Measuring skills in the age of agile work 

Insight into the workforce’s capability and capacity helps leaders pinpoint productivity challenges and efficiently flow talent with the right skills (or critical-to-develop skills) to the right opportunities. Yet less than half (47%) of employees today say their managers understand their current skills.^[1] Further, 48% of companies don’t consider skills insights for career pathing and talent mobility.^[2]

This skills “black hole” exists, in part, because of differing opinions regarding the approaches to skills measurement. On one side, HR tech leaders are bullish on the power of AI (and specifically skills inferences) to deliver skills at scale. On the other, assessment professionals, who have long relied on certified objective tests to measure an individual’s skills, raise concerns about the appropriateness of using AI-based skills inferences to inform pay decisions, succession moves or special projects assignment — because insights are often inferred from past experiences, prior jobs, or other data sources and may not be verified. These professionals question the advisability of going all-in and making talent decisions based on such data alone.

Although both arguments have merit, this healthy debate regarding the best approach to building skills insights has contributed to a lack of consistency in skills measurement, adding to the confusion surrounding how to combine different types of talent insights to improve decision-making.

Today, only 9% of businesses are taking advantage of tech to infer skills.^[3] And most anecdotally consider their approaches to understanding skills to be below par. This makes it tough to get a clear picture of skills across employee populations and job types — and often even tougher to compare data between internal and external talent pools. Agility suffers too, with only 27% of executives strongly agreeing that their workforce models are agile enough to pivot talent from one area of their business to another.^[4] The challenge becomes even more pressing with the growing need for solid data and trusted insights as businesses start to work in more agile ways that require building skills in the flow of work. Consider, for example, how AI drives the need for quicker upskilling due to the decreasing shelf life of technical skills. Or the fact that employees are seeking greater internal career mobility in pursuit of progression.

As it stands, most businesses measure skills based on current job (71%) or manager identification (68%), and almost half (49%) use employee self-reporting.^[5] Levelling up to Workforce 2.0, where human capability is unlocked through greater human–machine teaming, will depend on deploying the most appropriate skills measurement methods at the most appropriate time. While accepting that there is no silver bullet for improved agility or productivity, we can make better people and business decisions when we have the right insights available at the right time. 

As part of a wider talent insights strategy, any method of assessing skills allows for a greater understanding of employees’ capabilities and potential as well as their preferences, motivations and career aspirations. The data gathered can inform talent moves, experiences and learning needed to help unlock greater agility — not just today but tomorrow too. Data-backed assessment allows us to pinpoint areas for work redesign where insufficient skills exist. It can also inform strategic workforce planning by helping leaders identify potential bottlenecks and areas where the current strategy won’t deliver the necessary skills for projected business demands.

Making better talent decisions calls for a multifaceted and flexible skills measurement strategy to guide and govern skills data across the enterprise — with clear guidelines for how these data are used and deployed:

Although each method of collecting skills data has its respective merits, methods can also be combined to paint a more complete skills picture that strengthens talent decision-making and can better inform business strategy.

An organization’s skills measurement strategy and the intended outcome will determine when and how each method is used. Based on our experience with implementing skills-powered organizations, we’ve provided an overview below of how to make the best use of each approach:

Inferred skills can be identified by AI-based software that “tags” skills to an individual by scraping and interpreting existing data. Often, there are two layers to this method:

Initial scraped data from:

• Job history
• Resume/CV or professional job profile (e.g., LinkedIn)
• Personal profiles    

And enhanced by:

• Static data captured in human capital management software or other systems
• Dynamic HR data from learning, performance and recognition programs
• Unstructured data from business and collaboration software, including customer relationship management platforms, project software and emails/chat  

Some leading models are also attempting to infer the proficiency level associated with a skill, but current capabilities are still in their infancy. Instead, most employers use the rated skills method (described in #2 below) to indicate proficiency level.

In this method, the employee and/or their manager (and possibly their peers) rate the employee’s skill proficiency. This rating is typically based on which skills are critical for the current or future role. Rated skills data provide greater rigor than inferred skills in understanding an employee’s skill set and current proficiency levels compared to job expectations. Some employers combine proficiency scores inferred by AI with the self- and/or manager-rated skill proficiencies to improve “directional accuracy” (rather than expecting exact alignment).

Validated skills help organizations understand employees’ skill potential and performance through rigorously developed, validated measures. Validated skills potential measures are typically used when assessing behavioral rather than technical skills. They provide an understanding of an individual’s preferences and behavioral tendencies as well as how willingly and quickly an individual learns. 

Validated skills demonstration measures provide the opportunity for individuals to actively demonstrate a set of skills in a simulated or the actual work environment, such as a workplace exam (virtual or in person). Typically, technical skills tend to be measured using this method.  

In both assessed and demonstrated skills methods, proficiency is determined by benchmarking the individual’s results against the broader market (either at an industry or job family level on a regional or global scale). 

Effectively measuring skills is a key piece of the talent insights puzzle. With a strategic plan in place, you can enjoy a bird’s-eye view of workforce capability while working with individuals to craft personalized learning and career pathways based on business-critical skills. This also establishes a firm foundation for additional AI to aid personalization, nudging and coaching alongside greater clarity to support the validity and ethics of your decisions. Executing a clear skills measurement strategy enables better matching of work to individuals’ skills and motivations, can fuel agile flow-to-work models and will improve the accuracy of all talent decisions, including strategic workforce planning. 

Attracting and retaining top talent in an agile macro environment is crucial to success, and a considered skills measurement strategy is essential for staying ahead of the game. These are critical competencies for HR and leaders to master. After all, it’s your people that give your business a competitive advantage.

[1] Mercer’s 2024 Global Talent Trends Study.

[2] Data from Mercer’s Talent Management 2024 Survey.

[3] Data from Mercer’s 2024 Skills Snapshot Survey.

[4] Mercer’s 2024 Global Talent Trends Study.

[5] Data from Mercer’s 2024 Skills Snapshot Survey.