According to a report from Randstad, while employers are rapidly expanding their demand for 'AI Agent' skills—with job vacancies featuring these requirements rising by 1,587% in 2025—one in five workers still believe that AI will have no impact on their daily tasks. The implicit theory is that the gap is technical; people just need to learn the new technology to recover performance.

There is a different number to set next to that one.

BetterUp Labs found that workforce performance has declined 4–6% since 2019 across all industries, including those with high AI adoption. (Perry, 2025) Despite adding more tools in the last decade, productivity has fallen. (Unlocking the AI-Productivity Paradox in HR: Qualitative Insights Across Organizational Levels, 2025) If reskilling alone were the key solution, we would expect to see better performance outcomes. (Hasan et al., 2024) However, Deloitte points out a paradox: even as workers often feel unprepared, most organizations focus solely on making their employees fit, rather than rethinking how learning actually happens. (Schwartz et al., 2019) Deloitte highlights that more effective strategies may involve integrating learning into the daily flow of work and life. (Creating Value and Impact through the Alternative Workforce: Learning and Development, n.d.) However, data shows that the stated problem is not the root cause. (Eshete et al., 2025) According to Perceptyx, while concerns about artificial intelligence are present, many employees are more affected by feelings of disengagement, isolation, and uncertainty about the impact of their work. The 68% figure reflects that these issues are real for a significant portion of the workforce. (Most Employees Still Unprepared for AI at Work, Perceptyx Research Finds, 2025) It's just describing the surface, not the substrate.

Organizations trying to boost performance with AI repeat old moves: more training, more dashboards. But essentials like clarity and connection are mislabeled as 'soft variables' someone else will handle. For example, instead of simply rolling out new reporting tools, a team lead might start each week by meeting with their team to walk through how each project directly supports a specific client or function. Or, rather than relying only on digital updates, a leader can set up brief one-on-one check-ins to ensure each person understands who will be directly affected by their work and how it contributes to the bigger picture. These small, concrete actions address the variables that actually drive engagement and performance.

They are not soft variables. They are the variables.

Research shows performance improves more when people know why their work matters. Adam Grant’s call center study proved that a brief chat with a scholarship recipient before fundraising doubled results. No training changed—just a named, human recipient.

A separate line of research — Oswald, Proto, and Sgroi's piece-rate experiments at Warwick — found that induced positive mood produces a 10–12% productivity lift on identical tasks. Same person, same task, different condition.

According to Gallup, global employee engagement dropped for the second consecutive year in 2025, highlighting a sharp contrast with recent research gains and suggesting that changing workplace conditions could help reverse the decline in employee performance.

The mattering signal is not motivational speech. It is not a values statement on the wall, a quarterly all-hands, or a Slack message about how everyone's contributions are valued. Those are forms of broadcast, and broadcast doesn't satisfy the variable being measured.

Three features make a mattering signal actually land.

It is specific. Not "your work matters" but "this analysis is going to a particular person who is making a particular decision next Thursday."

It is before the task. Not at the year-end review. Not when the work is done. The signal has to live in the space where the work gets shaped.

It is to someone. The recipient of the work must be a named human, not "the business," "leadership," or "stakeholders." Abstraction is what dilutes the signal.

Most organizations communicate their purpose at scale to broad audiences at suboptimal times. Purpose is emphasized at the organization level but overlooked at the task level. This is not a skills issue; it is a conditions issue, stemming from a lack of vital signals that drive strong performance. (The overlooked essentials of employee well-being, 2018)

Most performance challenges are, at their core, issues of workplace conditions rather than skills, motivation, or technology. Yet interventions typically focus on training, incentives, or new tools—with disappointing results. Address the true conditions for sustainable impact. (Zhenjing et al., 2022)

According to research by Ibrahim Kehinde and Kelly Anderson, the mattering signal is just one of several important factors in the workplace, but it stands out for being freely available, quick to implement, and not dependent on external approval. It simply requires someone to explain, before work begins, why the task is meaningful to a real person. (People Mattering at Work: A Humanistic Management Perspective, 2022)

According to a report from AIProductivity.ai, companies spend 93 percent of their AI budgets on technology and just 7 percent on workforce preparation, underscoring that simple actions, such as delivering clear, specific signals to employees about their value, may offer a cost-effective alternative that executives should consider alongside current AI training investments. (93% of AI Budgets Go to Tech, 7% to Workforce Training - and That's the Real Problem, 2026)

As a leader, implement this immediately: select one upcoming project and, before work begins, individually communicate to the responsible team member exactly who will be impacted and why. According to a Gallup report, employees who use AI regularly often notice improvements in productivity, but broader impacts on how work is done across organizations are still not widely evident. (Kemp, 2026) Before moving forward, consider who benefits from this work, what decisions it influences, and why it is important. If your manager can't answer, that reveals your work's conditions.

For larger teams, this approach can scale beyond one-on-one conversations. Leaders can introduce short team rituals, such as starting project meetings by naming the individuals or groups affected by each deliverable and discussing the specific impact. Managers can be trained to regularly connect each team member's work to its recipient, embedding this practice into onboarding, project kickoffs, and regular check-ins. By making these signals part of the team's shared routines, the sense of mattering can reach everyone—not just those in direct contact with leadership.

The mattering signal is not the whole answer to a 4–6% decline in performance. (Reece et al., 2021, pp. 228-248) But it is a missing condition we are persistently failing to install, and the cost of installing it is approximately zero. To drive sustained improvement, leaders can combine the mattering signal with other essential conditions, such as ensuring clear role expectations and providing regular, constructive feedback. When leaders address these complementary factors alongside mattering, the impact on engagement and performance is even greater.

Improved productivity in the AI era will not result solely from training. Reestablish the foundational conditions that historically drove strong performance—most of which are unrelated to technology. (Eshete et al., 2025)

Reply now with the one condition you are most committed to fixing in 2026. Be specific. I want to hear your priority and your plan. Please share your response by leaving a comment directly on this document or by emailing me at [email protected]. Clear, actionable responses are welcome—let me know your next step.

I read every reply.

— Jennifer