Organisational Dysfunction of the Day

The customer we never met

Context: The team is building a product. They have a backlog, a product owner, and a roadmap. A user story ap describe what users need. Personas represent who those users are. Analytics track what people do. Every few months, there is a user research report from the UX team. The team works hard, ships regularly, and hits their sprint goals. They have never spoken directly to a customer, though. The product owner does that, and the UX researcher. But the engineers, the people making hundreds of small decisions every day that shape the product, have not. They are building for an abstraction. A persona on a wall, a ticket in Jira, a data point in a dashboard.

OST explains: An open system maintains its health by actively engaging with its environment. For a product team, the primary environment is the people using what they build. The DP1 bureaucracy is a closed system and mediates that relationship through roles: the product owner translates customer needs into requirements, the UX researcher translates behaviour into insights, and the engineer receives the output of both translations. Each translation loses something. The judgment, the context, the friction, the moment when a real person says something that changes how you understand the problem entirely. In DP2, the group owns the whole task, which includes understanding who it is for. Teams with direct customer access make qualitatively different decisions than teams working from second-hand accounts. Not because engineers are better researchers, but because unmediated contact with the environment is not a nice-to-have. For an open system, it is the condition for staying alive.

#OpenSystemsTheory #SocioTechnical #OrgDesign #agile

Organisational Dysfunction of the Day

The collaboration that isn't

Context: Two teams decide to work together. A shared initiative, a joint project, a problem neither can solve alone. There is enthusiasm on both sides, and a genuine excitement about what they might build together. Everyone has a picture in their head of what that looks like and what they want to get out of it, and everyone assumes the others have the same one. Who owns what, and what success looks like for each of them, is left open. Good intentions and regular syncs will be enough. For a while, they are. Then something goes wrong. Credit lands unevenly. Decisions get made by the louder team. One side feels their contribution has been absorbed rather than shared. Trust erodes quietly. The collaboration continues in name but not in spirit.

OST explains: Two teams are two social systems, each with its own purpose, its own design principle, its own internal logic. The shared purpose is assumed rather than agreed upon. There are two clean ways to work across that boundary. A clear transactional relationship: a contract, explicit deliverables, arm's length. Or deliberately create a new, shared system with a common purpose both sides have actually agreed to, a structure for how the collaboration works and is coordinated, and a design principle governing the joint work. What tends to happen instead is neither. The teams proceed as if goodwill substitutes for structure. The result is laissez-faire more often than not. No clear location of responsibility, no purpose anyone has committed to, no shared system. DP1 (bureaucracy) fills the vacuum, as it always does: the team with more power, more visibility, or a stronger brand quietly starts setting the terms. The other finds itself inside someone else's system without ever agreeing to join it. The same dynamic plays out between companies, just with invoices adding a harder edge to the same underlying confusion. The collaboration was real. The shared system never was.

#OpenSystemsTheory #SocioTechnical #OrgDesign #systemsThinking

Organisational Dysfunction of the Day

Working alone together

Context: The pandemic forced a global experiment nobody had planned. Offices emptied overnight, and something unexpected happened. People managed. Many thrived. The commute nobody missed, the open-plan office nobody pined for, the meetings that turned out to work fine as calls. When offices reopened, a significant number of people simply did not want to come back, at least not every day. The debate that followed framed this as a question of productivity, collaboration, or work-life balance. It missed the more uncomfortable signal. When given the choice, a large number of people revealed they prefer to work alone. That is not a statement about remote work. That is a statement about what going to the office actually felt like.

OST explains: In DP1, the bureaucratic structure, work is individualised by design. Tasks are broken into minimal parts, each assigned to one person, with coordination handled above. The office under DP1 is not a place of genuine collaboration; it is a place of proximity without connection. People sit near each other, manage their own tasks, protect their own position, and attend meetings where decisions have usually already been made. The six psychological requirements for productive work include mutual support and respect, and continual learning through real feedback from people you work with as peers. DP1 delivers neither. Of course, people prefer to stay home. Home is quieter, the coffee is better, and nobody is watching. In DP2, the group is the unit of work. Coordination, learning, and support all happen in the group, not above it. People in genuine DP2 structures report wanting to be with their colleagues because the group is where the work actually lives. The WFH debate is really a DP1 debate. The office stopped being worth the commute long before the pandemic. The pandemic just made it impossible to pretend otherwise.

#OpenSystemsTheory #SocioTechnical #OrgDesign #teams

Organisational Dysfunction of the Day

The external verdict

Context: The organisation wants an independent assessment of the codebase. An external firm is brought in, given access to the repositories, and asked to produce a report. Sometimes the teams are interviewed individually beforehand. Sometimes they are not involved at all until the findings land. Either way, the result is a document that describes their work from the outside, written by people who were not there when the decisions were made, did not carry the constraints, and will not be there when the changes need to happen. The teams read the report. Some findings are fair. Others miss the context entirely. A few are simply wrong. The usual response is a mix of defensiveness, resignation, and quiet relief that at least some things were not noticed. Nobody feels ownership of the conclusions. The report goes into a folder. Most of it stays there.

OST explains: Bringing in external experts to evaluate a team's work is a bureaucratic DP1 move, even when done with good intentions. Control and judgement sit outside the group, and the group receives a verdict rather than participating in an enquiry. The knowledge needed to assess a codebase well is distributed across the people who built it. An external reviewer can spot patterns and bring a comparative perspective, but cannot access the reasoning, the tradeoffs, or the organisational pressures that shaped every decision. Individual interviews are better than no involvement, but they reinforce exactly the individualisation that DP1 already produces, and in teams with existing tensions, they can surface conflicts that the interview format has no mechanism to resolve. What works is treating the review as a collective sense-making exercise. The team leads the process, names what it already knows is wrong, and uses external input as one source of perspective rather than as the source of judgment. In DP2 (self-managing teams), the group owns the quality of its own work. That ownership cannot be outsourced. An external consultant can be a useful mirror and a driver for change. They should never be the judge.

#OpenSystemsTheory #SocioTechnical #OrgDesign #review

Organisational Dysfunction of the Day

Team Topologies, the wrong way round

Context: Leadership has read Team Topologies. The insights are compelling: stream-aligned teams owning end-to-end delivery, platform teams reducing cognitive load, enabling teams building capability. A reorganisation is planned. Teams are renamed and restructured. The new topology is announced. People find themselves in stream-aligned teams that still wait for approval from the same architects, report to the same managers, and receive priorities from the same product managers who held the backlog before. The platform team is the old infrastructure team with a new name and a mandate to serve internal customers, though nobody agreed on what that means. The enabling team runs workshops that nobody has time to attend. Six months in, the cognitive load has not decreased. Delivery has not improved. Leadership concludes that the teams need to embrace the new model more fully. Another round of communication is planned, and product coaches are hired en masse to fix people.

OST explains: Team Topologies describes structural patterns that emerge from healthy organisations. Like DORA, it is a map of what good looks like, not a recipe for getting there. The patterns only function as intended when the teams operating them are genuinely self-managing, owning their work, coordinating among themselves, and making decisions without constant escalation. Imposing the topology from above while leaving the underlying design principle unchanged is a bureaucratic (DP1) move applied to a self-managing (DP2) framework. Stream-aligned teams become delivery units with a new name. Platform teams become service departments, and their internal customer model quietly recreates the same dependency it was meant to dissolve. Goodhart's law applies here as much as it does to DORA: the moment the topology becomes a target, it stops being a good topology. The book is right. The reorganisation missed the point.

#OpenSystemsTheory #SocioTechnical #OrgDesign #TeamTopologies

Organisational Dysfunction of the Day

Involvement theatre

Context: Some architects, project leads, or managers want to do things properly. They genuinely believe in collaboration and know the teams have valuable knowledge. So they organise workshops, like EventStorming sessions, design sprints, or other types of collaborative design workshops. People are invited, post-its go up, discussions happen, and there is real energy in the room. Then the session ends, the outputs are photographed, the facilitator disappears with the material, and a few weeks later, a design document or architecture proposal lands in the team's inbox. It looks nothing like what people thought they were building together. When questions are raised, the answer is that the workshop inputs were "taken into account." The teams learn quickly that the workshops are not really about designing together. They are about being consulted. Next time, fewer people will engage seriously. The post-its get sparser. The energy in the room is noticeably lower, even hostile.

OST explains: This is one of the most common misapplications of participative techniques in DP1 organisations: design authority is retained above, while the appearance of participation is layered on top to legitimise decisions already made or soon to be made elsewhere. OST explains why it fails on two levels. First, Bion's basic assumptions: the moment a person with authority enters the room, even a well-meaning independent facilitator, people shift into dependency or fight/flight, so the workshop never produces genuine collaborative design, regardless of how it is facilitated. Second, Fred and Merrelyn Emery were explicit that it is only when people design their own work that they develop the motivation, responsibility, and commitment to implement it effectively. A design imposed on a group, even one consulted, will never have the ownership that a design created by the group has. Involvement theatre does not just fail to produce good design; it actively corrodes trust in the collaborative process, making genuine participation harder to achieve each time. As Kurt Lewin warned, people cannot be trained for democracy by autocratic means.

#OpenSystemsTheory #SocioTechnical #OrgDesign #leadership

Organisational Dysfunction of the Day

Tyranny of the majority

Context: The workshop is wrapping up. There are too many ideas on the wall, and a decision needs to be made, so dot voting is introduced. Everyone gets three dots and places them on their favourites. The tallied results are clear, and the facilitator announces the winners. Most people seem satisfied. A few do not. Their priorities were lost, their concerns were not addressed, and the vote moved on before the reasoning behind the minority view was ever really heard. They leave with a decision they had no real part in making. In meetings, the same pattern plays out through a show of hands, anonymous polls, or simply the loudest voices drowning out the quieter ones. It feels democratic. It is not.

OST explains: Dot voting and majority voting are representative democracy mechanisms, which really are DP1 in disguise. They produce a winner and, by definition, a loser, and the minority does not just lose the vote; they lose the ability to influence the outcome, making them unlikely to commit to implementing something they actively disagreed with. Fred Emery called the alternative rationalisation of conflict: instead of forcing a resolution through voting, the group stays with the disagreement long enough to understand it. The goal is not the most popular option, but the option nobody has a reasoned objection to. A higher bar that takes longer, but produces genuine shared ownership that majority voting never can. A decision reached that way does not need to be enforced; people carry it forward because they helped make it.

#OpenSystemsTheory #SocioTechnical #OrgDesign #collaboration

Organisational Dysfunction of the Day

Professional leadership

Context: The organisation has invested heavily in leadership development. There are programmes, frameworks, coaching, 360-degree feedback, and a clear leadership model on the intranet. The managers are well-intentioned, many of them genuinely skilled, and they take their responsibility seriously. However, the teams below them are still not performing as hoped. Engagement is middling, decisions are slow, and the best people keep leaving. Some even because of their manager. Leadership quality is clearly not the bottleneck, so what is? In many industries, supervisors are expected to know the craft they oversee. Not so in most IT organisations, where the manager's job is people and process, not technology. That gap has been growing.

OST explains: The problem is not the leaders; it is the existence of the role itself. DP1 as bureaucracy requires leaders because control and coordination are handled by a layer above the real productive work, be it managers in the line, project managers, product managers, or even architects. You therefore need good ones, and training them makes sense within that logic. But no amount of leadership quality fixes the structural problem that the people doing the work are not in control of it. In DP2, the need for professional leadership largely disappears because coordination and control are handled by the group itself. The resources currently spent on developing leaders should instead be invested in developing the team's self-managing capacity. That is not a small shift; it is a fundamentally different theory of how organisations work. A DNA swap.

#OpenSystemsTheory #SocioTechnical #OrgDesign #leadership