Skills Volatility: An Emerging Challenge for Workforce Readiness

In this third edition of the Skills Work! newsletter, I want share that I have envisioned this newsletter consisting of 10-12 "editions" that together form a playbook you can leverage in your skills work -- and please be sure to share your experiences building your own skills strategies, too -- we can make this a collaboration.

In this edition, I introduce a dynamic we need to pay attention to as we implement skills strategies. I refer to it as skills volatility -- a period where the rapid reshaping of the skills required to perform specific job tasks change faster than companies and workers can update workflows, training, and performance measures. Skills volatility has two critical components:

• Skills Flux: When legacy and emerging skills must coexist in a single role impacting tasks within the role.
• Skills Instability: When roles rapidly shift or dissolve as new technologies reshape how the work gets done.

We see skills volatility mostly at the task layer and it creates a performance risk until capability, tooling, and incentives are realigned. Skills volatility comes into play primarily as new technology brings new ways of working. L&D historically struggles with being proactive enough to support task-level change, especially with frontline work. Since roles won't change as quickly as tasks will, mismatches between actual and perceived capability will appear. Since employers expect their workforces to face significant skill gaps (According to WEF 2025, employers expect 39% of core skills to change by 2030. They also estimate that 59% will need new training by 2030, but about 1 in 10 won’t receive the training), skills volatility needs to be considered in your skills strategy.

Let's Learn from Our Past

Evidence of performance value from formal training relies on outcomes and we have data from past technology transitions to inform our actions. A clear example is the migration to the cloud in the 2010s. Infrastructure tasks moved away from provisioning and patching to architecture and automation. And in less than a decade, 80% of business data was moved to the cloud. This was the single-largest change to IT departments since their inception. Within that domain, we saw significant skills volatility in cybersecurity engineering, for example, as every new digital surface deployed created heightened security-related tasks. We've felt the pressure of persistent workforce skills gaps in this area. The impact is still with us as companies are experiencing insufficient cloud-related skills gaps that cause roughly 50% of migrations to add 2+ years to completion. The lesson with the cloud migration is that we can't simply “lift-and-shift” training. To close skills volatility, L&D stood up role-based paths, labs, and incident simulations tied to real SLAs. I remember doing this at Yahoo in 2012. It's critical that we recognize that skills volatility peaks where tasks are routine, high-volume, or tightly coupled. Training will most likely fail if it ignores workflow recomposition. The training for the cloud migration didn’t stall because people couldn’t “use the tool”; they often stalled because the work changed and incentives/metrics didn’t.

With AI and automation being rapidly integrated across the enterprise (similar to the cloud migration, but not function-specific), there's a level of uncertainty that introduces questions for L&D to answer:

1. What role tasks are being (or may be) affected by automation and AI?
2. What are the actual vs. perceived capabilities of the workforce at the task level?
3. How confident are you that your skills strategy will lead to actual performance improvement as you work with the business to get to the task-level?

Skills volatility isn’t a story about jobs vanishing overnight; it’s about tasks being re-written faster than companies can update workflows and learning. Past shifts show the same trap: training the tool instead of re-wiring the work. The proof point that your skills strategy will move the needle on performance sits at the task layer -- mapped, measured, and tied to real outcomes. I recommend building a task-mapping strategy for your enterprise critical roles (ECRs):

• Choose 1-2 ECRs and identify 3-4 high-volume tasks for each ECR (at least one newly AI-affected).
• Build a task evidence map for each task that includes: What “good” looks like (definition of done); Signals available (logs, timestamps, quality/defect flags, customer outcomes); and the KPI targets (−10% cycle time, −20% rework, +5 pts quality)

From this, you can create a volatility index which points to where your focus needs to be for training. You'll obviously need to work with the business on this -- which is why a performance architect is a needed role on L&D. It may seem like nuanced work, and you will find ways to speed up the process and even integrate AI into the L&D workflow to assist, but dealing with skills volatility will be the armor in your skills strategy.