When Competent Leaders Cannot See the Whole System

Organizational Complexity, Analytical Distance, and the Limits of Managerial Visibility

An external analyst should begin an organizational engagement with a presumption that consulting practice can sometimes obscure: the people leading an organization generally know far more about their business than the analyst does. They understand its history, customers, workforce, constraints, technologies, regulatory environment, informal relationships, and the accumulated rationale behind decisions that may appear unusual when viewed without context. They also bear responsibility for consequences that an outside observer does not. Respect for that knowledge is not merely professional courtesy; it is an important epistemological constraint on organizational analysis. An analyst who begins by assuming that organizational difficulties exist because leadership has failed to recognize something obvious risks confusing analytical distance with superior understanding.

The converse, however, is equally important. Expertise in leading an organization is not the same as the capacity to observe that organization comprehensively as a system. Complex organizations contain more relationships, information, feedback processes, competing demands, and localized interpretations than any individual leader can continuously apprehend. Herbert Simon's work on administrative decision-making challenged assumptions of comprehensive rationality by demonstrating that decision-makers operate within unavoidable limits of information, attention, computational capacity, and time. Organizational actors therefore make decisions under conditions of bounded rationality rather than from a position of complete knowledge (Simon, 1997). This does not imply deficient leadership. It suggests that even highly capable leaders necessarily work from partial representations of the organizations for which they are responsible.

This distinction provides a more constructive starting point for organizational diagnosis. Persistent performance problems do not necessarily indicate that leaders are incompetent, employees are resistant, or organizations have failed to articulate sufficiently compelling missions or strategies. In many cases, competent people are making reasonable decisions within systems whose aggregate behavior has become difficult to interpret from any single position inside them. The analytical problem is therefore not simply to identify who made an incorrect decision. It is to understand how organizational purpose, culture, strategy, structures, routines, incentives, information, human behavior, and performance outcomes interact over time.

The Limits of Managerial Visibility

Organizations are deliberately divided into roles, functions, hierarchies, and specialized units because no individual can perform or directly supervise all organizational work. Specialization makes complex collective activity possible, but it also distributes knowledge. Senior executives tend to possess greater visibility into strategic priorities, external relationships, financial conditions, and enterprise-level performance, while managers and frontline employees often possess deeper knowledge of operational constraints, customer interactions, process variation, and the practical consequences of organizational decisions. Neither perspective is intrinsically superior. Each is structurally incomplete.

Karl Weick's work on sensemaking adds another dimension to this problem. Organizational actors do not merely receive an objective reality and report it upward unchanged. They interpret ambiguous circumstances using their experiences, identities, expectations, social relationships, and available cues. Sensemaking is therefore an active organizational process through which people construct workable interpretations of complex situations (Weick, 1995). Two groups within the same organization can encounter the same initiative and arrive at materially different assessments without either group being dishonest or irrational. An executive may understand a restructuring as the implementation of a deliberate strategic priority, while frontline employees experience the same restructuring primarily through changes in workload, decision authority, customer interactions, and daily routines. Both accounts may accurately represent different dimensions of the same organizational event.

For organizational analytics, this has an important implication: perceptual disagreement should not automatically be treated as measurement error. A difference between executive and workforce perceptions may reflect poor communication, but it can also reveal differences in access to information, organizational position, lived experience, incentives, or interpretation. The analytical task is not necessarily to determine which group possesses the “correct” view. It is to investigate what conditions could plausibly produce the divergence and whether other evidence supports one interpretation, several interpretations, or an explanation that neither group has fully articulated.

This is especially relevant when organizations attempt to evaluate alignment. A mission statement, strategic plan, organizational value, performance dashboard, and employee survey each provide information about the organization, but they describe different aspects of it. Treating them as interchangeable evidence obscures important distinctions. Mission expresses intended purpose. Strategy describes choices about how that purpose will be pursued. Organizational ethos reflects shared assumptions and normative expectations about how work should be conducted. Formal systems establish roles, processes, and controls. Employees experience and enact those systems in practice. Performance measures attempt to describe selected consequences. Whether these components are coherent cannot be established by examining any one of them independently.

Formal Organizations and Organizations in Practice

One of the persistent challenges in organizational analysis is the difference between how work is formally represented and how it is actually performed. Feldman and Pentland's (2003) work on organizational routines provides a particularly useful framework for understanding this distinction. They differentiate between the ostensive aspect of a routine—the generalized or abstract understanding of how the routine works—and its performative aspect—the specific actions undertaken by particular people in particular circumstances. Organizational routines are therefore not simply static procedures replicated mechanically; their performance can simultaneously reproduce, interpret, and modify the routine itself.

This distinction has substantial practical consequences. A policy establishes evidence of intended practice, but it cannot by itself establish performed practice. A training record demonstrates that instruction occurred, but not necessarily that the resulting behavior was adopted consistently. An automated workflow may encode a formal process while employees simultaneously develop informal workarounds to accommodate conditions that the system did not anticipate. Conversely, variation from a formal process should not automatically be interpreted as noncompliance or employee resistance. The variation may represent adaptive behavior that allows the larger system to function despite weaknesses in its formal design.

Strategy exhibits a similar distinction between intention and organizational realization. Mintzberg and Waters (1985) distinguished deliberate strategy from emergent strategy, demonstrating that realized strategy can include both intended patterns of action and patterns that arise through organizational learning and adaptation. The existence of a strategic plan therefore does not, by itself, demonstrate that the organization is behaving strategically in the manner described by the plan. The strategy that ultimately becomes visible in resource allocation, operational priorities, management attention, incentives, and everyday decision-making may differ from the strategy that was formally articulated.

This is one reason that statements such as “mission, ethos, strategy, and results should align,” while intuitively appealing, are analytically incomplete. The more consequential question is how alignment can be demonstrated. A mission may emphasize service quality while operating incentives reward throughput. A strategy may prioritize workforce development while staffing models leave supervisors little time for coaching. Organizational values may emphasize candor while employees perceive substantial interpersonal risk in reporting problems. A quality system may define a standardized process while operational evidence reveals extensive dependence on informal adaptation. These conditions do not necessarily indicate hypocrisy or poor leadership. They indicate that the organization contains relationships that deserve investigation.

Edgar Schein's work on organizational culture further supports this distinction. Schein (2010) conceptualized organizational culture as operating at multiple levels, including visible artifacts, espoused beliefs and values, and deeper underlying assumptions. Consequently, what an organization formally says about itself and what its members have learned to assume about how the organization actually operates may not be identical. This gap is not unusual; it is precisely the kind of organizational phenomenon that careful analysis should attempt to understand rather than simply condemn.

More Data Do Not Necessarily Produce More Understanding

Modern organizations frequently possess extraordinary amounts of information. Performance dashboards, financial reporting systems, employee surveys, quality audits, customer feedback platforms, compliance systems, incident databases, operational applications, project management tools, and increasingly sophisticated analytics platforms can produce a nearly continuous stream of measures. Yet an abundance of data does not eliminate the problem Simon identified. Indeed, data proliferation can create a different manifestation of bounded rationality: the organization may possess more information than its leaders have the capacity to integrate meaningfully.

The distinction between measurement and understanding is especially important in complex systems. Sterman's (2000) work in system dynamics emphasizes feedback, delays, nonlinear relationships, accumulation, and the difficulty human decision-makers experience when attempting to infer system behavior from isolated observations. Actions frequently produce consequences that emerge later, elsewhere, or through indirect pathways. An intervention that improves one metric in the short term may degrade another condition that is not immediately visible. Similarly, an intervention may appear ineffective because its intended outcome is delayed even though underlying conditions have begun to change.

Consider an organization in which primary customer-service measures remain stable after implementation of a major workflow change. Leadership may reasonably interpret that stability as evidence that the implementation is succeeding. At the same time, employees may report substantially greater workload, supervisors may notice growing dependence on informal workarounds, overtime may increase, and minor customer complaints may begin trending upward. These observations are not necessarily contradictory. One plausible interpretation is that employees are temporarily preserving external performance by consuming additional internal capacity. If that interpretation is correct, the apparent stability of the headline performance measure may conceal deterioration elsewhere in the system.

The point is not that this explanation should automatically be accepted. It is a causal hypothesis requiring evidence. The purpose of multidimensional analysis is precisely to distinguish a plausible story from a sufficiently supported explanation. Trends in overtime, staffing, employee perceptions, process exceptions, customer experience, throughput, rework, error rates, and subsequent performance might collectively strengthen or weaken the hypothesis. The appropriate analytical posture is neither to accept leadership's interpretation uncritically nor to assume that workforce dissatisfaction reveals the “real” explanation. It is to test competing interpretations against multiple forms of evidence.

This approach is consistent with Kaplan and Norton's work on strategy mapping, which emphasizes relationships among organizational capabilities, internal processes, customer outcomes, and financial or mission-level results rather than treating performance measures as unrelated indicators (Kaplan & Norton, 2004). Their work is particularly relevant because it shifts strategic measurement away from isolated metrics toward hypothesized relationships through which organizational assets and processes are expected to produce value.

However, even a well-designed strategy map remains a model rather than the organization itself. The existence of an assumed relationship between workforce capability, process performance, customer outcomes, and organizational results does not establish that the relationship is operating as expected. The model must remain subject to evidence. When observed results contradict the organization's strategic assumptions, the appropriate response is not merely to search for a weak metric. It may be necessary to reconsider the underlying theory of how the organization expects value to be created.

From Root Cause to Causal Configuration

The language of root-cause analysis is useful because it encourages analysts to move beyond symptoms. It can become problematic, however, when it implies that complex organizational outcomes always have a single discoverable origin. Many organizational problems are better understood as products of interacting conditions. Turnover, for example, may involve compensation, supervisory relationships, workload, role clarity, labor-market conditions, technology burden, psychological safety, scheduling, career development, organizational culture, or several of these factors operating together. The relative importance of each condition may also vary across units or employee populations.

James Reason's distinction between person-centered and systems approaches to human error is instructive. Rather than focusing exclusively on the individual who committed an error, a systems perspective examines the conditions within which human behavior occurs and the organizational defenses that shape whether errors become consequential (Reason, 2000). Nancy Leveson's systems-theoretic approach similarly challenges linear accident models in complex sociotechnical systems and emphasizes interactions, control structures, constraints, and broader system behavior (Leveson, 2012). Although both bodies of work emerged largely from safety science, their underlying caution is relevant to organizational analysis more generally: explaining a complex outcome exclusively through the most visible individual action can obscure the conditions that made the outcome possible or likely.

Soft Systems Methodology provides a complementary perspective. Checkland and Scholes (1990) developed their approach for complex human problem situations in which the definition of the problem itself may be contested. Rather than assuming that analysts can begin with a universally agreed-upon problem and engineer the optimal solution, soft systems thinking treats inquiry as a learning process involving multiple perspectives and interpretations. This orientation is particularly useful for organizational consulting because the presenting problem may not ultimately be the most useful unit of analysis. What leadership initially describes as an accountability problem may involve process design, information flow, resource constraints, conflicting goals, or ambiguous decision authority. What employees describe as a staffing problem may partially reflect unnecessary process complexity or technology friction. The purpose of analysis is not to invalidate either perspective but to construct a better-supported representation of the problem situation.

For that reason, multidimensional root-cause analysis should generally be understood as the disciplined development and testing of competing causal explanations rather than a search for a single convenient culprit. Quantitative trends may establish where and when a problem occurs. Qualitative evidence may identify recurring mechanisms or experiences. Process evidence may show differences between intended and performed practice. Organizational structure may reveal how decisions or information move. Customer evidence may demonstrate external consequences. Workforce perceptions may reveal conditions that operational measures do not capture. None of these sources is independently definitive. Their value increases when relationships among them can be examined.

When Disagreement Becomes Evidence

One of the most useful consequences of this approach is that disagreement among evidence sources does not have to be eliminated before analysis can proceed. Divergence can itself become an object of inquiry. If executives assess strategic communication positively while frontline employees report substantial ambiguity, the gap may indicate inadequate communication. It could also reflect legitimate differences in what each group needs to know, inconsistent middle-management interpretation, competing operational priorities, or a strategic message that is clear conceptually but difficult to translate into daily decisions. Each explanation generates different questions and requires different evidence.

The same is true of discrepancies between reported and observed practice. Edmondson's (1999) research on psychological safety and team learning demonstrated that interpersonal conditions can influence whether people feel able to speak openly about mistakes, questions, and concerns. This has an important implication for organizational analytics: the absence of reported problems cannot always be treated as evidence of the absence of problems. Reporting behavior is itself part of the system through which organizational information is produced.

Conversely, a high volume of reported problems does not automatically establish poor organizational performance. In some contexts, greater reporting may reflect increased visibility, stronger reporting systems, or greater confidence that concerns can be raised without adverse consequences. An organization that becomes more transparent about defects may initially appear to deteriorate on a metric that counts defects even while its capacity for organizational learning is improving. Interpreting such evidence requires understanding how the measure is generated rather than examining its numerical value in isolation.

This principle extends to perception surveys. Perception should neither be dismissed as “only subjective” nor treated as a substitute for objective performance evidence. Perceptions are data about how organizational conditions are being interpreted and experienced. Their analytical value becomes much greater when they can be compared across organizational levels, functions, time periods, operational indicators, and qualitative evidence. A large perceptual gap between leaders and employees is not itself a diagnosis. It is a signal that deserves explanation.

Analytical Distance Without Epistemic Arrogance

The potential value of an external analyst therefore arises from analytical distance rather than presumed superiority. Organizational insiders possess contextual depth; external analysts can contribute methodological structure, comparative reasoning, and a vantage point less embedded in established assumptions and routines. The objective should be to combine those forms of knowledge rather than elevate one over the other.

This is also why consulting models that depend heavily on authoritative recommendation can be problematic in complex systems. The analyst may identify relationships that are difficult for organizational insiders to observe, but the analyst remains dependent on those insiders to understand context, feasibility, history, and consequences. A technically elegant recommendation that ignores operational knowledge can easily produce a new layer of systemic complexity rather than resolving the original problem.

A more appropriate relationship is collaborative but not uncritical. The analyst should take leadership's interpretation seriously without treating it as incontrovertible. Workforce perceptions deserve serious consideration without being presumed inherently more authentic. Documentary evidence should establish what the organization has formally designed without being mistaken for proof of implementation. Quantitative relationships should inform causal reasoning without being casually converted into causal conclusions. Qualitative evidence should deepen interpretation while remaining subject to questions of representativeness and alternative explanation. The purpose is disciplined triangulation.

In this sense, analytical humility does not require analytical timidity. An evidence-based assessment should be willing to conclude that a prevailing organizational explanation is poorly supported when the evidence warrants that judgment. It should also be willing to conclude that an apparent problem is less consequential than initially believed, that a fashionable intervention lacks evidentiary justification, or that additional evidence is necessary before leadership acts. Humility concerns the limits of what the analyst can know; rigor concerns the quality of the conclusions that can reasonably be drawn from what is known.

From Diagnosis to Leadership Refocus

Organizational analysis becomes useful when it improves the quality of action. A highly sophisticated explanation that leaves leadership with no clearer understanding of where to focus is analytically interesting but operationally incomplete. Systems thinking therefore requires not only an explanation of relationships but also disciplined consideration of intervention.

Donella Meadows' work on leverage points is useful in this respect. Meadows (1999) argued that potential interventions within systems differ substantially in their capacity to alter system behavior. Changes to parameters may matter, but changes to information flows, rules, feedback structures, goals, or the capacity of the system to adapt can produce more substantial effects. Her framework should not be treated as a mechanical hierarchy applicable identically to every organization; its more important contribution is the recognition that the most visible problem is not necessarily the most consequential place to intervene.

Leadership refocus, therefore, should follow diagnosis rather than precede it. If poor customer response time is being produced primarily by unnecessary approvals, adding frontline employees may improve capacity without addressing the structural source of delay. If turnover reflects inconsistent supervision rather than market compensation, a pay adjustment may produce only temporary improvement. If employees are circumventing a technology platform because the official workflow does not accommodate actual operating conditions, additional compliance monitoring may intensify the workaround rather than eliminate it. Each intervention may be reasonable under one causal explanation and ineffective under another.

A pragmatic systems analysis should help leadership determine which conditions appear most consequential, which are merely visible, which can realistically be changed, and which require further evidence. The aim is not to produce an exhaustive catalogue of everything that could be improved. Organizations have finite attention and implementation capacity. Effective analysis should therefore reduce complexity for decision-makers without falsely simplifying the system itself.

Organizational Learning Requires Testing Assumptions

No organizational diagnosis should be treated as final. Even a rigorous analysis remains a model of reality based on available evidence, and interventions change the system being studied. The appropriate endpoint of analysis is therefore not certainty but a better-supported basis for action accompanied by a method for learning from what happens next.

Argyris and Schön's work on organizational learning is particularly relevant here. Their distinction between single-loop and double-loop learning separates the correction of deviations within existing governing assumptions from inquiry that questions and potentially alters those assumptions themselves (Argyris & Schön, 1978, 1996). An organization experiencing a repeated performance problem may become highly effective at correcting individual occurrences while never examining the structure that continually reproduces them. In such circumstances, greater operational discipline can coexist with limited organizational learning.

For example, if an organization repeatedly misses a performance target, single-loop responses might include additional monitoring, corrective action plans, retraining, or greater managerial oversight. Those interventions may be appropriate. However, if the problem persists, double-loop inquiry asks whether the target, process design, resource assumptions, incentive structure, technology, or underlying theory of performance requires reconsideration. The distinction is not that double-loop learning is always superior; many deviations genuinely require straightforward correction. The value lies in recognizing when repeated correction is evidence that the organization may need to examine the assumptions governing the system.

This is where measurement can become part of organizational learning rather than merely accountability. Establishing an evidence baseline before intervention allows leaders to evaluate whether subsequent changes produced the expected effects, created unintended consequences, or challenged the original causal model. Analysis then becomes iterative: evidence informs an explanation, the explanation informs an intervention, the intervention generates new evidence, and that evidence refines the organization's understanding of itself.

A Systems-Analytic Role for Inimetrica

The intellectual foundations described here are not unique to Inimetrica, nor should they be presented as though they are. Bounded rationality, sensemaking, organizational routines, emergent strategy, organizational culture, systems dynamics, soft systems methodology, systems approaches to error, leverage points, psychological safety, and organizational learning represent decades of scholarship across management, organizational behavior, systems science, safety science, and strategy.

The opportunity for Inimetrica is not to claim authorship of these traditions. It is to integrate their implications into a pragmatic method of organizational analysis.

That method begins by illuminating the problem. Organizational evidence is examined across multiple dimensions so that presenting symptoms can be differentiated from deeper conditions and plausible causal explanations can be compared rather than presumed. Quantitative performance, workforce perceptions, qualitative accounts, organizational structures, formal policies, operational practices, stakeholder evidence, and historical trends can contribute different forms of knowledge. The analytical objective is not simply to aggregate these sources but to understand the relationships among them.

The second task is to refocus leadership attention. Once the evidence has been synthesized, leaders need to understand which conditions appear consequential, where uncertainty remains, what may be symptomatic, and where intervention could plausibly have meaningful leverage. This is fundamentally a prioritization problem. Organizations rarely suffer from a shortage of potential improvement projects; they more often struggle to determine which problems deserve scarce leadership attention.

The third task is to realign the organization where evidence demonstrates meaningful disconnects. Alignment in this context should not be reduced to agreement with a strategic plan. It concerns coherence among organizational purpose, ethos, strategic priorities, resource allocation, structures, routines, incentives, workforce experience, customer or stakeholder outcomes, and the measures by which performance is evaluated. Kaplan and Norton's (2004) work demonstrates the importance of understanding relationships between strategic capabilities, internal processes, and results, while Mintzberg and Waters (1985) remind us that realized strategy may diverge materially from intended strategy. Alignment must therefore be examined empirically rather than assumed from formal documentation.

The fourth task is to intervene pragmatically. Recommendations should be proportional to the strength of the evidence and the organization's actual capacity for change. Not every problem requires transformation. Some require process clarification, some require a change in information flow or governance, some require additional capability, and some require a better measure. In other cases, the most responsible recommendation may be to collect additional evidence before changing anything.

Finally, the organization must learn. Intervention should create new evidence. Leaders should be able to determine whether expected improvements occurred, whether unintended consequences emerged, and whether the assumptions underlying the original diagnosis remain defensible. Organizational analysis is therefore better understood as a continuing learning cycle than as a report that produces a definitive answer.

This five-part synthesis—illuminate, refocus, realign, intervene, and learn—is not intended to replace established theories of organizational behavior or systems science. It represents a practical translation of those traditions into a sequence of questions that leaders can use to examine complex organizational conditions.

Seeing the System More Clearly

The central argument is ultimately modest. Competent leaders can struggle with organizational problems without being poor leaders because complex organizations exceed the observational capacity of any single participant. Leadership provides context, experience, judgment, and responsibility, but those strengths do not eliminate bounded rationality, distributed knowledge, competing interpretations, feedback delays, informal adaptations, or the distinction between intended and realized practice. Indeed, the closer a leader is to an organization, the greater the depth of contextual knowledge available to that leader and, simultaneously, the more difficult it may become to recognize assumptions that have become embedded in everyday organizational life.

Systems analytics cannot eliminate those limitations. Nor should it attempt to replace leadership judgment with an external model. Its more defensible purpose is to improve the informational conditions under which judgment occurs. By bringing different forms of evidence into relationship with one another, examining perceptual and operational discrepancies, testing competing explanations, and tracing connections among purpose, strategy, organizational practice, and results, analysis can increase the resolution with which an organization sees itself.

That distinction also changes the appropriate posture of the analyst. The analyst is not the expert who arrives to explain an organization to people who somehow failed to understand their own work. The organization and its leaders possess indispensable substantive expertise. The analyst contributes another kind of expertise: disciplined inquiry into complex systems and the relationships among evidence that those systems produce.

The most useful result of that relationship is therefore not an externally imposed answer. It is a stronger shared understanding of the organizational conditions that leaders are attempting to influence. Sometimes that analysis will uncover a problem that was previously invisible. Sometimes it will reveal that several apparently unrelated problems share a common structure. Sometimes it will challenge an accepted explanation. Sometimes it will provide evidence for what leadership already suspected but could not yet demonstrate. And sometimes it will show that the available evidence is not sufficient to support a confident conclusion.

Each of those outcomes can be valuable.

For complex organizations, better leadership does not always begin with another initiative. Sometimes it begins with seeing the existing system with enough clarity to understand where leadership attention can matter most.

References

Argyris, C., & Schön, D. A. (1978). Organizational learning: A theory of action perspective. Addison-Wesley. https://search.worldcat.org/title/394956102

Argyris, C., & Schön, D. A. (1996). Organizational learning II: Theory, method, and practice. Addison-Wesley. https://search.worldcat.org/title/35831782

Checkland, P., & Scholes, J. (1990). Soft systems methodology in action. Wiley. https://search.worldcat.org/title/21230229

Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350–383. https://doi.org/10.2307/2666999

Feldman, M. S., & Pentland, B. T. (2003). Reconceptualizing organizational routines as a source of flexibility and change. Administrative Science Quarterly, 48(1), 94–118. https://doi.org/10.2307/3556620

Kaplan, R. S., & Norton, D. P. (2004). Strategy maps: Converting intangible assets into tangible outcomes. Harvard Business School Press. https://www.library.hbs.edu/working-knowledge/mapping-your-corporate-strategy

Leveson, N. G. (2012). Engineering a safer world: Systems thinking applied to safety. MIT Press. https://doi.org/10.7551/mitpress/8179.001.0001

Meadows, D. H. (1999). Leverage points: Places to intervene in a system. Sustainability Institute. https://donellameadows.org/archives/leverage-points-places-to-intervene-in-a-system/

Mintzberg, H., & Waters, J. A. (1985). Of strategies, deliberate and emergent. Strategic Management Journal, 6(3), 257–272. https://doi.org/10.1002/smj.4250060306

Reason, J. (2000). Human error: Models and management. BMJ, 320(7237), 768–770. https://doi.org/10.1136/bmj.320.7237.768

Schein, E. H. (2010). Organizational culture and leadership (4th ed.). Jossey-Bass. https://www.wiley.com/en-us/Organizational+Culture+and+Leadership%2C+4th+Edition-p-9780470185865

Simon, H. A. (1997). Administrative behavior: A study of decision-making processes in administrative organizations (4th ed.). Free Press. https://www.simonandschuster.com/books/Administrative-Behavior-4th-Edition/Herbert-A-Simon/9780684835822

Sterman, J. D. (2000). Business dynamics: Systems thinking and modeling for a complex world. Irwin/McGraw-Hill. https://mitmgmtfaculty.mit.edu/jsterman/business-dynamics/

Weick, K. E. (1995). Sensemaking in organizations. SAGE Publications. https://us.sagepub.com/en-us/nam/sensemaking-in-organizations/book4988