A Skills Strategy Is Not a Technology Project. It Is an Organizational Shift

Cracking the Skills Code, Part 1

Skills strategy has been discussed for years, but most organizations that took explicit strategic steps were early adopters and pacesetters. They were the ones experimenting with business process, understanding market realities, and envisioning a new way of working with talent.

That is now changing.

We are clearly entering a broader phase of adoption as companies move from skills conversations and experimentation to more practical execution. In our client work, that shift often shows up as organizations moving from foundational stages of awareness and intrigue toward more integrated, skills-driven models of talent strategy and technology enablement. SAP SuccessFactors’ own messaging reflects this momentum, with product strategy and go-to-market efforts increasingly aligned to customer demand for a skills-based workforce.

What has changed is not just interest. It is capability.

For years, many of the strongest skills-first strategies were enabled by best-in-breed HR tech vendors. They were often the first to introduce AI into these domains in practical ways: large-scale skill ontologies, skill adjacencies, inferred capability, and matching across use cases like candidate-to-job, employee-to-role, employee-to-mentor, and career recommendations. Those innovations helped prove that skills-first strategy could work beyond theory.

That mattered because most organizations could not scale skills strategies manually or with the limited functionality of legacy HCM applications.

Static taxonomies, spreadsheets, inconsistent job titles, fragmented systems, and self-declared profiles were never going to be enough. AI became critical because it made it possible to normalize skills, infer capabilities from multiple signals, recognize related skills, and power better matching at scale.

But here is the hard truth:

Most skills strategies do not stall because of the technology. They stall because organizations treat skills only as a technology project when they must also involve an organizational shift. Emerging AI applications are necessary, but not sufficient.

That distinction matters.

When skills are framed only as a platform conversation, the focus quickly narrows to vendors, integrations, libraries, and architecture. Those things matter, but they are not the strategy.

The real work is organizational:

• defining what skills mean in your company

• deciding how they connect to hiring, development, mobility, and planning

• determining who governs them

• creating an experience that employees, managers, and recruiters actually trust and use - see the recent post from @Asaf Jackoby for a wholistic model on this point.

That is why a skills strategy succeeds or fails less on software and more on the coherence of the operating model around it.

Why this moment is different

The skills-first conversation is no longer being driven only by niche innovation.

Best-in-breed vendors still have an important role to play, and many will continue adding significant value to client journeys. But large HCM vendors, including SAP SuccessFactors, have also made meaningful progress over the years in supporting the skills-first strategies of their client base.

With the improvements over the last two years to Talent Intelligence Hub as a governance and standardization layer, Growth Portfolio as the employee skills experience layer, and downstream talent, learning, and recruiting modules increasingly consuming and returning that skills data, SAP now offers a much stronger foundation for customers who want to pursue this journey inside the core platform.

That creates a much more compelling opportunity than many customers had in the past.

What AI makes possible (and is still working on)

AI did not invent skills-first strategy.

But it has been essential in making it operational at scale.

The capabilities that matter most are now clear:

• large-scale skill ontologies to structure skills consistently

• skill adjacencies and equivalencies to recognize related capability and improve expectations of time to productivity

• inferred capability to move beyond self-declaration and manager validation guesswork

• dynamic profiles with insight derived not just from legacy HR data, but from signals created in the employee’s flow of work

• matching algorithms to activate skills across hiring, mentoring, mobility, learning, and career growth

• workflows that continue to enable the human in the loop to validate inference before it is final, which remains essential until confidence in these models matures

These are not just technical enhancements. They are what allow a skills strategy to function in real organizational conditions, where data is messy, language is inconsistent, and talent decisions happen across many systems and processes.

They are also increasingly important because the work itself is changing faster than traditional talent structures were ever designed to keep up with. Skill requirements inside jobs shift continuously as technology, processes, and business models evolve. But the underlying job architecture in many organizations still moves at the cadence of compensation review cycles, formal governance processes, and legacy role structures. Over time, the architecture becomes disconnected from the work.

That disconnect is not just a data problem. It becomes a talent experience problem and one of the recurring AI challenges we still see in client work.

My colleague, @Sally Elstad, has framed this particularly well through the distinction between job architecture and career architecture.

Traditional job architecture was largely built to support compensation, leveling, reporting relationships, and organizational control. Those purposes still matter. But they are not enough to guide modern talent decisions. Career architecture requires a different lens. It is about how people grow, how roles relate to one another, what adjacent moves are realistic, what skills bridge one opportunity to the next, and how individuals can navigate progression, lateral moves, and reinvention inside the enterprise.

That distinction becomes especially important in the age of AI.

If the data used to frame and guide AI matching is drawn primarily from compensation-driven job architecture, the recommendations often fail to be meaningful for individuals. The system may understand job codes, pay bands, or hierarchical relationships, but that does not necessarily help an employee understand what role they could move into next, what learning would matter, who might be a relevant mentor, or what adjacent capability could unlock a different future.

In other words, traditional job architecture may classify legacy work, but it often does not support navigation through work. In many of our clients, AI recommendations can provide reliable career trajectories in a traditional ladder framework but often come up short when the client expectation is more of a lattice.

Where many skills strategies go wrong

In my experience, most skills efforts do not fail from lack of ambition or good intent. They most often fail from a lack of coherent strategy and an incomplete shift away from legacy ways of working.

A few patterns show up repeatedly.

Starting with the use case and expected results

One of our clients was explicit from the beginning: “We want to improve retention of our silver-medal internal candidates by 40%.” Teams can spend enormous effort building or cleaning a skills list before aligning on what decisions the skills data is supposed to improve.

Confusing more data with more strategy

An organization can infer thousands of skills and still fail to improve hiring, mobility, development, or planning in meaningful ways. Hearts and minds still matter. Compliance is not advocacy, and failure to eliminate legacy rogue systems only delays real value.

Treating governance as the last step in the process

Governance is not the slow part of a skills strategy. It is the part that makes scale possible. Without decisions around structure, validation, standardization, flow, and ownership, the result is duplication, confusion, low trust, and rework.

Separating architecture from experience

The value is not in having a taxonomy or ontology. The value is in creating better talent experiences and better decisions because the core skills architecture exists.

Assuming legacy job structures are enough for modern mobility

Many organizations are trying to power career recommendations, mentoring matches, and internal mobility with architectures that were designed for compensation administration, not career navigation. Those are not the same thing. We still need better solutions from our best and brightest minds in this space.

What leaders should focus on first

Before expanding the technology conversation, leadership teams should align on four questions:

What business outcomes matter most?

Hiring precision? Internal mobility? Workforce planning? Development? All of the above?

What level of coherence is required now?Not every company needs a perfect enterprise-wide model on day one. But every company does need enough consistency to support trust and action.

Who owns what?Role structure, skills standards, validation, and downstream use all need clear ownership. We increasingly see previously separate HR teams coming together around both business process governance and system design decisions. These are interconnected, and leading HR teams are finding ways to break down internal walls that may have existed in the past.

What experience are we trying to create?What should be better for employees, managers, recruiters, and leaders because this strategy exists?

I would add a fifth question that is my current leading interest for the workforce of the future:

How will the emerging identification of work and task ontologies, integrated agents, and human work continue to shape our expectations for skills and the workforce?

More on that in Part 6 of the series.

The bottom line

The most successful skills strategies are not just implementing tools.

They are redefining how the organization sees talent:

• from static descriptions to dynamic capability

• from disconnected systems to coherent design

• from scattered signals to governed intelligence

• from job classification to career navigation

• from HR process to enterprise capability

Technology enables this shift.

AI accelerates it.

Best-in-breed innovation still matters.

And SAP SuccessFactors now offers a far stronger core foundation for clients that want to pursue skills-first strategies through TIH, Growth Portfolio, and downstream talent, learning, and recruiting integration than it has in the past.

The real effort is an organizational shift in strategy, technology, and ways of working.