Decentralized management in practice: The Edgeryders experience and what it means for social enterprises in general



Edgeryders started out as an online community. Collaborating on the web led us naturally to decentralization. People were holding many conversations at the same time. Some of these conversations led to action. The rule that emerged was this: whoever wanted action was responsible for taking the first step and leading. We ended up drawing legitimacy from action, and that made us a do-ocracy. A mantra started to circulate:

Who does the work calls the shots.

It turned out that organisations wanted to work with us. The only way to do this was for us to build a company on top of the online community, that could act as interface. So we did.

We thought we could it without changing our way of working. In fact, we very badly did not want to change it. This is how we did it: someone from our community had experience in business and finance. We asked them to be the “money guy”. They would keep track of legals and finance. We let them choose the jurisdiction they were most comfortable with (they chose the UK) and asked them to incorporate. We saw the legalities and procedures involved with running a company as boxes to tick, and we had just ticked them. Then we went back to work.


Companies were never meant to be do-ocracies. They evolved out of pacts between traders, who joined forces to finance and organize a caravan to some distant land. These pacts were meant to protect the interests of those entering them. This means that they need to have a conservative bias. Who can do what is staked out with great care. If you are not explicitly empowered to do or know something, you are, by default, forbidden to.

Services to companies are built around that bias. Accountants and banks will only talk to credentialed representatives of the company. Access to company data is out of the question for everybody else. Result: anyone wants to do anything around a company, they have to ask the few people (or even the only person) who are credentialed to take action.

This leaves most companies with a “centralization by default” position. Edgeryders was no exception. All our corporate services (Companies House, the accountant, the bank) were going through the same one “money guy”. The only way to have decentralised management in spite of all this was for this person to accept its whole burden. They would have needed to provide timely, transparent updates on what was going on. They did not do it, and in fairness it was not a nice job to have.

This led to a lack of transparency and the consequent inability of an information-starved board to make decisions. Frustration translated into conflict. It got quite ugly. We paid a heavy price, and it serves us right for not paying attention. The “money guy” refused to hand over company records and was finally expelled from Edgeryders LBG. Some money was unaccounted for. We suffered the humiliation of reporting our missing records to the tax authority and the police. We had to kick the person out of the Edgeryders company. We then needed to rebuild them starting from banking data – excruciatingly boring work.

We also needed a better solution to run things. A year later, I think we have one, made up of several components. My heart goes out to the many people out there who helped us pull through such difficult times, and pointed us towards some of those components ( @ivan was especially helpful). In the rest of this post I introduce and discuss them.


The following principles are “portable” to almost any organisation wishing to run in a decentralized way.

  1. Use the cloud. Internet access to business data and services enables decentralized work practices. Everyone can query them for the information she needs, whenever she needs it. Everyone gets the same picture, so there is less potential for misunderstandings. We use online banking, accounting, social media management, CRM, and of course our own platform to discuss things. Everything we can put online is online.
  2. Keep it simple. Solutions are worth nothing if people do not use them properly. We tend to go for barebones ones, simple enough for us to stick with them.
  3. Share responsibility. If you assign a key function to any one person, you break decentralization, because that person can no longer be routed around. We rotate roles, so that people gain the ability to stand in for each other. This is especially true of anything related to money. If you are an entrepreneur, you have to take responsibility for the business side of what you do. There is no "money guy". You and your partners are it.
  4. Don't cut corners. Regulation on companies is there for a reason. Resist the temptation to treat it as a series of boxes to tick. Doing things right has huge advantages for transparency and accountability. We do this by keeping wikis on management practices: how to account for operations, how to authorize payments etc. In a sense, wikis and documentation replace management: they are algorithms, and anyone who enacts them is managing.
  5. Write. The written word functions as a record of your activities. Get into the habit of making short written updates on what you are doing, and share them in the company's workspace. This buys you autonomy: if you are transparent about what you are doing, you take away the need for other to monitor you in case you are up to something potentially damaging. Also, writing allows your colleagues to choose their own level of involvement in what you are doing. This is important, because it pushes out to them the responsibility for intervening if they don't like it. You write "I'm going to do X". No one objects, so you do it. If then someone comes back complaining after the fact, you can point to your post and say "hey, this has been here for a month. You have had your chance to intervene." At the same time, of course, you must be honest and not drown your colleagues in a sea of information.

Specific solutions

We adopted the following solutions. No pretense of portability here; some are only available to UK companies, others are home made and exclusive to Edgeryders.

  1. E-government: Companies House. The UK's registrar of companies allows you to manage your company's records entirely online. Do yourself a favor and keep those records up-to-date.
  2. Accounting: FreeAgent. A sophisticated accounting cloud service (browser-based). You import your bank's transactions (via APIs), then do all your accounting on it. You can (and should) create many users, with different levels of access. I like the social implication of using it: it implies you do your own accounting. Someone must be able to explain each and every movement of money. You end up knowing your business well! And, with this low-level work done, you can generate sophisticated reports, do ratio analysis and so on. Only 20 years ago you needed to be a bank to have this kind of stuff! The only down side is that there is no way to give read-only access to people. If you can see something, you can change it.
  3. Accountant: Nimble Jack Accounting. After the "money guy" fiasco, we signed up with a London company called Nimble Jack. They have a brilliant business model: they live in symbiosis with FreeAgent. All their clients get an account in the basic package. This makes their work extra efficient, because the clients do all the low-level work of entering bills, invoices, and banking data. If you decide you want to use NJA too, please tell them Edgeryders sent you (we get a discount).
  4. Banking. This is an unsolved problem. If you are a small, global company doing most of your business overseas (and in EUR or USD), UK banking frankly sucks. It is expensive and the level of service is dismal. We are functional, just about. If you know of a half-decent bank in the UK, or, even better, a bank anywhere that offers competitive banking in EUR to a British company, please tell us about it! The only silver lining; most have APIs that FreeAgent can link to.
  5. Super simple CRM: We tried a Drupal-based CRM called RedHen, but it did not work for us. Way overkill. So, Matthias built a simple content type called Journal, with variables that can be updated via comment. The principle is the same we use for Tasks (think Issues in GitHub). Each Edgeryders project becomes a journal. Initially they are in the idea stage. As they progress, we leave comments saying "such and such happened today" and change their status to "negotiating" or "live" or "finished" or whatever. Take a journal and go through its comments: as you do, will also follow the related project through its cycle. If you want this for your own project on Edgeryders, you can have it.
  6. Forum: again. A specific, invitation-only project on is the conversation hub for Edgeryders-the-company. Yes, we discuss business mainly in writing. This is great for decentralization, because two people can start exploring a topic. If they need to involve others, that's easy: their conversation thread acts as a record and a recap for the new people.

At the end of the day, these are just standard pieces. They are not the reason why Edgeryders is a decentralized environment. We are decentralized because we choose to be, and we are ready to pay the price. We pay it gladly, because such are our values, and because it enables parallel processing and, as a result, far higher productivity. And we need any edge we can get to survive as a social enterprise.

But we need to stick with that choice, and renew it often. Organisations seem to flow towards centralisation like water flows downhill. Every time you don’t make the effort to share a piece of information, you are hoarding knowledge and power. Perhaps it’s true, as Loki once said, that hierarchy is our natural state.

What about you and your organisation? What’s your experience? Is there anything we can learn from you?

Photo credit: Smitty


Mentoring? Similarities and differences

Hey, I have started to dabble a little with the folks at Local Motors. My impression is that there are many similarities. Perhaps this would be a good address to put the mentoring bit for you and Noemi on the table again?

This might actually cut both ways - as I think they could really use a guy like you in the coming months. A very straight shooter who knows about networks, soft touches, complexity, and radical thinking. In my opinion it becomes fairly obvious in e.g. Yoitisi’s comments here.

If I’d win the competition I with the help of you guys I could perhaps branch some monies back to you - provided Nimble Jack can book that. Or we have to trade in kind - see above. :slight_smile:


Local Motors community

@trythis thanks for the references, interesting reads, now i haven’t dug a LOT in the way their company is designed, but are you thinking of applying to one of their competitions/ challenges creatively or are you approaching them to offer a service for their company?

PS very interesting to read threads in an up front technical product dev community :smiley:


Ideally both things together

Obviously, there is one going on in which I am involved and I have started bugging Matt and Natalia to see if they would be game (mostly because their Nepal experience makes some sense in this connection). But if you are interested I’ll keep you posted as well of course.

And honestly I don’t claim to know much about the internals at the company - but the “behind the scenes” stuff looks authentic and the vibe is pretty good (although cars really aren’t my cup of tea). They also perceive themselves as sort of concerned with bigger-picture (i.e. mobility in general, and co-creation).

In the mid to longer term this would potentially segway into my world domination plans a la MUSE. Airbus, you guys for network/social angle, LM more for the design, and physical hardware angle. But that is a perspective for step 3-4, not now.


Or let’s say

We get to know each other in this first one, and see if we would be a good match…

One the other hand there is no reason why we should not participate in one of their challenges - or let them host one for OpenCare.


Some critical reflections

Going by the invitation in the end, my two cents. As someone who participated in about half a dozen (failed) decentralization attempts, perhaps my experience warrants a more critical reflection or can serve as a word of warning. So, here goes.

  1. Why do you wish to decentralize? No, seriously. I know it's somewhat of an organisational fashion, but unless you really know why you are doing it, how could you do it properly? The business with the "finance guy" suggests that it's more about wanting to offload responsibilities to other people, when you don't like them, than to seriously tap the potential of a group. The do-ocracy and ad-hocracy smells the same to me and it is something that I saw again and again: wanting other people to handle "boring" stuff, but deeply controlling what really matters personally. Untill it goes wrong. I'd also suggest to at least use a first name instead of "finance guy", this is needlessly dehumanizing.
  2. What is the exact problem with "hierarchy"? Sure, associating it with a fictional bad guy is funny and all, but it is not exactly an argument. Hierarchy can indeed imply authority (delegated or involuntary), but it can also imply leadership, responsibility, etc. An organisation is not only about "who does what", but also about "who decides what", "who takes responsibility for what", "who is responsible for assigning such responsibility and/or decision making rights", etc etc. My own experience tells me that organisations that try to avoid explicit leadership and try to distribute responsability typically fail as they grind to a halt as soon as there is any kind of truly profound difference of opinion. Again, your story about essentially ex-communicating a key player in Edgeryders, shows clearly who the true "bosses" are. I highly doubt that this decision was taken in full consensus with the community, so it's a perfect example of "do what I say, but not what I do". The resolution to this essential dissonance is - in my experience - quite simple: just own up to the leadership and assume responsibility. After that, it's possible to give people more or less room to maneouver, but at least it is honest about what kind of situation they commit to.
  3. What is the price you think you will pay by having a (soft, non-dictatorial) hierarchy? (Which I think you implicitly have by the way.) I know there are numerous pitfalls in hierarchical organisations (which is the overwhelmingly majority of all organisations), but have you considered mitigating those without trying to reinvent the way humans have organised groups for ages? Are there pitfalls (a "price") that you believe you would be unable to avoid? Are there examples of organisations that are hierarchical to a degree and yet you still admire?
  4. Do you feel that people are still confused about how Edgeryders works? Why would that be? Is it possible that this has to do with the fact that the presentation doesn't match the execution? I know it is easy to blame the outside world for "not understanding the way you work", but you typically attract very smart people, so what gives?
  5. What are you trying to achieve with your transparency efforts? Is this the only pre-requisite for decentralization? Does it have inherent benefits that would work in other styles of organisations? Again, I think it is important to realise why you are doing it. If it is only to enforce decentralisation, then the message you are sending in your organisation is that information and communication is indeed still the best form of power in Edgeryders. (I've actually witnessed some backchanneling that directly supports this thesis, so quite a worrying element.) Transparency can be a way to "de-power" people, but only if information becomes trivial to come by. As long as there are ways to hide information, people can and will - to my personal horrifying experience - abuse this gap all the way to the top. So, how to obtain and maintain a true culture of transparency? I don't think API's can do that.



Hey Thomas, that’s a pretty grim view. Not that I don’t see your point – it’s close to the one made by Loki in the video! I am going to pick your questions 1 and 4.

As for question  1: as I wrote, we started out as decentralized. That’s easy, when you are not a company. All of these contortions were meant to preserve the way of working that was already there, and (we think) is efficient and comfortable. A lot of it was trying to avoid disruption. For me personally there was also an unsatisfactory personal experience working in large buraucracies.

As for question 4: no, people do not understand how Edgeryders works. I don’t blame them. Hell, I don’t fully understand how Edgeryders works. We are making it up as we go, trying to keep the principles and the practices that we like and make things possible for a very marginal, zero-debt organisation. And of course presentation does not match execution; execution is full of holes. But in fairness, this is also true of formal hierarchies, which should have full accountability and control, but are in fact in SNAFU mode more often than not.