3 Organizational Models To Decentralize Your Company - Radically

Written by
- 11 min read

For a few years I argued that most pioneering firms on our Bucket list move beyond traditional multi-layer hierarchies via organizational models focused primarily on principles of communal sharing or market pricing. But a new round of interviews suggests they use a third model to organize their radically decentralized workforces: namely, a focus on the principle of reciprocity.

This work builds on my previous PhD work on large firms that organize without a formal middle management layer. I dubbed these "middle managerless organizations" or MMLOs for short. (I shared this two years ago in long reads here and here).

Radically decentralized firms

In earlier pieces I showed MMLOs are moving away from the multi-layered hierarchies that are still the norm in many firms today.

Instead, MMLOs radically decentralize the workforce into 'loose hierarchies' of just two layers, using alternative models based on small entrepreneurial units staffed by highly autonomous peers.

article photo

However, it’s important to note that MMLOs are not completely without hierarchies or bosses. They still have a top management team (CEO, founder, owner...) that enjoys ultimate decision-making authority.

That is, MMLOs still include top-management layers and formal structures. But they aim to delegate to the lowest levels of the firm.

Highly autonomous employees make most day-to-day decisions without going through a chain of command.

Of course, this is not new. Scholars have described these variously as:

  • 'Non-hierarchical organizations' (Herbst in 1976),
  • 'Loose hierarchies' (Malone in 2004),
  • 'Boss-less organizations' (Puranam et al. in 2014),
  • 'Self-managing organizations' (Lee & Edmondson in 2015).

What is new, is the adoption of information technology to scale MMLOs massively - up to tens of thousands of peers.

A new model

Where my earlier PhD research was mainly focused on understanding the organizational structures of these firms, my latest research is now focused on understanding the individual behaviours of, and the nature of relationships between, members of these firms.

My earlier PhD research was focused on understanding the structures of these firms. My latest research focuses on understanding the individual behaviours of, and relationships between, members of these firms.

Which is why I interviewed people from various MMLOs: i.e. Haier, Buurtzorg, Viisi, Centigo, ner Group, Encora, VkusVill. I wanted to better understand the behaviours and relationships in these organizations.

This last round of interviews influenced me to iterate to a new model (which I propose later in this post). It is still based on the old model I proposed one year ago.

But the new one tries to resolve an uncomfortable tension I felt in the old model, which was focused on making a distinction between collaboration (communities) and competition (markets).

This was because this simple dichotomy (community vs. market) could not, for example, comfortably place firms run by sociocratic (or holacratic) principles in the old model.

In this new model, I finally can.

Important concepts

The new model is not only based on using MMLO structures to organize people in radically decentralized forms. It is also highly influenced by the work of Thomas Malone (MIT) and Alan Fiske (UCLA).

It draws on two important concepts:

  • Malone's concept of hyperconnectivity, as a way to connect people via IT on unprecedented scales.
  • Fiske's concept of relational models, as a way to cooperate and to coordinate social interactions.


In his latest book, 'Superminds', MIT Professor Thomas Malone shows how computers offer hyperconnectivity. They can connect people at massive scale, and in rich ways not previously possible.

Malone shows that by using new information technologies, like the Internet, the costs of communication and coordination can be drastically reduced. In turn, this dramatically increases the size of groups that can work and make decisions together, and access their diverse backgrounds, skills, and capabilities.

Moreover, Malone argues the trend of declining communication cost will increasingly motivate firms to decentralize and allow members to act autonomously. Why? Because cheap access to information allows more people to make sensible decisions for themselves, their team and their firm.

These advances, and especially hyperconnectivity, put the MMLOs studied in a situation where a radically decentralized way of organizing work was not only possible, it was also desirable.

In fact, hyperconnectivity enables MMLOs to radically decentralize without sacrificing economies of scale. It affected the firms studied in at least two ways:

  • First, IT enabled an increase in the size of the firm, because of cheaper communication between thousands of autonomous members.
  • Second, IT enabled decreased costs of decision making, because it provided everyone in the firm with enough information to make sensible decisions for themselves, instead of waiting for orders from a superior.

As a result, some of these firms (i.e. Buurtzorg, Haier & VkusVill) were able to scale their ‘loose hierarchies’ to very large sizes: indeed, up to tens of thousands of employees.

They did not, however, only rely on new IT tools to enable decentralization on massive scales, they also relied on different relational models to guide the behaviours of firm members.

Relational models

Anthropologist and UCLA Professor Alan Fiske pioneered the concept of relational models with his Relational Models Theory (RMT). According to RMT, people are naturally social, using relational models to structure and understand social interactions.

article photo

According to RMT, all relational models can be analyzed via four elementary models, which are intuitively simple and commonsense:

  • Authority ranking (arrange into a hierarchy),
  • Communal sharing (have something in common),
  • Market pricing (use of ratios),
  • Equality matching (maintain egalitarian relationships).

RMT argues people use these four models to organize much of their everyday social cognition, and as such also interactions in firms.

To understand which relational models MMLOs use to organize these social interactions, we first need to understand how multi-layered hierarchies rely on authority ranking as the dominant relational model.

Model 1. Hierarchy (Authority Ranking)

In the 'Hierarchy Model' relationships are guided by 'authority ranking'. This implies hierarchies in which individuals or groups are placed in relatively higher or lower relationships.

Hierarchies not only specify how many ranks there are, but also who belongs to which rank—because people higher in the hierarchy typically have some control over those who are lower.

Moreover, members ranked higher enjoy prestige and privileges not shared by those who are lower. Often, they also have a duty of protection and care for those beneath them.

At the risk of oversimplifying, I would argue that many traditional firms rely on 'authority ranking' as their dominant relational model for social interactions between members. Simply, this is because they rely mostly on multi-layered hierarchies to organize work between members.

In this view, members primarily use their respective positions in the hierarchy (as superior or subordinate) to frame relationships and guide behaviours with others in the firm.

In multi-layered hierarchies people with authority make decisions and others are required to follow. Often, this entails a strict bureaucracy (i.e formalized job descriptions) specifying the decisions people at different levels can make.

Alternative Models

Based on my research on MMLOs, and influenced by the concepts of hyperconnectivity and relational models, I identified 3 alternative models to the traditional multi-layered hierarchy. That is, how large firms can scale as ‘loose hierarchies’ using new IT tools and focus on meaningful relationships.

article photo

These three alternatives differ from the status-quo of multi-layered hierarchies in the way peers relate to each other. In these, peers can only persuade others, but never enforce their will on others.

As a result, relationships are based on principles of ‘community sharing’, ‘market pricing’, and ‘equality matching’. Let’s explore each model in turn.

Model 2. Community (Communal Sharing)

In the 'Community Model' relationships are guided by 'communal sharing'. This implies a firm in which the behaviour of members is based on belonging to the same community.

In this model, peers share a strong common identity that acts as a kind of social contract. Peers are not differentiated from each other and are expected to help others in the community whenever they can.

Decisions are made mostly via informal consensus or shared norms. Thus, performance of the firm emerges from interactions and relationships between autonomous peers bound by a community of shared values.

Communal sharing & hyperconnectivity

In firms with this model IT is mostly used to reduce the cost of locating, and communicating with, other members of the community.

IT makes it easier to assemble large groups, and easier to involve others throughout the firm to help in solving specific problems. Doing so may also generate better solutions.

Examples of communal sharing

The 'Community Model' can, for example, be found at online communities like Wikipedia and other open source projects. The most obvious example of a MMLO operating with communal sharing as the dominating relational model is the Dutch Buurtzorg.

Learn more about how Buurtzorg works in this animation:

Model 3. Market (Market Pricing)

In the 'Market Model' relationships are guided by 'pricing'. This implies a firm where behaviour of its members is based on cost-benefit analysis.

In this model peers are rewarded for contributions, and reward others for the use of their contributions. This includes suffering losses when one's calculations prove to be incorrect.

Peers enter freely into transactions with each other, with decisions made mostly by agreement among partners—buyers and sellers.

As such, performance of the firm emerges from transactions and relationships between autonomous peers buying and selling on an internal market.

Market pricing & hyperconnectivity

In firms operating with this model, IT is mostly used to create internal markets and bidding webs where peers can compete with each other for resources.

IT allows peers to choose whether or not they want to work on a task, by allowing all members to share information with, and potentially transact with, other members on the electronic internal market.

Moreover, IT allows MMLOs to manage the complex information exchanges required to operate true internal markets within firms. Mostly, these exchanges are formalized in actual (electronic) contracts.

Examples of market pricing

This 'Market Model' can, for example, be found at online platforms like Airbnb, and similar.

The most obvious examples of MMLOs operating with market pricing as the dominating relationship model are the Swedish Centigo, and the Chinese Haier.

Learn more about how Haier works in this animation:

Model 4. Sociocracy (Equality Matching)

In the 'Sociocracy Model' relationships are guided by 'equality matching'. This implies a firm in which behaviour of members is based on balanced reciprocity.

In this model, peers attempt to achieve and sustain a fair and equal balance of contributions. The status of peers increases when they make contributions that others value.

When there is no balance, peers try to track the imbalance in order to calculate how much correction is needed, with decisions mostly made by consensus or consent.

As such the performance of the firm emerges from transactions and relationships between autonomous peers who are trying to ensure a balance in reciprocity.

Equality matching & hyperconnectivity

In firms with the sociocratic model, IT is mostly used to make it easier for its members to self-select a portfolio of tasks that best fits their skills and capabilities.

IT makes it easier for groups to select the members who are best for a task from the entire firm, reducing the transaction costs of finding, selecting and working with peers on a similar task.

Moreover, IT allows MMLOs to create electronic reputation systems to establish credibility among members.

Examples of equality matching

This 'Sociocracy Model' can, for example, be found at online platforms like CouchSurfing, and other peer-to-peer sharing platforms.

The most obvious examples of firms operating with equality matching as the dominating model are the Russian VkusVill, and firms operating with holacratic and sociocratic ways of working, like the Dutch Viisi.

Learn more about how Viisi works in this animation:

No pure models

I do not argue that the four organization models just described are pure forms. And it would be too simple to claim there are only four kinds of organizations. In fact, the MMLOs I studied used a mix of all different relational models with one being dominate.

For example, beside using either the 'communal sharing', 'market pricing' or 'equality matching' as the prime relational model in the firm, all MMLOs also use a relationship model based on 'authority ranking' between top-management and the other members of the firm.

This simple example shows how one model can be embedded into, or substituted for another to create a hybrid. In principle, there is no limit to such embedding, allowing an infinite number of hybrid models.

article photo

Next steps

In a future post I will describe how these three alternative organization models impact and influence the relationships between different members of the MMLOs.

I will also try to explain why some people fit well into these firms, while others struggle...

Stay tuned.

Learn more about Buurtzorg, Haier & Viisi in the online courses we've created with them. Visit the Corporate Rebels Academy to learn more.

Subscribe to our newsletter