what you're asking someone to do in a job and then the skills that you require for them to do that. And then comes, you know, an asset that you can use during the hiring phase. It makes it a lot easier, you know, during the interview process and in terms of identifying candidates and potentially assessing candidates.

[00:24:05] And then as you go through the whole process of bringing someone into the organization and then assessing their performance, again, it comes back to giving you that language that you could talk about in terms of developing employees with skills. And then the last step is to figure out what that does to your compensation strategy and what that does to pay for specific roles. Some, many skills will be baked into the job that you're using as a benchmark.

[00:24:31] And then some of those skills won't be and they'll be worth extra money either as you're hiring or to incentivize your current workforce to learn those skills. So that seems like the linchpin step after you've got the skills and you've got your jobs cataloged in them and annotated with those skills is figure out what that does for comp. Yeah. So we have been doing work on looking at skills based pay here and you've been looking at it for some time now, probably.

[00:25:02] So before we talk about like kind of what that's going to look like for pay scale. This is we didn't probably talk about this, so I'm putting you on the spot. Like you and I've had some interesting conversations around some of the things that we've observed about how skills shows up and roles and roles as they stack in terms of a career framework. I think one of the interesting observations you made is as you get more senior.

[00:25:27] The skills stop kind of being called out and they again tend to be inferred in the overall responsibilities of that person. Do you want to talk about that a bit? Sure. Yeah, that's something we're seeing in the data is that as you're hiring, let's say you're hiring entry level software engineers, data scientists, factory workers, those job descriptions and the requirements for those roles,

[00:25:52] even beyond the job description, really do kind of amount to a bag of skills of like, can you use this tool? Can you make this widget? Can you operate this part of the line? And as you, at least in the white collar side, as you go more senior and look at someone like Ruth, for instance, it's hard to sum up the impact that Ruth makes on our company as a collection of the skills that she has.

[00:26:19] And you could even imagine another person in Ruth's job or in her role that brings different skills to the table and also has a big impact. So it doesn't, it gets a little murkier at those more senior levels. You know, what makes a good people manager? How do we describe that? How do we quantify it? What makes a great leader or a great visionary?

[00:26:41] It's tough to do that in a couple of words and have those couple of words be something that's consistent across the organization or across different organizations. So really challenging there. The other thing that's kind of interesting that I think people don't necessarily realize when they start thinking about skills is that even if you think of your employees or a person as a bag of skills,

[00:27:07] the impact that learning a new skill has on the ability of that person to drive impact for your organization isn't, it doesn't stack linearly. So the example I use with my team who are all data nerds is once you know one of the SQL, which is a querying language for databases, and there's like six different types of SQL, but they're all pretty similar. Once you know one of those, I'll hire you.

[00:27:32] I don't care if you know the one that we're using because you can learn that one really quickly as long as you've got that framework in place. And so if you first learn MySQL and then we hire you and you learn PostgreSQL and SnowflakeQL and OracleQL, then great. That's great. But the impact of learning that third database querying language isn't as strong as the impact of learning that first one.

[00:27:57] And so it does make the math more fun and more interesting to think about how, which, what is the combination of skills look like and what was the value of certain combinations of skills. And it goes the other way too. You can imagine two skills that on their own are worth something, but when you put them together are worth a lot more because you can, you can really drive bigger impact there. So the math is really fascinating too.

[00:28:25] So it's not a case of like, I, you know, skill one gets me this amount, skill two gets me this amount. It really like varies. It varies by context a lot. And so let's talk a little bit about, you know, what we're going to be doing here at PayScale. We obviously, as we said, we've been thinking about this problem for quite a while and we believe it's our role to bring the solutions to market that help solve these challenging problems for you all. So Sarah, do you want to give us a kind of. Yeah.

[00:28:56] I'm really excited that, that we've. Through the transfer, the technology. Transformations that we've seen over the last couple of years, we can get at this problem in a new and really powerful way. So right now we're building a skill differential engine that will help organizations figure out how much to pay for a skill or a combination of skills in a specific role. And we're, we're expecting to launch this in the summer and I'm excited about it.

[00:29:24] We're getting really good feedback so far from the comp experts that we've been talking to and sharing the numbers with. Like, oh yeah, that, that makes sense. That job. So the breakthroughs have been meaningful in terms of what we can do in this area. And I'm excited to build tools that will help people in their day jobs do this better. And it's part of our kind of broader like mission, which is to deliver compensation intelligence to the right people at the right time.

[00:29:52] So, you know, historically, yep, we've been great at giving you the price for the job. But now as we think about delivering intelligence to you, we're thinking more about like what are the other insights that we can do that will give you a more informed decision when you're trying to make a decision about what price to apply for a job. So, yeah, we give you the market price. We're also going to be layering in demand data.

[00:30:16] So, you know, you'll be able to understand how many people are posting job adverts for that role, how in demand is that. So you'll be able to see that at the time that you're pricing jobs. And then the skill, skill information is the same. You'll be able to layer that on, bring that into your pricing experience and see what impact that has on, on, you know, where you want to end up in a decision about pay.

[00:30:38] And so it's kind of building out, I like, I call it the 360 view, like when you're making a pay decision, you're no longer making it kind of like in a narrow binary kind of view. We're getting to give you a much more broader view of the different types of information to make the right pay decision. It's an exciting future, right? The reward unicorn. I think we're getting closer and closer. We're getting closer. We're getting closer. Yeah. Great. Well, thank you very much for joining me today, Sarah.

[00:31:05] We're excited to see how this evolves and this ongoing topic that's been one of our favourite ones, as I said earlier, to debate. Thank you all for listening in. If you have any experience around evolving skills-based pay, let us know. We'd love to hear from you. And if you'd like to find out more about like what we're thinking about skills or connect with Sarah or I, we're up for the debate. We're up for the discussion. We would love to learn more as well. As always, we'd love to hear from you.

[00:31:35] So if there's anything that you would like to get in touch with, you can email us at coffee at payscale.com. That is coffee at payscale.com. Or, of course, you can join the conversation on LinkedIn and follow Sarah and I or reach out to Sarah and I on LinkedIn. I know we're connected to many of you who do listen in to the podcast. But a great discussion today, Sarah. Thank you so much for joining me. And we will see you all again very soon. Thanks. Always a pleasure.