Organizational Transformation
A few thoughts on digital and agile organizational transformations
I’ll describe my ideal enterprise organizational model as a healthy cross-pressure between two elements. On one side, a clear chain of command with the ability to execute, and on the other, an ambiguous and evolving “Spotify” model. These two organizational concepts are not mutually exclusive, and I provide my reasoning and resources that led me to my conclusions.
Motivation
Many companies embrace Agile transformation journeys, and change agents quickly realize that focusing on software development or delivery methods is a minor source of impact compared to organizational design.
I am genuinely curious about the topic, follow public resources about companies that inspire me, and use my writing to reflect on lived experience. As a Designer/PM who is enthusiastic about complex systems, I evolved my professional journey from creating buttons and application specs to systems and platforms. I confirm that everything and everyone working inside of a large enterprise depends on its organizational design. The way how we staff teams, with what roles, how the roles are incentivized to collaborate [1], and what organizational clarity we provide to employees is design. It affects the outcome beyond what an individually contributing designer, product manager, or software engineer can influence. This organizational design is executed by a relatively small group of people, aka leadership or top management. The leadership gets advice from various change agents and coaching roles, and my observations have a similar advising aspiration.
Hierarchies
Hierarchy, of course, is not just an organizational construct. It is a phenomenon intrinsic to the complexity of the natural world. Indeed, all biological organisms are made up of systems — circulatory, skeletal, and respiratory — which themselves comprise many subsystems. Our mental processes are also often hierarchical, especially when we perform complicated tasks.
[3] — Why hierarchies thrive
I tend to think about myself as progressive, but my worldview is challenged whenever agile enthusiasts speak about hierarchies, implying the exact same definition is used in politics, agile software development, biology and just about everywhere else. My immediate association with organizational hierarchy is information and competence, not a power or corruption. Yet, it is increasingly common to consider organizational hierarchy as a sign of legacy, corruption, and inefficiency [4]. With Agile trends, we aspire to find leaner and flatter organizational models, aiming to bring more individual independence and the ability to respond to a change coming from outside constraints.
If a problem is solved and many people involve themselves in the solution, then hierarchy must and will arise. If the problem is real, then the people who are best at solving the problem at hand should arise on the top — that is not power, that is the authority that propels accompanied ability. In fact, genuine authority constrains the arbitrary exercise of power, especially with full transparency about individual responsibilities.
The corruption of hierarchy and power is often strongly objected to by organizational transformation agents as negative, and it is true it might happen. But it is important to distinguish between a functional and productive hierarchy (and people who make it so) and the degenerate shell of legacy institutions, say a Balkan country immigration bureau. Making this distinction requires the capacity and the willingness to observe and differentiate, rather than reliance on generalized beliefs without the effort spent to identify the root cause. In the case of the immigration office, organizing the departments into non-hierarchical squads might not influence the corruption and bureaucratic overhead as it was never the root cause.
Agile and Frameworks
The only constant of digital world is change, and most of the teams are in the middle of digital transformation involving Agile narrative. I learned not to argue about what it means [5], but still, I see codified frameworks such as Scrum, SAFe, or Spotify used as direct blueprints in completely unrelated domains.
As a former civil structural engineer, I take inspiration from different industries or even art. I remain surprised to see how agile enthusiasts are unable to consider abstract patterns instead of step-by-step frameworks. For example, I doubt that the Spotify model, an effective model for the Swedish SaaS music streaming business, can be applied across any other domain such as professional software, 3D modeling libraries, or replacing ITIL standards. I also noticed experts suggesting a narrative of the construction industry or army as a direct example of the opposite spectrum, a legacy, siloed, waterfall and inefficient. But this is not the only inspirational abstract pattern I can synthetize.
In the army, a highly independent team is making pivotal decisions under fire and respond to a change autonomously, often without direct contact, inspection, or approval by the leadership. Indeed, the mission is committed upfront and is non-negotiable, but the execution is nowhere to be a waterfall task factory. Similarly, the building industry, where digital experts might naively perceive a waterfall execution of an exact blueprint turned into a complex building overnight. But this blueprint goes into the ether the first day on-site with completely different weather conditions as anticipated and half of the site workers showing up. We can consider ourselves lucky that software development outcomes do not depend on enemy fire or weather conditions and should use this to our advantage.
If you think about patterns from army, engineering, or industrial design as inappropriate to be applied to knowledge work, I suggest considering differences between workflow and work execution [6].
I appreciate the concept of external advisory. Still, I can’t help myself but think of the Dunning-Kruger effect whenever Spotify or SAFe organizational models are suggested by coaching experts coming from different domains. Every gap or problem is explained away one by one, considered an unpredictable discovery and a positive sign of learning and progress. But these discoveries are predictable, and I wonder if the change agents know the direction of the journey or lack the skills to discover the root cause.
Scrum and Squads
Without a single engineering manager responsible for the engineers on a team, the product manager lacked an equivalent peer. There was no single person accountable for the engineering team’s delivery or who could negotiate prioritization of work at an equivalent level of responsibility.
[8] — Failed Squad goals
The vanishing responsibilities of traditional organizational models, especially in talent and performance management, are often replaced with various models of “Centre of Excellence” aka Tribes, Squads, Chapters, Guilds…and the infinite number of other configurations. While this is an excellent approach to hone the craft, coaching, and connecting the experts across the organization, it is unlikely to scale as a universal organizational model unless we all plan to develop a music streaming services and cap enterprises at 5k employees. What is lacking is a clear “chain of command” or execution hierarchy in a complex system, and I don’t see these two approaches as mutually exclusive concepts — see my final conclusions.
Final Conclusions
Our world is not simple and can’t be simplified into binary categories of good and evil. Our democracy might not be perfect but seen as a constant and healthy competition between conservative and liberal, left and right, well-founded institutions, and progressive agility. I hope to not live in a world of wars when the existence of one only comes with erasing another. I don’t see hierarchies as a negative or evil element of organizational culture because I don’t associate hierarchy with corruption. I struggle when I see artificial ambiguity introduced by agile frameworks that only cover existing hierarchies under new mythology. These hierarchies are not clearly exposed with known balancing mechanisms to avoid corruption of power. Instead, the teams are educated that there is no management, everyone is a leader (or designer), and that system-level challenges can be coached away one-by-one by individual contributors [10].
My ideal enterprise organizational model is drawn on learnings from NASA Challenger disaster story described in Range [11]. I believe a healthy — digital, agile, progressive, modern, and effective organization needs both — an informal, ambiguous “Spotify” organization for creative work with safe space for dissent, test & learn, and craft hyper-specialization that evolves and changes every time it needs to. At the same time, the only way to send a rocket to the moon is a coordinated effort of many individuals within a complex system — a chain of command that relies on clear mission, prioritization, and middle management. A chain of execution with hierarchies builds on authority, expertise where merit is transparent and called out for any signs of corruption.
These two organizational concepts are not mutually exclusive and shall be in healthy tension. One does not come at erase of another.
Resources
[1] — What Is Incentivized Is Policy — All Else Is Lip Service
[2] — What is the Conway’s Law?
[3] — Why hierarchies thrive
[4] — Vanishing middle manager
[5] — Avoid Arguing about Agile
[6] — A world without email
[7] — Dunning-Kruger Effect
[8] — Failed Squad goals
[9] — There Is No Spotify Model for Scaling Agile
[10] — Research on Emotional Dynamics in Radically Democratic Organizations
[11] — Range: Why Generalists Triumph in a Specialized World
