A new region gets added to a report.
It’s a routine change. The underlying data has been connected, the pipelines extended, the logic adjusted to account for regional differences. The model refreshes, the numbers populate, and the report goes out as expected.
Then the feedback starts.
Nothing specific at first. No clear error message, no obvious discrepancy. Just a general sense of discomfort.
“The numbers don’t look right.”
It’s a familiar moment for anyone who has worked with reporting, analytics, or operational systems. The initial response is equally familiar. The team goes back through the process, step by step. Data validation checks confirm that source tables match expected outputs. Pipeline runs are reviewed for failures or anomalies. Transformations are re-examined. Even the report formatting is scrutinized, in case the issue is simply presentation.
Everything checks out. And yet the feedback continues. The numbers may be technically correct, but they are not trusted. And without trust, correctness doesn’t resolve the problem.
At some point, the approach shifts. Instead of continuing to validate the system in isolation, the focus turns outward. A simple, user-facing feedback mechanism is added directly to the report. It doesn’t require a separate email, a meeting, or even a clear articulation of the issue. It just needs to be easier than doing nothing.
Something changes almost immediately.
Feedback becomes specific. Patterns emerge. Questions that were previously vague resolve into identifiable gaps; not errors in the underlying data, but differences in interpretation. Mismatches in expectation. Context that existed in one part of the organization but not another.
The issue was never just accuracy. It was interpretation, communication, and trust.
And once those began to align, the problem that had resisted technical validation began to resolve itself.
Situations like this are easy to dismiss as isolated incidents. A quirk of a new region. A temporary gap in understanding. Something that better documentation will fix.
But they aren’t isolated. They are structural patterns that repeat across organizations.
The challenge isn’t that organizations struggle to produce correct outputs. It’s that what is produced, what is understood, and what is acted upon are shaped by the layers and relationships those outputs pass through. A decision can be clear at the point it is made. A system can be accurate at the point it is built. But as both move through the organization, across functions, priorities, and perspectives, they change.
The question isn’t so much why this happens, but how.
Hierarchy: Transformation, Not Authority
We tend to think of hierarchy as a question of authority: who reports to whom, who makes decisions, who is accountable for outcomes. In practice, hierarchy functions less as a chain of command and more as a sequence of transformations.
At the outer edge, closest to reality, are the people doing the work. The front line interacts directly with customers, systems, and operations as they exist… not as they are described. Their focus is immediate and concrete.
As information moves inward, it reaches the first layer of management. The role here is not simply oversight, but translation. Managers take fragmented, situational inputs from the front line and convert them into patterns that can be communicated upward. At the same time, they take direction from above and convert it into actionable work for their teams.
Above them, directors operate at a higher level of abstraction. Their responsibility is to define priorities: which problems matter most, how resources should be allocated, which tradeoffs are acceptable. Their context is comparative rather than situational.
At the executive level, the focus shifts again. Here, decisions are directional. Executives balance organizational goals against external pressures, financial realities, and long-term positioning. The information they receive has already been aggregated, interpreted, and shaped by the layers below.
At each step, something subtle but important happens. The decision, the signal, and the context; none of it passes upward unchanged. It is filtered, compressed, and reframed to match the needs of the layer receiving it. What matters at the front line on a given Tuesday is not what matters at the director level during a business review.
Every layer adds interpretation, and none of them are wrong for doing so. They are operating within the context required of them. The result, however, is that by the time information reaches the point where decisions are made, it no longer looks like it did at the edge. And when those decisions move back outward, they undergo the same process in reverse.
This is not a failure of execution, but a natural consequence of how organizations are structured to process information.
Functions as Languages
Hierarchy explains how information moves through a reporting chain. It doesn’t explain how it changes when it moves sideways.
Organizations are divided into functions, Finance, Legal, IT, HR, Operations, Product, with each responsible for a different aspect of the business. These functions do more than divide work. They divide interpretation.
A revenue initiative, for example, does not remain a revenue initiative as it moves across the organization. To Finance, it becomes a question of cost, return, and variance. To Legal, a question of risk, exposure, and defensibility. To IT, a question of feasibility, scalability, and system constraint. To HR, a question of policy, compliance, and workforce impact.
None of these perspectives are incorrect, and each is necessary, but they are not the same. The original intent is preserved in principle and transformed in meaning. What began as a single idea becomes multiple parallel interpretations, each valid within its own context, each incomplete without the others.
Because these functions operate concurrently rather than sequentially, these interpretations don’t resolve into a single unified view. They must coexist.
The result is not confusion so much as divergence. The same decision looks different depending on where you stand. Once that divergence exists, it doesn’t need to be created again — but it still needs to be managed.
Interactions Define Value
If hierarchy explains movement, and functions explain language, interactions explain outcomes.
Organizations operate at the boundaries between layers and functions, where one group’s output becomes another group’s input, and where each side defines success differently.
Consider Sales and Production. Sales receives demand from the market; its output is revenue. Production receives those commitments as requirements; its output is fulfillment. A deal that is valuable to Sales may be disruptive to Production. A delivery schedule optimal for Production may be unacceptable to the customer Sales has committed to.
Or consider Product and IT. Product translates market demand into feature requirements, measured by value delivered and adoption achieved. IT receives those priorities as requests against a system, measured by stability, scalability, and feasibility. A feature that creates value for Product may introduce complexity for IT. A constraint necessary for system stability may limit what Product can promise.
Or Finance and Operations. Finance receives activity as numbers (transactions, costs, or forecasts) and produces insight measured in predictability. Operations receives work as reality in terms of orders, resources, and constraints, and produces execution measured in throughput. A decision that improves margin on paper may create friction in practice.
In each case, the pattern is the same. Inputs are transformed into outputs. Outputs are measured against a definition of success. And those definitions do not perfectly align across boundaries.
This is not a flaw in the organization. It is the organization. Misalignment is not a failure of communication or competence, but instead a natural consequence of dividing work into functions that each optimize for a different dimension of the same system.
Once you begin to see this, a shift occurs. Problems that appear as breakdowns start to reveal themselves as interactions; relationships to be understood rather than errors to be eliminated.
Bridges Across Boundaries
If interactions reveal where misalignment occurs, cross-functional groups exist because those boundaries cannot resolve themselves.
These groups are often described simply as teams, but that framing understates what they actually do. A team produces outputs within a defined scope. A cross-functional group operates at the edge of multiple scopes simultaneously, receiving inputs shaped by different incentives, translating between different definitions of success, and trying to produce something coherent from material that wasn’t designed to fit together.
Consider an architect. They don’t own a single system; instead, they hold the relationships between systems. When one team’s technical decision creates a constraint for another team’s roadmap, the architect is the person who has to make that visible, and then navigate the competing priorities of both sides without the authority to simply override either. The work is less about designing systems than about managing the friction between them.
Product managers operate in similar territory. They receive market signals from one direction and delivery constraints from another, and their job is to find something that satisfies both well enough to move forward. The tension between what customers want and what the organization can actually build is not a problem to be solved once, but the permanent condition of the role.
Program and project managers live at the boundary of timelines and dependencies. Their work is largely invisible when it succeeds and impossible to miss when it fails. They don’t produce outcomes directly; they create the conditions under which other groups can produce outcomes together.
What these roles share is that they absorb tension the structure would otherwise leave unresolved. They don’t eliminate the underlying divergence; the incentives, constraints, and definitions of success that create friction still exist. What they provide is a place for that friction to be held and worked through, rather than ignored.
Which is also why their effectiveness is always partial. They are operating inside the same system they are trying to bridge.
Incentives, Time, and Power
Up to this point, the system can appear mechanical: layers translating inputs, functions reshaping decisions, and interactions creating predictable tension. But translation is not neutral. It is shaped by incentives, constrained by time, and filtered through uneven access to information.
These forces move through the entire organization, quietly shaping how decisions are interpreted and acted upon, and they interact with something the system depends on but cannot mandate: the willingness of people to say what is actually true.
At every boundary between layers, functions, and teams, the system relies on people to communicate honestly. It needs to know where assumptions break, where constraints will cause friction, where outcomes will differ from expectations. When people don’t feel safe doing that – because the incentive is to soften bad news, or because past experience suggests honesty carries a cost – the system loses its only reliable feedback loop.
Issues are delayed or left unspoken. Signals from the edge fail to reach the center with the fidelity needed to act on them. The organization doesn’t simply become misaligned; it becomes blind to its own drift.
With that in mind, consider the other forces.
Every layer, every function, is not just processing information; it is optimizing for what it is measured against. Sales is driven by revenue. Operations by throughput and cost control. Finance by margin and predictability. The same decision enters each of these contexts and is reshaped, in ways both large and small, to align with how that group is rewarded. A pricing change might represent an opportunity for growth in Sales, a margin risk in Finance, and an operational disruption in Production. Each interpretation is valid, just as each is incomplete. And each pulls the decision in a slightly different direction.
Alignment is difficult to sustain not only because of communication gaps, but because of how success is defined differently across the system.
Time adds another dimension. Not all parts of the organization operate on the same horizon. Sales lives in the immediate: this deal, this quarter. Operations works in the near term: delivery schedules, capacity, stability. Executives stretch further into quarters and years. The board may emphasize financial cycles that favor short-term performance or long-term return depending on the market and the moment.
A decision that is correct in one timeframe can be harmful in another. An aggressive revenue push may strengthen short-term results while creating long-term strain on operations. A cautious operational stance may protect stability while limiting immediate growth. Time becomes a hidden dimension of misalignment; the tension is not just about what is right, but when.
Finally, consider where decisions are actually made. The people closest to reality have the most detailed and accurate information. They see the friction, the exceptions, and the details that don’t survive compression into summaries. Conversely, the people farthest from that reality have the most authority. Information moves inward; authority moves outward. They do not always meet in the same place.
As information travels upward, it is compressed; detail becomes summary, variance becomes pattern, complexity becomes something that can be communicated at scale. That compression is necessary, but it comes at a cost, and it is often the nuance that determines whether a decision will actually work in practice.
Incentives shape what people optimize for. Time shapes what they prioritize. Power shapes where decisions are made. And as decisions move through those forces, they change; not always because they were misunderstood, but because they were processed exactly as the system requires.
The Shape of the System
At a glance, organizations are usually described as pyramids: layers of authority, decisions flowing downward, and information flowing upward. But that model hides more than it explains. What we’ve been describing is not a straight line.
It’s a loop.
At the outer edge is reality: the front line, the systems, the customers, and the day-to-day work as it actually exists. This is where signals originate, where outcomes are felt, and where constraints are real and immediate.
As those signals move inward, they are translated, layer by layer, function by function, into something that can be coordinated, prioritized, and acted upon at scale. Detail is compressed. Local context is generalized. What is messy becomes navigable.
Near the inner ring, decisions are made. Not in direct contact with reality, but in response to it as it has been interpreted, reshaped, and compressed along the way. Those decisions then move back outward, undergoing the same process in reverse. They are translated again at each layer, and adapted again at each function, until they reach the edge where work actually happens.
And at the center sit the constraints that define the entire system: ownership, capital, and governance. The structural limits that shape what is even possible to decide.
Seen this way, the organization is not a pyramid. It is closer to a torus; a continuous loop where information moves inward and decisions move outward, and both are transformed along the way.
This matters because a torus has no privileged point – every position on it is simultaneously downstream of something and upstream of something else. The front line is downstream of executive direction and upstream of the signals that will eventually shape the next decision. The executive is downstream of compressed information and upstream of the direction that will be translated by every layer below. No one in the loop is simply a sender or simply a receiver. Everyone is both, and the quality of what passes through any given point depends on what happened at every point before it.
This is also why intent alone is never sufficient. Organizations do not produce outcomes based on what was meant. They produce outcomes based on what survives the journey; what is preserved, what is reshaped, what is lost, and what is reinforced as decisions move through functions, incentives, and time.
Which reframes the role of nearly everyone inside the system. Most roles are not defined solely by authority. They are defined by the inputs they receive, the outputs they produce, and the quality of the relationships that shape both.
Which brings us back to where we started.
The reporting problem was never really about data accuracy. It was about the interface between people and the system, and the pathways through which signals could move, be interpreted, and be corrected. The feedback mechanism didn’t fix the data. It repaired a pathway.
When those pathways are weak, the system drifts. When they are strong, the system adapts.
Not because the structure has changed. But because what moves through it has.
Abstract Foundations 
Leave a comment