Key takeaways:
- Skills-based pay, or differentiating compensation by specific skills, not just role, has resurfaced roughly every generation since Hay Points in the 1940s.
- It keeps stalling for the same reason: inventorying skills and maintaining reliable benchmarks for them is a heavy administrative lift.
- AI is reviving interest again (should an "AI-native" engineer or an AEO-literate marketer be paid differently?), but the underlying cost problem hasn't gone away.
- The more realistic near-term shift: benchmark role descriptions evolve to bake in new skill expectations, rather than comp teams building standalone skills-based pay programs.
- In the meantime, job-postings analysis can give a useful directional read on where skill premiums already exist.
When you’ve been in the compensation field long enough, there are certain topics that resurface every few years. Customized employee rewards is one of them. Skills-based compensation is another that is currently having a moment.
What is skills-based pay?
Skills-based pay is a framework under which employee compensation is differentiated both by role and by the skills employees use to perform in that role. For a software engineer, this could mean differentiating pay by programming language. For a manufacturing employee, this could mean a license or certification to operate specialized equipment.
What is the history of skills-based pay?
In the 1940s, Edward Hay, founder of the Hay Group (now part of Korn Ferry), developed a job evaluation system to look at three key components of each job at an organization:
- Know-how (knowledge, skills, and experience required to perform a job)
- Problem solving (the thinking required by a job)
- Accountability (the job’s impact on end results)
Employers would complete detailed questionnaires providing all of this information, and those responses would be translated into points, which could then be used to develop a market pay rate. That system, officially called the Hay Job Evaluation Framework and more commonly known as Hay Points, was designed to help organizations determine how to compensate employees more objectively.
Why hasn’t skills-based pay caught on?
As you can imagine, the administrative burdens associated with a system like Hay Points or any skills-based compensation system are significant. Organizations need to inventory all skills associated with each role, maintain that inventory in a system of record, and then determine the pay implications for each skill. As part of the annual benchmarking process, the company then needs to review all of the above in addition to the conventional compensation market pricing that most companies complete every year.
In a world where compensation teams are strapped for time and resources, the administrative burdens attached to skills-based pay have never been feasible at scale.
Will AI impact skills-based pay?
With AI transforming the nature of work for many employees, senior managers at companies are once again exploring skills-based compensation. Software engineers are expected to use new coding tools; marketers are expected to think about AEO in addition to SEO, product managers and designers are using AI to ship code. Should pay recognize these new expectations?
Unfortunately, the problems inherent to skills-based compensation programs haven’t gone away. The administrative requirements for tracking and benchmarking these skills still exist, even if what they measure has changed.
What is more likely is that over time, the way we think about benchmark role descriptions will change. If new AI-related skills are truly foundational to a role, then this assumption should be baked into how a data source describes a role, and pay can be set accordingly.
Exploring potential skills-based pay premiums
Despite all of the challenges associated with managing skills-based pay at scale, it can still be useful to analyze how specific skills impact pay from time to time.
A great way to accomplish this is to look at public job postings in locations where the disclosure of pay ranges is required. If published pay ranges are meaningfully higher when specific skills are mentioned in job postings, this can be a signal that a skill carries a premium.
For example, let's look at job postings for warehouse workers with and without forklift skills. To run a quick analysis, we asked Pave Agent to pull 273 live job postings with disclosed pay ranges for warehouse workers in the U.S. Among the 39 postings that specifically mentioned forklift driving as a required skill, average hourly wages were 17% higher, suggesting a material premium for forklift skills.

Is skills-based pay right for your company?
Skills-based compensation keeps resurfacing because the logic is sound: pay should reflect what people actually do, not just their title. But it keeps stalling for the same reason it stalled in the 1940s: inventorying and maintaining skills data at scale is expensive, and reliable benchmarks for specific skills are still hard to come by.
What's different now is the cost of getting a directional read. Comp teams don't need a full skills-inventory program to start—a lightweight, postings-based pulse check can show where the market is already paying a premium, long before a formal program would be justified. That's a much lower bar to clear, and it's worth clearing before deciding whether a bigger investment in skills-based pay makes sense for your org.
Josh Steinfeld is Pave's Director of Data Strategy, bringing over 25 years of experience in the compensation space. Josh started his career as a consultant at Willis Towers Watson, focusing on executive compensation and working directly with Compensation Committees and CEOs on benchmarking, plan design, and corporate governance issues. He then joined the Compensation team at Google, leading compensation for YouTube, Area 120 (Google's corporate incubator), and Google's corporate functions. Prior to joining Pave, Josh spent nearly five years at Carta leading product strategy for Carta Total Compensation.









