Getting skills right: Why demand and supply must always be on

This article was written by Renée Smith, Senior Managing Director and Global Solutions Lead, Skills at WTW, and Daniel Torrens, Head of Partnerships at TechWolf. It was originally published on wtwco.com and is republished here with permission.

‍

Skills create value when organizations continuously connect talent demand and supply. The strongest performers turn skills data into live intelligence that improves workforce, pay and talent decisions.

Every leadership team is being asked two deceptively simple questions: What skills will we need to ensure future business success? What skills do we already have? There isn’t a confident answer in most organizations.

AI, automation, new work models and persistent talent shortages have made both questions urgent. They’ve also exposed how quickly the usual approaches to skills become outdated. A taxonomy gets built then left to drift. Employee profiles are launched, but managers don’t trust them. Workforce plans describe future roles with no reliable link to the organization’s current capabilities.

The gap is rarely one of ambition. More often, skills have been treated as a discrete project with a beginning and an end — even though the business they support keeps changing. Skills create value only when the demand and supply for them are connected, and when that connection is kept current. Organizations are being called to shift from treating skills as a periodic exercise to a capability that runs continuously and leaders rely on in their decision making.

A proven business case, unevenly executed

WTW’s 2025 Skills Survey: From Concept to Currency, sets the business case: Across 1,238 organizations representing 21.6 million employees, those that use skills effectively across multiple use cases reported materially better outcomes than their peers. These organizations see:

• 2 times higher employee productivity
• 1.5 times stronger retention of key talent
• 1.4 times better attraction and retention
• 1.7 times stronger financial performance

But not many organizations are realizing those gains. Fewer than 30% of participants reported using skills effectively to support even a single primary use case. The reasons tend to be practical rather than strategic:

• A taxonomy is built but not maintained
• The organization focuses on demand without developing a reliable view of supply
• Workforce plans are produced that name future roles, but those roles can’t be connected to current capability
• The reference architecture exists, but it rarely reaches the systems where work and talent decisions are made

Skills only matter when they connect to work

Jobs aren’t disappearing. That’s one point our research has consistently found. Jobs remain the backbone for how work, rewards and talent decisions are made. What is changing is that jobs are increasingly powered and refreshed by skills. A skills-first approach doesn’t require abandoning job architecture; it strengthens architecture with a clearer view of the capabilities the strategy demands.

But, however well designed, a skills list creates little value on its own. Skills become useful when they are connected to the work that needs to be done, the roles through which that work is organized, and the business outcomes leaders are trying to achieve. That connection begins with demand: Translating strategy, operating model shifts, AI adoption and market signals that come together into a forward-looking view of the skills the organization will need.

It’s also important to note that demand on its own isn’t enough. Organizations also need a credible view of supply: The skills employees already have, the skills they’re building and the adjacent capabilities that could support redeployment. Without that visibility, workforce planning remains largely theoretical.

The limits of a point-in-time approach

Many organizations still approach skills as a project: A taxonomy build, a refresh of job profiles, a market scan, a self-reported skills campaign. Each creates a useful foundation — and each begins to age as soon as it’s finished.

A helpful comparison is the difference between a dictionary and a search engine. A taxonomy is the dictionary. It defines the language of skills, creates consistency and gives structure to jobs, careers and learning. That’s all valuable, but it’s static. A dictionary can’t tell you what questions the business is trying to answer: Which skills are rising in demand? Where do capabilities already exist? What action should we take next?

In addition to taxonomy, organizations need something closer to a search engine with a recommendation layer: A live capability that interprets context and connects signals to point leaders, managers and employees toward the most relevant next decision. In practice, that means treating skills intelligence as an operating layer inside workforce planning, talent movement, learning, rewards and work redesign — not a reference document that sits to one side.

At its best, that capability tracks demand by job, work activity and business priority, then builds a sharper view of supply across individuals and the wider workforce. Once in place, leaders can make better decisions about whether to build, buy, borrow, redeploy or automate a given capability, how much to invest in learning, and whether a genuinely scarce skill warrants differentiated pay.

Where continuous intelligence fits

A continuously updated skills intelligence layer addresses this directly. Better data gives leaders a clearer picture of what they already have, where gaps exist and how the work itself is changing. It also gives leaders something concrete to act on when they weigh whether to build, buy, borrow, redeploy or reward a given capability.

This capability matters most for organizations pursuing AI transformation. AI is changing tasks faster than traditional job documentation can capture. Some work will be automated, some augmented and some will depend more heavily on human judgment, creativity and relationships. Continuous intelligence provides the signal organizations need to translate those shifts into workforce planning, role redesign, internal mobility and targeted development.

Technology can surface these signals, but people still must act on them. Organizations make the design choices that matter by identifying:

• Which skills are most important
• How skills connect to jobs and levels
• How proficiency is defined
• How data is governed
• How managers and employees use skills in real decisions

Expertise and continuous intelligence must work together for any of this to move from plan to practice.

Implications for skills-based pay

The demand-and-supply challenge is especially visible in the context of pay. Most organizations aren’t replacing job-based rewards models wholesale. Rather, they’re selectively exploring where skills should inform pay and where doing so creates genuine value.

That selectivity is important. Skills-based pay works best when the skills concerned are scarce, high in value and credibly measurable. Without reliable intelligence on demand and supply, skills-based pay can easily become too broad to be meaningful or too subjective to defend.

Continuous intelligence strengthens the foundation for skill-informed rewards. It shows which skills are genuinely scarce, whether demand for them is rising or cooling, and whether the organization already holds supply it hasn’t spotted. That helps rewards leaders make more disciplined decisions about whether value is best recognized through base pay, premiums, incentives, career progression, development investment or deployment opportunities.

A practical path forward in 6 steps

No organization needs to activate every skills use-case at once. The better starting point is identifying the decisions that skills should improve, then build the minimum foundation needed to support those decisions with confidence.

1. Start with the work. Clarify the strategic priorities, work changes and role shifts that create demand for new or different skills.
2. Strengthen the job and skills foundation. Use job architecture as the organizing backbone, then define skills at the level of granularity you can actually use.
3. Build demand-and-supply visibility. Connect required skills to jobs and work while developing an evidence-based view of individual capability and adjacent skills.
4. Prioritize use cases. Focus first on decisions where better skills intelligence will create measurable value (e.g., workforce planning, internal mobility, upskilling, work and role redesign, selective skills-informed rewards).
5. Choose the right technology and data sources. Determine which intelligence platforms, HR systems, work data, market signals and employee inputs will give you a reliable, ethical and scalable view of demand and supply.
6. Govern and refresh continuously. Establish ownership, validation routines, data ethics, employee transparency and refresh cycles so skills intelligence stays trusted over time.

From skills projects to skills intelligence

The skills agenda has matured. The task now is to make skills visible, trusted and useful enough to improve real business and people decisions. Taxonomy still matters as reference architecture, but it’s only the starting point. Organizations that pull ahead are the ones that turn that architecture into live intelligence that’s kept current as work changes. And this intelligence is what informs the daily decisions leaders face.

Handled this way, skills give an organization a genuine advantage in how it plans and moves its people, rather than remaining an ambition on the side. Leaders spot capability earlier, redeploy talent faster and put their investment where it counts.