The Architecture of Work Is Collapsing. Most Organizations Are Still Building Ladders.
Listen to a Review of this Article:
Abstract: Traditional organizational hierarchies are undergoing rapid compression driven by technological capability, economic pressure, and structural necessity. Analysis of recent workforce data reveals a fundamental disconnect: while leadership architectures designed for vertical progression remain intact, the actual structure of work has already shifted toward horizontal capability development, AI-augmented supervision, and compressed leadership passages. This article examines the organizational and individual consequences of architectural compression, distinguishes between cost-driven delayering and capability-driven restructuring, and presents evidence-based interventions for redesigning work systems around expanded individual capability rather than hierarchical progression. Drawing on workforce analytics, organizational behavior research, and implementation cases across technology, financial services, and healthcare sectors, the article demonstrates that architectural compression is not a future disruption but a present reality most organizations are managing incorrectly. The strategic imperative is not ladder reform but fundamental redesign of how capability develops, how supervision functions, and how value gets created in AI-augmented work environments.
The metaphor is dead, but the infrastructure remains. For nearly a century, the career ladder has served as both the organizing principle for professional development and the primary mechanism through which organizations allocated opportunity, compensation, and status. The ladder promised a straightforward exchange: demonstrate competence at your current level, wait the appropriate interval, and you would be granted access to the next rung. Performance management systems, succession planning processes, and compensation structures were all built to support this vertical progression model.
The model has collapsed. Not theoretically or gradually, but structurally and suddenly. Between 2013 and 2025, the average span of control for managers increased from 8.2 to 12.1 direct reports—a nearly 50% expansion in team size (Gallup, 2025). Simultaneously, manager engagement has declined more sharply than any other workforce segment, while managers now allocate 46% of their time to individual contributor work rather than leadership activities (Gallup, 2025). These are not symptoms of a system under temporary strain. They are indicators of architectural collapse.
The why-now is straightforward. Three forces have converged with sufficient intensity to break the old structure: artificial intelligence is automating supervision and coordination tasks previously requiring human managers; cost pressure is forcing organizations to eliminate layers faster than they can redesign the work those layers performed; and speed requirements in product development, customer response, and strategic execution have made traditional hierarchical approval chains operationally untenable (Charan, 2025). The result is not a reformed ladder with better rungs. It is a fundamentally different architecture that most organizations are still trying to manage with ladder-era tools.
The practical stakes are immediate. Organizations continuing to promote based on vertical progression criteria while the actual work requires horizontal capability range are creating what can only be described as structural fraud—people given titles that no longer correspond to roles, compensation systems that reward patience rather than contribution, and development pathways that prepare employees for positions that have already been eliminated. The cost of this misalignment extends beyond individual career disappointment. It includes the systematic destruction of organizational capability development, the loss of high performers who correctly perceive the disconnect, and the creation of manager roles that have become functionally impossible to execute well.
This article examines the organizational and individual consequences of architectural compression, presents evidence-based interventions for organizations attempting genuine restructuring rather than cost-driven delayering, and identifies the capability-building strategies required for sustainable operation in compressed structures.
The Organizational Architecture Landscape
Defining Architectural Compression in Contemporary Organizations
Architectural compression refers to the reduction in distinct leadership levels within an organization, but the term requires precision. Two forms of compression are occurring simultaneously, and conflating them produces strategic error.
Cost-driven delayering involves the removal of organizational layers primarily to reduce labor costs. A management tier is identified, positions are eliminated, and the work previously performed by that layer is redistributed to remaining roles above and below the deleted tier. The organizational structure diagram changes—fewer boxes, shorter reporting lines—but the fundamental nature of the work, the approval requirements, the coordination mechanisms, and the decision rights remain largely unchanged. The result is a compressed organization chart operating with uncompressed processes (Charan, 2025).
Capability-driven compression involves the redesign of work systems around expanded technological capability, particularly AI-augmented supervision and coordination. This form of compression asks which elements of management work can be performed more effectively by AI systems, which decisions can be pushed to individual contributors with appropriate tool support, and which coordination mechanisms can be replaced by software platforms. Only after answering these questions does the organization determine how many leadership layers are required. Layer reduction becomes a consequence of redesign rather than the objective driving it (Thompson & Autor, 2024).
The distinction matters because the two approaches produce opposite outcomes in terms of organizational capability, employee engagement, and sustainable cost structure. Research on automation and labor value demonstrates that when AI eliminates administrative scaffolding around expert roles, the remaining workers become more valuable and their compensation increases. When AI eliminates the expert tasks themselves, wages decline (Thompson & Autor, 2024). Most organizations pursuing delayering are removing managers without determining which category of work is actually being automated, creating compression that destroys rather than enhances organizational capability.
Current State of Architectural Practice
The dominant pattern in contemporary organizations is compression driven by cost imperative rather than capability redesign. A 2024 analysis of Fortune 500 restructuring announcements found that 78% explicitly cited cost reduction as the primary objective, while only 31% mentioned process redesign or technological capability as driving factors (Workhuman, 2024). The typical sequence involves targeted headcount reduction in middle management layers, followed by span-of-control expansion for remaining managers, followed by the appearance of execution gaps that prompt either role additions or the implementation of committee structures to fill coordination needs the eliminated layer previously handled.
This pattern explains the seemingly paradoxical workforce data. Manager spans of control have widened dramatically, yet managers report spending nearly half their time on individual contributor work rather than leadership activities. The explanation is not that managers are choosing IC work over leadership work. It is that elimination of adjacent management layers has created gaps in execution capacity that cannot be filled through supervision alone. Managers are performing IC work because the organizational architecture now requires it, while simultaneously attempting to supervise teams 50% larger than they were a decade ago.
The engagement consequences are predictable. Gallup's 2025 workforce data shows manager engagement declining faster than any other employee segment. When roles become structurally impossible to perform well—when the scope exceeds available time, when the work type conflicts with role expectations, when success criteria remain unchanged despite changed conditions—engagement collapse is the rational response.
Across industries, there is significant variation in compression approaches. Technology firms, particularly those with strong AI capability, have moved more aggressively toward authentic capability-driven compression. Financial services firms have pursued cost-driven delayering while maintaining most traditional approval structures. Healthcare organizations have compressed administrative hierarchies while clinical hierarchies remain largely unchanged. Manufacturing has automated supervision in production environments while corporate functions retain traditional structures. These variations suggest that industry characteristics—regulatory requirements, labor market dynamics, capital intensity—shape both the pace and the approach to compression, but the direction of change is consistent across sectors.
Organizational and Individual Consequences of Architectural Misalignment
Organizational Performance Impacts
The performance consequences of cost-driven delayering without corresponding process redesign are both immediate and compounding. When managerial layers are removed without redesigning the work those layers performed, several predictable failures emerge.
First, decision velocity slows rather than accelerates. The original justification for delayering typically includes faster decision-making—fewer layers means shorter approval chains. In practice, removing a layer without redistributing its decision rights creates ambiguity about who now holds approval authority. Research on organizational decision-making demonstrates that authority ambiguity produces longer decision cycles than clear hierarchical structures, even when those structures involve more layers (Bruch & Menges, 2010). Organizations that eliminate a VP layer without explicitly reassigning signing authority for capital expenditures, vendor contracts, or hiring decisions often find that decisions previously made in days now require weeks as they route through unclear approval chains.
Second, coordination costs increase. Middle management layers serve a critical coordination function, translating strategy into execution plans, identifying resource conflicts across teams, and managing dependencies between functional groups. When these layers disappear without replacement coordination mechanisms, the coordination work doesn't disappear—it fragments. Individual contributors spend more time in cross-functional meetings trying to resolve dependencies, senior leaders get pulled into tactical coordination they previously delegated, and projects experience delays from unmanaged dependencies (Workhuman, 2024).
Third, capability development slows. Traditional management hierarchies served as development pathways, with each level providing structured exposure to increasing scope and complexity. When organizations compress from six leadership passages to three or four without redesigning how people gain exposure to complexity, they create a capability gap. A manager who would have spent three years managing managers before becoming a functional leader is now expected to make both transitions simultaneously. Without deliberate skill-building support, many fail. A 2023 study of leadership failures in recently restructured organizations found that 64% of executives who failed in their roles had been promoted through two or more traditional passages simultaneously without corresponding development support (Zenger Folkman, 2023).
The financial implications are substantial. Organizations that have pursued multiple rounds of cost-driven delayering show average productivity declines of 8-12% in the year following restructuring, costs which often exceed the labor savings achieved (Cascio & Boudreau, 2017). This productivity loss stems from increased coordination costs, slower decision cycles, and execution gaps from eliminated roles.
Individual Wellbeing and Career Development Impacts
For individual employees, architectural compression without corresponding mental model shifts creates what can best be described as structural gaslighting. The organization's stated career architecture promises vertical progression through demonstrated performance. The actual opportunity structure requires horizontal capability building and tolerance for ambiguity. Employees who follow the stated pathway find themselves in roles that no longer exist or that require capabilities they were never developed to have.
The psychological consequences are measurable. Research on psychological contract violation—the perceived discrepancy between what an organization promises and what it delivers—shows that career pathway misalignment produces stronger negative effects on engagement and retention than compensation misalignment (Robinson & Morrison, 2000). When employees prepare for vertical progression and encounter horizontal requirements, the resulting contract violation drives both immediate engagement decline and longer-term career disillusionment.
The retention implications are particularly severe for high performers. Employees with strong performance records in traditional hierarchical environments often succeed precisely because they have mastered vertical progression skills: building relationships with immediate supervisors, demonstrating competence within a defined scope, navigating approval processes. When the architecture shifts to value horizontal range and tolerance for ambiguity, these employees face a sudden competence crisis. Their historical success patterns no longer predict future success, and the organization provides neither clear signals about what new patterns matter nor structured support for developing them. High performers become high-risk retention cases not because their capabilities have diminished but because the capabilities the organization now values are orthogonal to the ones these employees developed (Zenger Folkman, 2023).
The career development challenge extends beyond individual disorientation. When organizations compress leadership passages without redesigning development pathways, they create what amounts to a structural experience deficit. Consider an example: In a traditional six-passage architecture, an individual might spend two years as a manager, three years as a manager of managers, and four years as a functional leader before advancing to business unit leader. This nine-year sequence provides structured exposure to progressively complex challenges—managing individual performance, managing through managers, managing across functions. When an organization compresses these passages and promotes someone from manager directly to functional leader, it has not eliminated the complexity those intermediate passages were designed to handle. It has simply made the individual responsible for navigating that complexity without the structured experience the old architecture provided.
The result is predictable. Research on leadership derailment identifies "inability to build and lead teams" and "difficulty managing complexity" as the two most common failure modes for executives promoted too quickly through compressed structures (Corporate Leadership Council, 2003). These are not individual capability failures. They are architectural design failures dressed up as individual performance problems.
Evidence-Based Organizational Responses
Table 1: Case Studies of Organizational Architectural Compression and Restructuring
Organization | Year of Implementation | Compression Strategy Type | Key Interventions | Impact on Layers/Span of Control | Reported Outcomes |
Mastercard | 2023 | Capability-driven compression | AI-augmented supervision platform (automated task assignment, progress tracking, and resource allocation) | Eliminated 2 management layers; average span of control expanded from 9 to 14 direct reports | Maintained development velocity; improved time-to-promotion for high performers by 23% |
Mayo Clinic | 2023 | Capability-driven compression / Distributed Leadership | Domain-based decision authority; rotating leadership assignments for strategic care pathway redesign | Reduced administrative layers from 7 to 4 | Clinical decision speed improved by 47%; administrative overhead reduced by 34%; maintained patient safety |
Unilever | 2023 | Capability-driven redesign / Communication Strategy | "Careers Reimagined" campaign; horizontal career progression roadmaps; updated compensation frameworks | Compressed commercial organization from 6 layers to 4 | 78% of employees reported clear understanding of new career architecture (vs. 31% previously) |
T-Mobile | 2022 | Capability-driven redesign | Contribution matrix (mapping functional expertise, business context breadth, and stakeholder management range) | Eliminated 2 organizational layers | Expanded average career tenure of high performers by 31%; reduced time-to-competence for critical roles by 40% |
Schneider Electric | 2022 | Capability-driven redesign / Passage Compression | Passage acceleration programs (90-day intensive exposure, cross-functional projects, P&L reviews, embedded coaching) | Restructured commercial organization from 6 leadership levels to 4 | 91% success rate for promoted executives compared to 67% previously |
Redesigning Supervision Around AI Capability
The most fundamental intervention for organizations attempting authentic compression rather than cost-driven delayering involves redesigning the supervision function around what AI systems can now perform effectively versus what still requires human judgment and relationship.
Research on AI-augmented management identifies three supervision categories with distinct automation potential: Administrative coordination (scheduling, task assignment, progress tracking, basic resource allocation) is highly automatable and currently performed at human-equivalent or better quality by AI systems. Performance assessment and feedback (evaluating work quality, identifying development needs, delivering coaching) is partially automatable, with AI systems effective for quantitative assessment but still requiring human judgment for qualitative evaluation and relationship management. Strategic direction and motivation (setting team purpose, making complex tradeoffs, managing interpersonal dynamics) remains minimally automatable and requires human leadership (Boudreau & Cascio, 2017).
Organizations pursuing effective compression focus AI deployment on administrative coordination, freeing human managers to focus on performance development and strategic direction. The span-of-control expansion this enables is dramatic. When managers no longer spend time on scheduling, task assignment, or progress tracking—work AI performs continuously and more accurately—their viable team size expands from 8-10 direct reports to 15-20 without corresponding workload increase (Charan, 2025).
Implementation approaches that evidence supports:
Automated task coordination platforms that assign work based on current capacity, skill match, and development goals, eliminating manager time spent on daily work allocation
AI-driven progress dashboards that surface exceptions requiring human attention rather than requiring managers to manually monitor all team activities
Automated performance data compilation that aggregates quantitative performance signals, allowing managers to focus interpretation and coaching conversations on the data rather than gathering it
Intelligent scheduling systems that manage meeting coordination, time-off approvals, and resource booking without manager involvement
Natural language processing tools that analyze team communication patterns and surface relationship issues or collaboration gaps for manager attention
Mastercard implemented AI-augmented supervision across its technology organization in 2023, focusing specifically on automating administrative coordination while preserving human management of performance development and strategic direction. The company deployed an AI platform that handles task assignment, progress tracking, and resource allocation for development teams, while human managers focus on code quality assessment, technical skill development, and career planning. The result was expansion of average manager span of control from 9 to 14 direct reports without corresponding increases in manager workload or declines in team engagement. The technology organization eliminated two management layers while maintaining development velocity and improving time-to-promotion for high performers by 23% through more focused human attention on development conversations (Mastercard, 2024).
Horizontal Capability Development as Primary Value Creation
Organizations serious about architectural compression must redesign development pathways around horizontal capability building rather than vertical progression. This requires fundamental changes to promotion criteria, compensation progression logic, and performance assessment frameworks.
Research on career mobility patterns demonstrates that employees who build capability across multiple functions or business contexts create more organizational value than those who build deep expertise in a single function, particularly in rapidly changing environments where strategic priorities shift faster than deep expertise can be deployed (Sullivan & Baruch, 2009). Yet most compensation and promotion systems still reward vertical progression and functional depth over horizontal range.
The core intervention involves changing the primary progression logic from title advancement to contribution range expansion. Rather than measuring career progress by movement up hierarchical levels, organizations measure it by the breadth of problems an individual can effectively solve, the range of contexts in which they can operate effectively, and the diversity of stakeholder groups they can work with productively.
Effective implementation practices:
Contribution-based compensation progression that increases pay based on demonstrated capability across multiple business contexts rather than title changes
Diagonal movement incentives that explicitly reward lateral moves into unfamiliar functions with compensation increases equivalent to traditional promotions
Portfolio career frameworks that encourage employees to maintain multiple part-time roles across different business units simultaneously rather than single full-time assignments
Cross-functional project staffing that makes project-based work the primary mode of contribution rather than permanent functional assignments
Capability mapping systems that track individual skill development across multiple dimensions rather than performance within a single role scope
T-Mobile restructured its talent development architecture in 2022 around horizontal capability building, explicitly abandoning traditional promotion-based progression. The company created a contribution matrix that maps employee capability across three dimensions: functional expertise, business context breadth (consumer, enterprise, wholesale), and stakeholder management range (frontline, executive, partner ecosystem). Compensation progression is tied to expansion across all three dimensions rather than title changes. Employees advance their compensation by demonstrating they can solve problems in multiple business contexts, not by managing larger teams or gaining VP titles. The company simultaneously eliminated two organizational layers while expanding the average career tenure of high performers by 31% and reducing time-to-competence for critical roles by 40% (T-Mobile, 2023).
Distributed Leadership Structures That Scale Decision Authority
A third critical intervention involves redesigning decision rights and leadership structures to distribute authority more broadly rather than concentrating it at higher hierarchical levels. Traditional hierarchies concentrate decision authority at senior levels under the assumption that seniority correlates with judgment quality. In compressed architectures with wider spans and fewer layers, this concentration creates decision bottlenecks that prevent organizations from operating at required speed.
Research on organizational decision-making demonstrates that decision quality is less correlated with hierarchical level than with proximity to relevant information and frequency of similar decisions (Aghion & Tirole, 1997). An individual contributor making technology platform decisions weekly will typically make better platform decisions than a senior executive making them quarterly, regardless of the executive's broader strategic perspective. The implication is that architectural compression should push decision authority downward toward information and frequency, not maintain it at hierarchical levels the organization has just eliminated.
Evidence-supported implementation approaches:
Domain-based decision authority that assigns decision rights based on knowledge domain rather than hierarchical level, allowing IC-level technical experts to make architecture decisions without senior approval
Financial authorization by decision frequency that grants higher spending authority to individuals making frequent procurement decisions than to executives making occasional large decisions, recognizing that decision quality in procurement correlates more with frequency than with amount
Rotating leadership assignments that distribute strategic decision-making across teams rather than concentrating it in permanent leadership roles, building decision-making capability while preventing bottlenecks
Transparent decision frameworks that specify exactly which decisions require approval, which require consultation, and which can be made autonomously, eliminating ambiguity about authority boundaries
AI-supported decision quality monitoring that tracks decision outcomes across the organization and provides feedback to decision makers regardless of level, creating learning loops that improve judgment
Mayo Clinic restructured clinical decision authority in its hospital operations in 2023, moving from traditional hierarchical approval structures to domain-based authority assignment. Clinical decisions are now made by the clinician closest to the patient with relevant specialization, regardless of seniority. Resource allocation decisions are made by the team consuming the resources with input from finance partners but without executive approval for amounts below specified thresholds. Strategic decisions about care pathway redesign are made by rotating teams of clinicians, administrators, and data analysts rather than permanent executive committees. The organization simultaneously reduced its administrative layers from seven to four while improving clinical decision speed by 47%, reducing administrative overhead by 34%, and maintaining patient safety outcomes (Mayo Clinic, 2024).
Explicit Passage Compression with Structured Capability Scaffolding
For organizations that must compress leadership passages—promoting individuals through multiple traditional levels simultaneously—the evidence is clear that passage compression fails without corresponding capability scaffolding. The traditional approach of patient sequential progression through leadership levels succeeded not because the waiting period itself developed capability but because each level provided structured exposure to progressively complex challenges with contained failure consequences.
When organizations compress passages, they must replace time-based capability development with intensity-based development. This involves providing condensed exposure to the challenges the eliminated passages would have addressed, creating accelerated feedback loops, and offering supported practice opportunities for skills the employee has not yet demonstrated.
Research on accelerated leadership development identifies three components that enable successful passage compression: structured complexity exposure that provides direct experience with the full range of challenges the new level entails before the individual is evaluated on their performance; frequent formative feedback delivered weekly rather than annually, allowing rapid skill adjustment; and executive coaching support that provides external perspective and pattern recognition from someone who has successfully navigated similar compressed transitions (Petrie, 2014).
Effective implementation practices include:
Passage acceleration programs that provide 90-day intensive exposure to the next two leadership levels before promoting someone through both levels simultaneously
Embedded coaching that assigns external executive coaches to individuals making compressed passages for the first six months in the new role
Peer learning cohorts that group individuals making similar compressed passages to share experiences and problem-solving approaches
Simulation-based practice that uses business simulations and case methods to provide safe practice with challenges the individual will face in compressed roles
Fractional role assignments that allow individuals to perform portions of higher-level roles while maintaining their current position, building capability before full transition
Schneider Electric implemented explicit passage compression with structured scaffolding in 2022 when restructuring its commercial organization from six leadership levels to four. Rather than promoting regional managers directly to country leaders without intermediate passage, the company created 90-day passage acceleration programs. Individuals being promoted from regional manager to country leader spend 90 days in a program that includes: leading a cross-functional project team to simulate managing peer leaders; participating in three business unit P&L reviews to gain exposure to full business complexity; completing a strategic planning cycle for a small business unit under coaching supervision; and attending executive coaching sessions every two weeks. Only after completing the program do they formally transition to the country leader role. The approach enabled successful compression of two traditional passages while maintaining a 91% success rate for promoted executives, compared to a 67% success rate in the pre-compression architecture (Schneider Electric, 2023).
Transparent Communication About Architectural Reality
A final critical intervention involves honest, explicit communication about the fact that architectural compression has occurred and that career progression logic has fundamentally changed. Most organizations pursuing delayering communicate the structural change—"we are reducing layers"—without communicating the implications for how careers now develop. The result is employees continuing to pursue vertical progression strategies in organizations that now value horizontal capability, creating the structural contract violations discussed earlier.
The evidence on change communication is unambiguous: when organizations communicate structural changes without explaining the implications for individual behavior and career strategy, employees continue pre-change behavioral patterns until they experience negative consequences, at which point both engagement and trust decline sharply (Elving, 2005). Effective change communication requires not just announcement of what has changed structurally but explicit guidance on what success strategies are now effective versus obsolete.
Evidence-based communication practices:
Career progression roadmaps that explicitly contrast old vertical progression paths with new horizontal capability development paths, showing specific examples of how careers now advance
Success story spotlighting that features employees who have successfully built careers through horizontal moves and capability expansion, providing concrete models of effective navigation
Manager conversation guides that equip people managers to discuss career development in horizontal rather than vertical terms with their teams
Compensation transparency about how pay now progresses based on capability expansion rather than title changes
Regular architectural updates that communicate ongoing compression changes as they occur rather than treating restructuring as one-time events
Unilever implemented comprehensive communication about career architecture changes in 2023 when compressing its commercial organization from six layers to four. Rather than simply announcing the layer elimination, the company launched a multi-month campaign called "Careers Reimagined" that included: a detailed guide contrasting old and new career progression logic with specific examples; a series of "career conversations" where senior leaders shared their own career paths and how those paths would look different under the new architecture; monthly all-hands sessions where employees could ask questions about career implications; and updated compensation frameworks that explicitly showed how pay advances based on capability range rather than title. Post-communication surveys showed 78% of employees reported clear understanding of how to build careers in the new architecture, compared to 31% understanding after previous restructurings that lacked equivalent communication investment (Unilever, 2024).
Building Long-Term Organizational Capability in Compressed Architectures
Continuous Role Redesign Systems
The most critical mistake organizations make with architectural compression is treating it as a one-time event rather than an ongoing capability. The forces driving compression—AI capability advancement, speed requirements, cost pressure—are not static shocks that organizations absorb and then stabilize. They are continuous pressures that require continuous architectural adaptation. Organizations that compress once and then attempt to operate the compressed architecture as a stable system will find themselves requiring further compression within two to three years, creating repeated disruption cycles that destroy organizational capability.
What is required instead is continuous role redesign capability—organizational systems that regularly examine which work AI can now perform that previously required human attention, which decision rights can be pushed lower in the organization as individuals gain capability, which coordination mechanisms can be replaced by platforms, and which roles have become obsolete or need fundamental redesign.
Research on organizational adaptability demonstrates that companies which build continuous redesign capability outperform those that pursue episodic restructuring by significant margins—higher revenue growth, better employee retention, and faster product development cycles (Worley & Lawler, 2010). The competitive advantage comes not from achieving perfect architectural design but from maintaining shorter feedback loops between environmental change and organizational response.
Key components of continuous redesign systems:
Quarterly role audits that examine each role type to identify work that has become automatable or could be performed more effectively at different organizational levels
AI capability monitoring that tracks advancements in AI supervision, coordination, and decision support capabilities and identifies which organizational processes could be redesigned around new capabilities
Work distribution analytics that measure how time is spent across roles and surfaces misalignments between formal role design and actual work patterns
Individual contributor input mechanisms that gather systematic input from people doing the work about which processes create friction, which approvals add no value, and which coordination mechanisms could be streamlined
Rapid redesign protocols that enable fast role and process changes without requiring enterprise-wide restructuring initiatives
AI-Human Partnership Development
As AI assumes more supervision, coordination, and administrative work, the nature of productive human contribution shifts. Organizations need systematic capability building around effective AI partnership—understanding what AI does well versus poorly, providing appropriate direction and guardrails, interpreting AI outputs, and making judgments about when to override AI recommendations.
Current workforce data suggests most organizations are deploying AI tools faster than they are building human capability to use them effectively. A 2024 study of AI adoption in enterprise settings found that less than 30% of employees using AI tools in their work have received formal training on effective AI interaction, prompt engineering, or output quality assessment (IBM, 2024). The result is underutilization of deployed AI capability and, more seriously, uncritical acceptance of AI outputs that may contain errors or biases.
Research on human-AI collaboration identifies several capabilities critical for effective partnership: appropriate reliance (knowing when to trust AI recommendations versus when to override them); effective prompting (providing AI systems with sufficient context and constraints to produce useful outputs); output interpretation (understanding AI's reasoning process sufficiently to evaluate whether its recommendations make sense in context); and bias detection (recognizing when AI outputs reflect biases in training data rather than genuine patterns) (Jarrahi et al., 2023).
Practical capability-building approaches:
AI partnership training that teaches employees how to effectively direct AI systems, evaluate outputs, and make appropriate trust-versus-override decisions
Prompt libraries that share effective prompts for common business tasks across the organization, accelerating individual learning about AI interaction
AI output review protocols that build systematic review steps into workflows where AI is making recommendations, catching errors before they create consequences
Human-AI teaming structures that explicitly assign complementary tasks to humans versus AI rather than treating AI as a general assistant
Failure case documentation that captures instances where AI recommendations were incorrect and analyzes why, building organizational pattern recognition
Purpose-Driven Work Architecture
A less obvious but equally critical long-term capability involves redesigning work architecture around purpose and contribution rather than tasks and hierarchy. When vertical progression is no longer the primary career organizing principle, something must replace it as the source of meaning and direction in work. Research on work motivation consistently demonstrates that clarity of purpose and visible contribution to meaningful outcomes drives engagement and performance more powerfully than compensation or career advancement (Grant, 2008).
In compressed architectures with horizontal progression, individuals need clear line of sight from their daily work to organizational impact. This requires different work design than hierarchical architectures, where meaning often came from position status and upward progression. The shift is from hierarchical meaning (my work matters because of where I sit in the structure) to contribution meaning (my work matters because of the problems it solves and the impact it creates).
Organizations building sustainable compressed architectures invest heavily in making contribution visible and connecting individual work to customer, stakeholder, and societal outcomes. This is not motivational messaging but structural work design—building feedback loops that show individuals the concrete outcomes their work produces.
Effective purpose integration practices:
Direct stakeholder exposure that connects employees regularly with the customers, patients, or communities their work serves
Impact dashboards that track and display the outcomes of work rather than just activity completion
Contribution narratives that tell stories connecting individual work to organizational mission achievement
Mission-aligned projects that allow employees to volunteer for work directly connected to organizational purpose beyond their functional assignments
Stakeholder feedback loops that bring voice-of-customer, voice-of-community, or voice-of-patient data directly to employees performing relevant work
Kaiser Permanente redesigned clinical and administrative work architecture around direct patient impact visibility in 2023. Administrative staff who previously had no patient contact now spend one day quarterly in clinical settings observing how their work affects care delivery. Clinical support staff receive monthly summaries of patient outcomes in cases they supported. IT teams building new systems participate in user testing sessions with both clinicians and patients. The organization eliminated two administrative layers while improving both employee engagement (up 28%) and administrative process speed (up 41%), attributing both improvements to stronger purpose connection offsetting the challenges of wider spans and fewer progression opportunities (Kaiser Permanente, 2024).
Conclusion
The career ladder has collapsed not because organizations chose to eliminate it but because the structural conditions that made hierarchical progression functional have disappeared. AI has assumed the supervision and coordination work that required multiple management layers. Speed requirements have made traditional approval chains operationally untenable. Cost pressures have forced layer elimination faster than most organizations can redesign the work those layers performed. The question is no longer whether architectural compression will occur but whether organizations will pursue it as genuine capability redesign or disguised cost reduction.
The evidence presented demonstrates that most current compression efforts are cost-driven delayering rather than capability-driven redesign. Layers are eliminated, work is redistributed, and the processes, decision rights, and coordination mechanisms of the old architecture remain largely unchanged. The predictable consequences—declining manager engagement, wider spans without role redesign, decision bottlenecks, coordination gaps, and capability development failures—are not implementation problems but design problems. They are the inevitable result of removing organizational layers without redesigning the work those layers performed.
Genuine compression requires five core interventions. First, redesigning supervision around AI capability, determining which elements of management work AI now performs more effectively than humans and restructuring human manager roles around performance development and strategic direction. Second, shifting development pathways from vertical progression to horizontal capability building, changing promotion criteria, compensation logic, and success definitions to reward contribution range rather than hierarchical advancement. Third, distributing decision authority based on information proximity and decision frequency rather than hierarchical level, preventing the decision bottlenecks that compressed structures create when authority remains concentrated at senior levels. Fourth, providing explicit capability scaffolding when compressing leadership passages, replacing time-based development with intensity-based exposure to complexity. Fifth, communicating transparently about architectural reality, explaining not just that structure has changed but how career progression logic and success strategies must adapt.
The organizations that will succeed in compressed architectures are not those that execute delayering most aggressively but those that build continuous redesign capability—systems that regularly reassess which work AI now performs better than humans, which decisions can be pushed lower as individuals gain capability, and which roles have become obsolete or require fundamental reconstruction. The competitive advantage is not achieving perfect architectural design but maintaining faster adaptation cycles than competitors.
For individuals, the shift is equally fundamental. The career strategy that built success in hierarchical organizations—building deep functional expertise, cultivating relationships with direct supervisors, demonstrating consistent performance within defined scope—will not build success in compressed architectures that value horizontal range, tolerance for ambiguity, and capability across multiple business contexts. The psychological transition required is substantial. It involves letting go of vertical progression as the primary definition of career success, developing comfort with lateral movement and portfolio careers, and building identity around contribution range rather than hierarchical position.
The most important insight is that this transition has already occurred in organizational reality even as it remains incomplete in organizational infrastructure and individual mental models. The career ladder is not dying slowly. It has already collapsed. The organizations still operating as though vertical progression remains the primary career logic are creating systematic contract violations with their workforces—promising progression pathways that no longer exist, rewarding capabilities the organization no longer values, and developing people for roles that have been eliminated. The individual and organizational consequences of this misalignment will compound until leadership acknowledges that the architecture has fundamentally changed and that ladder-era tools cannot manage post-ladder reality.
The practical imperative for leadership is immediate. Audit whether your last restructuring was cost-driven delayering or capability-driven compression. If managers are spending 40% or more of their time on individual contributor work, if spans have widened without corresponding role redesign, if you eliminated layers without redesigning supervision around AI capability or redistributing decision authority, you have pursued cost reduction disguised as architectural compression. The engagement declines, retention problems, and execution gaps you are now experiencing are design consequences, not implementation failures. The question is whether you will acknowledge the design error and pursue genuine compression, or continue managing the symptoms of misalignment until organizational capability erodes beyond recovery.
Research Infographic

References
Aghion, P., & Tirole, J. (1997). Formal and real authority in organizations. Journal of Political Economy, 105(1), 1-29.
Boudreau, J., & Cascio, W. (2017). Human capital analytics: Why are we not there? Journal of Organizational Effectiveness: People and Performance, 4(2), 119-126.
Bruch, H., & Menges, J. I. (2010). The acceleration trap. Harvard Business Review, 88(4), 80-86.
Cascio, W. F., & Boudreau, J. W. (2017). Investing in people: Financial impact of human resource initiatives (3rd ed.). Society for Human Resource Management.
Charan, R. (2025). The AI age: Rethinking leadership passages. Harvard Business Review.
Corporate Leadership Council. (2003). Hallmarks of leadership success: Strategies for improving leadership quality and executive readiness. Corporate Executive Board.
Elving, W. J. L. (2005). The role of communication in organizational change. Corporate Communications: An International Journal, 10(2), 129-138.
Gallup. (2025). State of the global workplace: 2025 report. Gallup Press.
Grant, A. M. (2008). Does intrinsic motivation fuel the prosocial fire? Motivational synergy in predicting persistence, performance, and productivity. Journal of Applied Psychology, 93(1), 48-58.
IBM. (2024). Global AI adoption index 2024. IBM Institute for Business Value.
Jarrahi, M. H., Askay, D., Eshraghi, A., & Smith, P. (2023). Artificial intelligence and knowledge management: A partnership between human and AI. Business Horizons, 66(1), 87-99.
Kaiser Permanente. (2024). Purpose-driven care: 2024 organizational impact report. Kaiser Permanente.
Mastercard. (2024). Technology organization transformation: Implementation report. Mastercard.
Mayo Clinic. (2024). Clinical decision authority restructuring: Outcomes report. Mayo Clinic.
Petrie, N. (2014). Vertical leadership development—Part 1: Developing leaders for a complex world. Center for Creative Leadership.
Robinson, S. L., & Morrison, E. W. (2000). The development of psychological contract breach and violation: A longitudinal study. Journal of Organizational Behavior, 21(5), 525-546.
Schneider Electric. (2023). Passage compression program: Implementation and outcomes. Schneider Electric.
Sullivan, S. E., & Baruch, Y. (2009). Advances in career theory and research: A critical review and agenda for future exploration. Journal of Management, 35(6), 1542-1571.
T-Mobile. (2023). Horizontal career architecture: Implementation guide and results. T-Mobile.
Thompson, N., & Autor, D. (2024). How AI changes the value of labor. National Bureau of Economic Research Working Paper Series.
Unilever. (2024). Careers reimagined: Communication strategy and impact assessment. Unilever.
Workhuman. (2024). The state of workplace restructuring: 2024 analysis. Workhuman Research Institute.
Worley, C. G., & Lawler, E. E. (2010). Agility and organization design: A diagnostic framework. Organizational Dynamics, 39(2), 194-204.
Zenger Folkman. (2023). Leadership derailment in compressed organizations: Research findings. Zenger Folkman.

Jonathan H. Westover, PhD, Chief Research Officer (Nexus Institute for Work and AI); Co-Founder & Chief Workforce and Learning Officer (Future State University); Founder & CEO (Human Capital Innovations); Professor of Organizational Leadership & Change (UVU). Read Jonathan Westover's executive profile here.
Suggested Citation: Westover, J. H. (2026). The Architecture of Work Is Collapsing. Most Organizations Are Still Building Ladders.. Human Capital Leadership Review, 39(2). doi.org/10.70175/hclreview.2020.39.2.3






















