Hacker News Reader

Comment by whack | Hacker News Reader

It's worth keeping in mind that this is a company of 40 people. Half the size of most primitive hunter-gatherer tribes which had no formal organization whatsoever. At sizes that small, everyone knows each other, everyone knows what everyone else is doing, and people can make decisions through consensus after a round of informal discussions.

The problems only start once the "tribe" grows beyond 150. That's when the number of people involved is too great for the human brain to process the entire social network. Hence why formal organisations and hierarchy are now needed.

It's great that startups and small companies are innovating with different organizational styles. Many of the structures found in big corporations are overkill for small organisations where everyone knows each other. Just don't expect any success stories from these startup experiments to scale up to bigger companies though.

View on HN

> The problems only start once the "tribe" grows beyond 150.

The problems also start when there's a potentially extremely large, but uncertain to happen, reward looming over the horizon. In crude terms, when there's a huge money bag that everyone could potentially snap up. Then all sorts of perverse incentives start to manifest.

I'm thinking of stereotypical Wall St. companies, and also of many Silicon Valley "startups" in recent years.

The no-leader scenario is stable when everyone is looking forward to having a regular job as a middle class citizen. It's also stable when EVERYONE is motivated by ideals other than money - great scientific breakthrough, big humanitarian cause, etc. That's when people cooperate with each other easily, there's the rule of the law, civilization as we understand it.

But when huge piles of money could potentially materialize on the table, things revert to a much more primal state.

Agreed. I've noticed that organizations that often times claim to be 'leaderless', just have other implicit ways of exerting power and influence. Truly leaderless is difficult, and requires humility from all it's members.

This seems like an opportune moment to remind people about the Tyranny of Structurelessness article. If you haven't come across it before, it's definitely worth a read.

http://www.jofreeman.com/joreen/tyranny.htm

I was just about to post this. I find Freeman's description of successful "unstructured" groups to be an uncanny description of the idealized early-stage startup, despite her very different context.

There's possibly a good analog in data. A while ago it seemed very popular to talk about "schemaless" data systems, and I heard the perfect commentary: "If you think your data is schemaless, what that really means is you don't know what the schema is, and that's bad."

What people probably meant by schemaless: It's very easy for things to have a slightly different schema than they used to if that's needed. Things that don't all conform to the same schema can still be handled well. And it should be understood that sometimes you do need a better defined schema.

So my point? Don't focus on being 'leaderless' or 'structureless'. Focus on being flexible enough to allow new substructures in the company as needed. Make sure there's a system in place that allows people to coordinate with groups that don't have the same structure as them. And recognize when you do need a bit of structure to keep scaling.

Yep. It only works as long as everyone knows each other. I've been arguing this point for years with my anarchist friends (both left and right varieties), who seem unwilling to grapple with basic concepts like scale and externalities.

I don't think straight hierarchies are the only possible organizational structure but it does seem like more people want to be told what to do than to think for themselves. .

My experience working with anarchist friends is that something as simple as leadership creates a certain hierarchy. Obviously if there's nothing formal everyone is pretty cool, but in the organisation of a society and a longer time frame I can see how it can lead to our historical material conditions.

If you're interested, I've been toying with the idea of an organizational structure that's loosely but explicitly modeled on the human body. My basic idea is that since bodies are complex interdependent structures that mostly Just Work, social structures that explicitly recapitulate human anatomy should be inherently more robust.

Would this open the social structure to all sorts of disease states in a similar way to the human body?

Then you would need teams within the social structure who's role it is to seek out pathogenic agents / sub-structures and terminate them. An immune system, if you will. Analogy could be a police force to deal with auto-immune issues, and a military to deal with foreign invaders.

Well, in a way, our social / structural systems already do mimic bacterial colonies: reproduce by consuming all available resources until the host system dies or the infection is cleared.

Voluntary association could produce hierarchies, and it's interesting to think about what sort of societies would result. But classically government is defined as a monopoly on violence, and that formulation essentially rules out the possibility of non-coercive government, at least so long as men are capable of violence.

> but it does seem like more people want to be told what to do than to think for themselves

I just want to have some clarity around what my role is and where my responsibilities begin and end. Necessarily sometimes I'm required to reach beyond my regular duties.

In that context, then, being told what to do vs thinking for myself is a false dichotomy. The two are a nested-hierarchy, whereby I think for myself within the role required.

>I just want to have some clarity around what my role is and where my responsibilities begin and end.

This. There are two basic ways of looking at things when the organization is fluid. One, "everything is my job", which is a recipe for burnout. Or two, "I decide the boundaries of my job." That doesn't work very well either, because some people will draw the boundaries so small they never have to do anything, and others will draw them so large they don't do anything well.

I prefer hierarchy, because good boss is able to give me autonomy and responsibility without me having to constantly fight for it.

Leaderless means being bullied by the most controlling team member unless I constantly defend turf and permanently play politics.

The actual number is a bit higher than 150 (290 according to Bernard-Killworth [1]), but the real cost is all the social grooming required to hold a large leaderless group together becomes an increasing drag on the organisation's efficiency.

The ideal leaderless company size is going to be a function of the efficiency gained by having increased specialisation (each person can focus on fewer things) against the cost of social grooming. Having a hierarchical structure allows the cost of social grooming to be constrained vertically rather than across the whole company.

1. https://en.m.wikipedia.org/wiki/Dunbar's_number

"That's when the number of people involved is too great for the human brain to process the entire social network. Hence why formal organisations and hierarchy are now needed."

Or innovate on a divisional or matrix-style organization where it's a series of independent units or teams of teams. Not too different from how we make complex programs integrate well. Decomposing them properly with complexity in the right spots lets the integration or supervisory part be small and simple. Could be something to apply to real-world organizations there.

Gore & Associates (makes of GoreTex) operate like this. Whole groups of worker-owned co-ops in Northern Italy (Emilia-Romagna) also operate more or less this way, in a loosely-coupled network.

The trend in modern employment toward more individual empowerment, less hierarchy, and greater profit share points toward a co-op future for a large band of white collar jobs. Or it would, if they weren't so heavily discriminated against in the US tax code.

Acess to capital is the real killer

There are ways around that. Cooperatives can raise cash through subordinated capital campaigns (essentially an investor-backed loan at a capped rate of return) and other methods. Even capital-intensive industries like manufacturing have big, successful worker cooperatives operating in them.

Honestly, too many people make too big a deal out of the inability for cooperatives to trade ownership for cash. And yet bootstrapping a start-up is somehow still possible, acceptable, and laudable. And I won't even get into all the perverse incentives introduced by stock ownership and the secondary market.

The tax issue, though, is the deal killer. Being double taxed by the federal government on your income is a non-starter. Until worker cooperatives get treated the same as producer and consumer cooperatives, I can't see them taking off in the US.

Poptels experience in the UK shows the potential downside (I say that as an ex member )

http://www.andrewbibby.com/socialenterprise/poptel2.html

> too many people make too big a deal out of the inability for cooperatives to trade ownership for cash

It doesn't need to be impossible to raise cash, it just needs to add enough incremental difficulty and expense to outweigh the other benefits of being a cooperative.

Maybe that's the natural size for a company and growing beyond that is simply an aberration!

It might be the natural size optimal for efficiency but not for survival (in the natural selection sense). The capacity of an organization grows with headcount even with diminishing efficiency per capita.

The "natural" size of a company is determined by the relationship between transaction costs and the costs of organization. See this seminal paper for a fuller exploration of the topic: http://onlinelibrary.wiley.com/doi/10.1111/j.1468-0335.1937....

I would not be surprised, though, if there is a sudden increase in organizational costs above a certain threshold due to the limitations of human social capacities.

I assume that your number of 150 comes from Robin Dunbar.[1]

[1] https://en.wikipedia.org/wiki/Dunbar's_number

I actually don't recall where I read that. It was one of the following 2 books:

https://www.amazon.com/Sapiens-Humankind-Yuval-Noah-Harari/d...

https://www.amazon.com/Tribe-Homecoming-Belonging-Sebastian-...

I wouldn't be surprised if the book derived that number from the link you gave either.

> The problems only start once the "tribe" grows beyond 150. That's when the number of people involved is too great for the human brain to process the entire social network. Hence why formal organisations and hierarchy are now needed.

That's an interesting claim! Can we find any places where it doesn't hold up? Yep! In fact, there's a great book called Reinventing Organizations[1] where the author studied successful workplaces that have >100 people each. It's still a research project how to make these organizations work well, but the author and the community around that book are trying to gather working techniques because the benefits can be phenomenal (and that's empirical!)

[1] http://www.reinventingorganizations.com/

This is also a firm that basically consists of agile priests and people who (I assume) like experimenting with organization and management.

Lindy Beige on YouTube discusses the size of military units and how it relates to that. Platoons have 30 men (with smaller squads), a Company 100 men or so but not over 150 for the same reasons.

People tend to build a small business and aren't interested in growing number but profit.

Basecamp has been about 17 years standing, and currently has just 51 people[0]. Everyone seems happy.

I am an export in economy but most of problems companies born to solve doesn't require big size corps. At least in my country, SMEs are driving the GDP.

[0] https://m.signalvnoise.com/why-we-choose-profit-e511efc4dcb9...

Still. There are 40 people companies where some people don't have the slightest idea what others are doing (anecdata: I am working for such a company at the moment).

What if you never allow it to grow over 40? What if those bossless creatures form a swarm, with diplomats, connecting to other "tribes"? A swarm of headless amoebas, hunting down and eating a giant.. its a mad world - we might see it.

Reply on news.ycombinator.com