The Alignment Illusion
Systems Praxis: Why agreement in the room produces disagreement in the work — and what the system is actually producing
The meeting ends with nods. Everyone agrees. The strategy is clear, the priorities set, the path forward understood. Two weeks later, three different interpretations execute simultaneously, and the fourth stakeholder has reverted to what they were doing before the meeting happened. The primer establishes that alignment is not a property of conversation — it is a property of purpose, visible only through what the system produces, never through what its members say they will produce.
The intervention follows from the diagnosis. And most organizations diagnose this wrong.
Agreement lives in the meeting. Alignment lives in the outputs.
Teams can use identical language while holding incompatible underlying assumptions. Research on shared mental models has documented this extensively. “Strategic priority,” “customer focus,” “quality” — each participant confident in shared understanding while their mental models diverge on every operational dimension. The meeting creates surface agreement. The work reveals the divergence underneath.
This is not a failure of communication clarity. Studies in organizational behavior show that even highly specific instructions produce interpretation variance. The more detailed the alignment meeting, the more material available for divergent interpretation. Precision does not solve the problem. Precision gives the problem more surface area.
Coordination research shows the pattern from another angle. Even with genuinely shared goals, individuals optimize locally. A sales team and an engineering team may both want product success. Their local optimization — close deals quickly, build robust architecture — creates friction that is invisible in planning meetings. Each team leaves believing alignment exists because the goal was shared. The friction appears only when outputs collide.
The standard interventions fail for the same structural reason: they optimize for agreement rather than for what agreement was meant to produce.
Weekly syncs, monthly reviews, quarterly planning sessions — each ritual is designed to create agreement. None is designed to produce alignment, because alignment only appears in future outputs that the ritual cannot access. The ritual works perfectly. It produces agreement. Agreement was never the problem.
Research on organizational routines reveals why additional process often deepens dysfunction. Routines become decoupled from their intended purposes. The meeting happens because the meeting is scheduled. Attendance substitutes for engagement. The ritual’s existence becomes evidence that alignment is being addressed — which reduces pressure to notice that alignment is not appearing in outputs.
There is a compounding effect. Each failed alignment attempt erodes trust in alignment processes generally. Participants arrive at the next meeting with reduced investment. They have learned that agreement in the room does not predict coordinated execution. So they agree more easily — why fight for precision that will not survive contact with actual work — and the gap between agreement and alignment widens further.
What gets called an “alignment problem” is usually a system confusing the input for the output.
The default frame treats alignment as an input problem: get the inputs right — clear goals, shared understanding, documented agreements — and aligned outputs follow. This frame is so pervasive it becomes invisible, the water organizational fish swim in without noticing.
But alignment is not an input. It is an output property. It cannot be installed at the front end and expected to persist through execution. It can only be observed — or its absence noticed — in what the system actually produces.
Research on high-reliability organizations reveals a consistent pattern. These systems do not achieve coordination through extensive pre-work consensus. They maintain coordination through continuous attention to what the system is producing — a preoccupation with small deviations before they compound. Studies of improvising teams — jazz ensembles, surgical units, emergency responders — show the same. They do not align through prior agreement. They align through real-time attention to what is being produced and rapid adjustment based on what they observe.
Different stakeholders describe this problem differently and none of them are wrong. Senior leaders see a communication failure. Middle managers see a follow-through failure. Teams see a direction failure. Each description is accurate from inside the frame it comes from. None of them is complete. Every alignment session changes the dynamic around alignment without reaching the structural condition that keeps producing divergence. More effort applied to the current approach deepens what it was meant to resolve.
If this reframe has explanatory power — if it accounts for what the standard approach cannot — there is a way to work with it directly.
The Output Lens Protocol addresses where attention goes — redirecting a system’s gaze from agreements to what it actually produces. (Free)
The Undistorted Information Protocol addresses what happens between observation and sharing — the discipline of offering a system information it can work with rather than conclusions it must accept or resist. (Paid)
Agreement lives in the meeting. Alignment lives in the outputs.
The system will keep producing what the structure makes probable. The question is not how to create better agreement. The question is what the system is actually producing — and whether anyone is looking.
🎧 Thank you for reading.
© 2025 Victor Nuñez / LABKOM Co. Ltd. Thailand. All rights reserved.
Further Reading
Cannon-Bowers, J. A., & Salas, E. “Reflections on Shared Cognition.” Journal of Organizational Behavior, 2001.
Weick, K. E., & Sutcliffe, K. M. Managing the Unexpected. 3rd ed., Wiley, 2015.
Feldman, M. S., & Pentland, B. T. “Reconceptualizing Organizational Routines as a Source of Flexibility and Change.” Administrative Science Quarterly, 2003.