Post
JA EN

Employee Well-being Can't Be Bought with Programs: The Organizational Levers That Design the Seven Factors

Posted
By
14 min read
Employee Well-being Can't Be Bought with Programs: The Organizational Levers That Design the Seven Factors
  • Intended readers: Engineering managers and tech leads wrestling with retention and performance, HR professionals whose engagement programs aren’t moving the needle, and executives thinking about organization design
  • Prerequisites: None. A rough sense of what a correlation coefficient means will help, but it’s explained as we go.
  • Reading time: about 11 min

Overview

Organizations pour resources into improving employee engagement: richer benefits, recurring surveys, more 1-on-1s. Yet global engagement rates have sat stubbornly around 20% for years—Gallup’s 2024 report put the figure at 23% (based on 2023 data)1—and most of these programs produce only small to moderate gains.2 The reason is straightforward: programs change the periphery of work; they don’t change the substance.

What determines whether work is good is not salary or perks but seven intrinsic qualities of the work itself—autonomy, achievement, focus, clarity, variety, connection, and contribution. (The personal side of this argument is developed in a companion article; this piece focuses on what organizations can do.) These seven factors aren’t just something individuals can craft for themselves—organizations can design them in.

Job design interventions that actually change the substance of work show positive effects in 71% of studied cases.3

This article describes three organizational levers for building the seven factors: (1) designing the work itself (job design), (2) creating the conditions—psychological safety and autonomy support—that let employees reshape their own work (job crafting), and (3) hiring people whose needs match what the organization can genuinely supply (needs-supplies fit). It closes with the AI era question: how to direct the slack that AI creates toward the seven factors rather than letting it dissolve into more tasks.

The conclusion in brief: employee well-being isn’t something you buy with programs. It’s something you build through design.

Part 1: Why Programs Alone Don’t Move the Needle

Engagement initiatives tend to target the work’s periphery—compensation, benefits, social events, surveys. The data consistently show that these levers don’t reach far enough.

Global engagement has hovered in the low twenties for years. Gallup’s 2024 report recorded 23% (2023 survey data).1 A systematic review and meta-analysis of workplace interventions found an overall effect of Hedges g ≈ 0.29—small to moderate—with no significant difference between intervention types (job design, training, health promotion).2 There is no single program that changes everything.

Engagement surveys are particularly misunderstood. A survey is a diagnostic instrument, not a treatment. Research on post-survey follow-up shows inconsistent results.4 Worse, if employees are asked for input and nothing about their work actually changes, the message they receive is “speaking up is pointless”—trust erodes and participation rates fall in subsequent rounds.

To be fair about the evidence: programs are not completely useless. Benefits and training can improve short-term satisfaction, and bottom-up redesign efforts have modest positive effects.5 But the consistent finding is this: the factors that most strongly predict engagement are those closest to the substance of work—discretion, feedback, skill utilization.6 No amount of peripheral polish moves the needle if the work itself doesn’t change.

And the person with the most influence over that substance is not HR—it’s the manager. Gallup estimates that roughly 70% of the variance in team engagement is explained by the manager.1 The real battleground is how frontline managers design work, not which programs the organization deploys.

Part 2: Building the Seven Factors Through Job Design

flowchart TB
    L["The 3 organizational levers<br>① Design the work<br>② Enable crafting<br>③ Hire for fit"] --> G["Employee well-being<br>(the seven factors)"]
    X["Programs / surveys alone"] -. "limited effect" .-> G

Since Hackman and Oldham, the research consensus has been that work is not something handed to employees—it is something that can be designed. Job characteristics explain 34% of the variance in job satisfaction, a figure from a meta-analysis covering hundreds of thousands of workers.7 A systematic review of interventions that actually changed the content of work found that 71% of 55 studied cases produced positive outcomes.3 Job design works.

Translating the seven factors into organizational actions:

  • Design for contribution: Make beneficiaries visible. In Grant’s field experiment, fundraisers who spent just five minutes meeting a scholarship recipient they had supported raised roughly 2.7× more the following month (+171%).8 For engineering teams, this means routing user feedback directly to developers, making it visible whose feature solved which problem.
  • Design for clarity: Clear, challenging goals outperform vague ones by an effect size of d ≈ .5–.8.9 OKRs and sprint goals are implementations of this principle. Equally important: don’t let role ambiguity fester—confusion about who owns what is a direct drag on both clarity and achievement.
  • Design for achievement: Make small progress visible. In a diary study, 76% of participants’ best-mood days included a “small win.”10 Small, mergeable units of work, visible progress boards, and frequent deployments are structural achievement-delivery mechanisms.
  • Design for autonomy: Delegate decisions about how and what to the team wherever possible. Among the seven factors, autonomy has an outsized effect.

One important caveat: whether a job redesign effort succeeds depends less on what is changed than on how it is implemented.3 Changing the org chart without changing how work actually flows produces no benefit. That is precisely why the second lever matters.

Part 3: The Organization That Enables Job Crafting

The seven factors can also be built by employees themselves. Actively reshaping the tasks, relationships, and meaning of one’s own role is called job crafting.1112 The companion article argues that individuals should do this for themselves. But whether crafting actually happens depends on whether the organization allows it.

The foundational condition is psychological safety. Edmondson’s classic study found that psychological safety correlated strongly with team learning behavior (reported r ≈ .80), and that learning behavior in turn drove team performance.13 A meta-analysis integrating 117 studies confirmed that psychological safety is consistently and positively associated with learning, voice, and job satisfaction.14 Google’s Project Aristotle, which analyzed 180 teams, identified psychological safety as the factor most predictive of team effectiveness.15

Psychological safety is, in effect, the ignition condition for crafting. In cultures where failure is punished, no amount of exhortation to “take on challenges” will cause people to pursue difficult work voluntarily. One study found that psychological safety predicted challenge-oriented crafting (β ≈ .16) among permanent employees, but the effect disappeared for workers in more precarious positions.16 Without safety, challenge crafting doesn’t happen.

The second lever is leader autonomy support. When managers support rather than micromanage subordinate autonomy, job satisfaction correlates at ρ ≈ .56.17 Micromanagement cancels out the autonomy that job design intended to grant.

Two practical design notes. First, job crafting interventions have a small average effect, but that average hides an important pattern: the gains are concentrated among employees who were crafting the least to begin with.18 Designing these programs for the more passive members of the team, rather than the already-engaged, is the rational allocation. Second, transparency is non-negotiable. Crafting has a shadow side—when people quietly reduce demands without coordinating, the burden often shifts to colleagues, breeding conflict and lowering team output.19 If crafting is encouraged, make it visible: who is taking what on, and who is putting what down.

Part 4: Hiring for Needs-Supplies Fit, Not Personality Type

Even with well-designed work and a crafting-friendly culture, if what the organization can offer and what a candidate needs are fundamentally misaligned, the mismatch won’t close after hiring. Recruitment is therefore also a design variable.

First, what to stop doing: don’t screen on personality type instruments like the MBTI. The MBTI produces a different type classification for 39–76% of respondents on retest20 and has no demonstrated predictive validity for job performance.21 Hiring on the basis of “this type is a good fit” is closer to astrology than assessment.

What to look at instead is needs-supplies fit—the match between what a candidate wants from work and what the organization can genuinely deliver. This fit correlates with job satisfaction at ρ ≈ .61, stronger than any single job characteristic.22 In practice, this means honestly auditing the seven factors your organization actually provides: How much autonomy can you really grant? What does growth look like here? Is the meaning of the work visible? Disclose this honestly, and use it as a matching surface during hiring. Overselling leads to misfit attrition—in a needs-supplies framework, “this isn’t what I was told” is the worst possible post-hire discovery.

Part 5: Organization Design in the AI Era

AI has added one new variable to the design problem.

AI absorbs the routine implementation layer of engineering work, freeing up capacity. But if organizations don’t design for that capacity, it defaults to “more tasks.” Telemetry analysis by Faros AI found that even when individual productivity rose with AI-assisted development, organization-level delivery metrics did not improve, and bugs per developer actually increased.23 What the freed time gets pointed at is an organizational design decision. Without a framework that channels the slack toward higher-order challenges—verification, architecture, system-level oversight—the slack doesn’t convert into the seven factors.

Here again, organizational support is the dividing line. AI adoption can push people toward either approach-oriented crafting (seeking out new challenges) or avoidance-oriented crafting (retreating from difficulty). What separates the two is not individual personality—it is organizational AI knowledge sharing.2425 Study groups, internal documentation, paired practice: an environment where people can learn to work with AI turns anxiety into agency.

Organizations should also design against skill hollowing. Developers who relied heavily on AI coding tools showed measurable skill degradation within three months, and employment in high-AI-exposure roles for early-career engineers dropped by roughly 13% in some reports.26 Drawing explicit lines around what is delegated to AI, sustaining code review culture, and maintaining mechanisms that keep foundational skills alive are protective design choices that organizations, not individuals, need to make.

Finally, the most commonly overlooked point: the transition that AI is forcing—from implementation to orchestration—feels threatening when it is imposed from above, and feels meaningful when the person chooses it themselves. The same change in role can be experienced as loss or growth depending on who holds the steering wheel. The best design an organization can offer is to hand that wheel to the people doing the work.

Summary

Programs change the periphery of work. Design changes its substance. When engagement doesn’t move, the usual culprit is an organization that keeps adjusting the periphery while leaving the substance untouched.

The three organizational levers for building the seven factors of well-being are: design the work itself (job characteristics explain 34% of satisfaction variance; 71% of redesign interventions show positive effects), enable employees to reshape their own work (psychological safety and autonomy support are prerequisites), and hire for match rather than type (needs-supplies fit, not personality instruments). In the AI era, add three more moves: direct freed capacity toward challenge rather than volume, build environments where AI knowledge can spread and reduce anxiety, and return the steering wheel for role transitions to the people making them.

Individuals craft their work. Organizations design the conditions that make crafting possible. Neither alone is sufficient. Employee well-being is not bought—it is designed.

References

Footnote numbers correspond to in-text citations. Correlation coefficients (ρ, r) and effect sizes (d, g) match the values used in the article body.

  1. State of the Global Workplace 2024 (Press Release) — Gallup (2024). 2023 survey data; global engagement rate 23%. [Reliability: Medium-High (proprietary research, large sample)] ↩︎ ↩︎2 ↩︎3

  2. Building work engagement: A systematic review and meta-analysis investigating the effectiveness of work engagement interventions — Knight, C., Patterson, M., & Dawson, J., Journal of Organizational Behavior (2017). [Reliability: High] ↩︎ ↩︎2

  3. How work redesign interventions affect performance: An evidence-based model — Knight, C., & Parker, S. K., Human Relations (2021). [Reliability: Medium-High (systematic review; no pooled effect size reported)] ↩︎ ↩︎2 ↩︎3

  4. Following up on employee surveys: A conceptual framework and systematic review — Huebner, L., & Zacher, H., Frontiers in Psychology (2021). [Reliability: Medium-High] ↩︎

  5. A meta-analysis of bottom-up work redesign / job crafting interventions — Björk, J. M., et al., Frontiers in Psychology (2021). [Reliability: Medium-High] ↩︎

  6. The relative importance of various job resources for work engagement: A concurrent and follow-up dominance analysis — Hakanen, J. J., Bakker, A. B., & Turunen, J., BRQ Business Research Quarterly (2021). [Reliability: Medium-High] ↩︎

  7. Integrating motivational, social, and contextual work design features: A meta-analytic summary — Humphrey, S. E., Nahrgang, J. D., & Morgeson, F. P., Journal of Applied Psychology (2007). [Reliability: High] ↩︎

  8. Impact and the art of motivation maintenance: The effects of contact with beneficiaries on persistence behavior — Grant, A. M., et al., Organizational Behavior and Human Decision Processes (2007). [Reliability: High (randomized field experiment)] ↩︎

  9. Building a practically useful theory of goal setting and task motivation: A 35-year odyssey — Locke, E. A., & Latham, G. P., American Psychologist (2002). [Reliability: High] ↩︎

  10. The power of small wins — Amabile, T. M., & Kramer, S. J., Harvard Business Review (2011). Diary study underlying the book The Progress Principle. [Reliability: Medium-High (observational study without control group)] ↩︎

  11. Crafting a job: Revisioning employees as active crafters of their work — Wrzesniewski, A., & Dutton, J. E., Academy of Management Review (2001). [Reliability: High] ↩︎

  12. Development and validation of the job crafting scale — Tims, M., Bakker, A. B., & Derks, D., Journal of Vocational Behavior (2012). [Reliability: High] ↩︎

  13. Psychological safety and learning behavior in work teams — Edmondson, A. C., Administrative Science Quarterly (1999). [Reliability: High] ↩︎

  14. Psychological safety: A meta-analytic review and extension — Frazier, M. L., et al., Personnel Psychology (2017). [Reliability: High] ↩︎

  15. Understand team effectiveness (Project Aristotle) — Google re:Work. [Reliability: Medium (non-peer-reviewed internal research; treated as near-primary source)] ↩︎

  16. Psychological safety, job crafting, and employability: A comparison of permanent and temporary workers — Plomp, J., et al., Frontiers in Psychology (2019). [Reliability: Medium-High] ↩︎

  17. Leader autonomy support in the workplace: A meta-analytic review — Slemp, G. R., Kern, M. L., Patrick, K. J., & Ryan, R. M., Motivation and Emotion (2018). [Reliability: High] ↩︎

  18. Effects of a job crafting intervention program on work engagement among Japanese employees: A randomized controlled trial — Sakuraya, A., et al., Frontiers in Psychology (2020). [Reliability: High] ↩︎

  19. The dark side of job crafting (Academy of Management Proceedings) — Abukhait, R., et al. (2019). [Reliability: Medium] ↩︎

  20. Cautionary comments regarding the Myers-Briggs Type Indicator — Pittenger, D. J., Consulting Psychology Journal: Practice and Research (2005). [Reliability: High] ↩︎

  21. A 25-year review and psychometric synthesis of the Myers–Briggs Type Indicator (MBTI) – Form M — Erford, B. T., et al., Journal of Counseling & Development (2025). [Reliability: High] ↩︎

  22. Consequences of individuals’ fit at work: A meta-analysis of person-job, person-organization, person-group, and person-supervisor fit — Kristof-Brown, A. L., Zimmerman, R. D., & Johnson, E. C., Personnel Psychology (2005). [Reliability: High] ↩︎

  23. Key takeaways from the DORA 2025 report — Faros AI (2025). Telemetry analysis of productivity and quality metrics in relation to the DORA 2025 report. [Reliability: Medium] ↩︎

  24. Organizational AI adoption and approach vs. avoidance crafting — Liu, Tian, Li, & Tan, Frontiers in Psychology (2025; published online January 2026). [Reliability: Medium-High (Chinese sample, 3-wave longitudinal)] ↩︎

  25. Digital-AI transformation and job crafting — Sha, & Chai, Frontiers in Psychology (2025). [Reliability: Medium-High (Chinese sample, longitudinal)] ↩︎

  26. New Future of Work Report 2025 — Microsoft Research (2025). [Reliability: Medium-High] ↩︎

Career & Learning
Organization Design Job Design Job Crafting Psychological Safety Engagement Management Software Engineers AI Era Careers
This post is licensed under CC BY 4.0 by the author.