A rich resource on organisational development for complexity
During December 2025, Stefan Norrvall and Richard Claydon collaborated on a series of articles that explored organisational design for organisations working with complexity.
The articles are of particular interest to me as we grapple with how to improve The Chancery Lane Project’s organisational structures in response to the seismic changes underway across law caused by artificial intelligence.
I have been intrigued and impressed by this series of articles, which are deeply relevant for how we do organisational development for organisations working with complex subject matter, and explore the structures that enable, or disable, how work actually gets done.
However, the writing is often dense and inaccessible, filled with complex language like “axiological complexity” and “regulatory grammars” that can be alien to many, myself included. To bridge this gap, I decided to create a NotebookLM, which is a small language model provided by Google, to help people access and work with this thinking, transforming these remarkable insights into a practical interface for design and decision-making.
You can make your own Notebook by uploading links to the 6 essays and then interacting with them using the interface provided. The 6 links are at the bottom of this article. You can also do this in ChatGPT’s modules, if you are a paid subscriber, or any of the other small language model services provided by LLMs.
These resources made accessible
Here is a summary of the six articles in the article series, translated into plain English.
Part 1: The Hierarchy of Work Complexity Is Inescapable (Even at Buurtzorg). This essay challenges the idea that “flat” companies work simply because they removed of managers. It argues that you cannot actually remove complexity from an organisation; you can only choose where to put it. Using the famous Dutch nursing company Buurtzorg as an example, the author explains that the company didn’t delete hierarchy; it moved the hard work of coordination into software, architecture, and small, autonomous teams. The essay concludes that successful self-management requires rigorous design, which requires making sure teams are small enough to solve their own problems without needing supervisors.
Part 2: Where Does the Work Really Live? This article translates the complex theories explained in part 1 into a simple model for executives called Run, Serve, Change:
The author argues that many companies are broken because they force the wrong people to carry these loads—for example, making partners in law firms try to do all three at once. It suggests that to be like Buurtzorg, companies should let teams own the “Run,” use coaches to “Serve,” and have a small, dedicated group handle the “Change”.
Part 3: Organisational Physics, Moral Gravity Author. This essay argues that you cannot just copy a company’s structure (its physics) without understanding its values (its moral gravity). It compares Buurtzorg, which is built on “care” and “solidarity,” with the Chinese giant Haier, which uses a similar structure but is driven by “market competition” and “entrepreneurship”. The author warns that professional services and tech firms often try to copy the “flat” structure of these companies but fail because their internal culture is too focused on individual competition and maximising profit. To succeed, leaders must design systems that fit both their economic reality and their moral goals.
Part 4: Designing a Team for Run, Serve, Change. This is a story about a fictional leader named Maya who tries to apply these theories to her leadership team. Instead of letting her country managers just focus on their local profits (“Run”), she redesigns their jobs so they also look after regional projects (“Serve”) and help redesign the system (“Change”). The essay defines hierarchy not as status, but as a shift in focus: junior roles should focus on delivery, while senior roles should focus on development. It ends with a warning that even if a leader designs a perfect team, the wider organisation might punish them if its performance reviews still rely on old, outdated metrics.
Part 5: Lewin, Rewritten. This piece revisits the work of Kurt Lewin, a founder of change management often famous for the “Unfreeze-Change-Refreeze” model. The author argues that businesses teach this model incorrectly, treating employees like blocks of ice to be melted and reshaped by managers. The real Lewin believed that change happens by helping teams see the “field of forces” they work in. In this view, “unfreezing” isn’t about breaking people’s resistance; it is about helping a team realise that their current way of working is broken so they can fix it themselves. Change is not a one-time project, but a continuous cycle of fixing the work from the inside.
Part 6: The Confusion Tax. The final essay defines the “Confusion Tax” as the price organisations pay for bad design. It argues that most companies are split into functional silos (like Marketing, Finance, and IT), which creates artificial complexity because no single team owns a whole outcome. To fix the disconnect between these silos, companies hire layers of managers to act as “glue,” which slows everything down. The essay suggests reducing this tax by designing “load-coherent” domains—teams that have the full authority and resources to finish their work without constantly needing to ask for permission or forgiveness.
The links you need to set up your own Notebook, or other small language model
Here are links to the 6 articles. Upload them to notebookLM, create a module in ChatGPT, or use another LLM to sit your small language model on, which you can then use to interrogate the body of work.
https://richardclaydon.substack.com/p/where-does-the-work-really-live
https://richardclaydon.substack.com/p/organisational-physics-moral-gravity
https://richardclaydon.substack.com/p/designing-a-team-for-run-serve-change
https://richardclaydon.substack.com/p/lewin-rewritten
https://www.linkedin.com/pulse/confusion-tax-stefan-norrvall-qdfrc/