Why some workforce skills have not bounced back since 2019

Despite unprecedented access to modern technology and productivity tools, we see signs of strain across organizations: individual contributors seem disengaged, managers are overwhelmed with decisions, and senior leaders struggle with direct reports who have agency, yet hesitate to make bold choices. People grumble about these human-skill weaknesses, but lack the data to support their intuition and make a case for change.

Between 2019 and 2025, BetterUp tracked the trends of human-centric skills and performance of over 351,000 individuals and discovered something striking: without exception, human-centric skills sharply dropped between 2019 and 2021 and have not recovered to pre-pandemic levels nearly five years later.

Human-centric skill erosion is not only a concern for future workforce readiness; it is already extracting a measurable performance toll. We observe drops across every dimension of work effectiveness as high as 6%. Based on these declines, if employees today were evaluated by the performance standards of 2019, almost 75% of them would be in a lower performance decile than they are today.

Though no single factor accounts for these declines, several converging pressures likely contribute: prolonged post-pandemic stress, economic insecurity, geopolitical uncertainty, hybrid work fragmentation, rising burnout, generational shifts in employee expectations, and education and training institutions neglecting human-centric skills. Concerns are also rising regarding cognitive offloading to AI and technology eroding skills. Just like all other economic measurements, human-centric skills are susceptible to external shocks and societal changes.

Ironically, as these skills decline, they’re more critical than ever. According to the World Economic Forum, creativity and problem-solving (bundled as a skillset) emerges as the most highly monetarily-valued skill, while curiosity and lifelong learning are key survival qualities in VUCA environments, enabling workers to keep pace with accelerating change.

As AI automates routine cognitive work and reshapes entire industries, these human capabilities – creativity and problem-solving, as well as curiosity and lifelong learning – emerge as organizations’ essential differentiators against competitors. These skills are not only safe from AI displacement, but likely multiply AI’s positive impact in a workforce.

Despite these skills’ importance, organizations underinvest in them. They are among the least frequently recognized skills at work and appear in only 2% of job postings. These capabilities declined by 4% during pandemic years and have yet to bounce back.

Specifically, the skill of curiosity and lifelong learning is rated as having the lowest competency in workforces by employers. Our BetterUp data helps us understand why: curiosity and lifelong learning is the hardest skill to build. It takes eight months of coaching for 50% of under-skilled learners to achieve basic competency, and nearly 24 months for 90% of learners to actually acquire the skill, much less master it. A combination of low visibility and investment, difficult-to-build experiences and external shocks is a recipe for starving the skills that fuel competitive advantage.

Individual contributors (ICs) experienced the steepest drops in human-centric skills, particularly in creativity and problem-solving, curiosity and life-long learning, resilience and agility, and leadership and social influence. This erosion is plausibly driven by prolonged social isolation and leads to several consequences.

Weakened ICs not only underperform, but also deplete their managers. Managers now have to mentor this population more deeply and solve their problems while shouldering increasingly larger, more geographically scattered teams – teams that are being mandated to use AI. We see a symptom of this cycle already in the form of workslop: low effort AI-generated output that lacks the substance to meaningfully advance a given task and shifts the burden onto the receiver. In an AI-augmented workplace, the impact of individual capability can compound exponentially. Every individual counts.

Historically, leadership responses fall short. Off-the-shelf learning courses and workshops may start conversation, but likely won’t build sustained behavioural change. Workers need continued engagement, long-term investment and personalized adaptation to their specific contexts.

Science points to a need for an organization-wide adaptive system for human transformation. In a peer-reviewed study looking at coaching interventions on 391 participants, people who participated showed significant human-skill improvements, and the same interventions across over 90,000 people helped professionals bounce back from a slump during and after the pandemic, even exceeding 2019 performance and skill levels.

Strategies to implement human-skills investment include:

These recommendations should prompt leaders to look more deeply at their organizational systems and answer open research questions – such as how to create the best targeted interventions for disproportionately affected employees, or to what extent various factors drive workplace skill declines. As AI reshapes work, leaders will need solid research to define the evolving relationship between technological adoption and human skills.

Understanding the decline in human-centric skills requires examining the systems, structures and blindspots that allowed this crisis to develop in the first place. Only then can an organization buttress the workforce for performance in the face of future challenges.

Leaders around the world are making billion-dollar decisions about AI adoption, workforce restructuring and talent development, while operating with little understanding of the human skills at the centre of these transformations. Understanding what is eroding human capabilities and performance, and how to reverse it, is how we ensure the next transformation delivers on its investment.