Social Network from first principles

Proof of X by Julian Lehr has had a rare effect in me. Since I have read it, almost a month ago, every other day my conscious self is driven back to it. If you haven’t read it, you should.

Julian opens with:

At their core, social networks are primarily about one thing: Building social capital through signaling.

And continues with:

Their primary role is to distribute signaling messages at scale and transform them into quantifiable social capital (in the form of likes and followers).


Social networks are therefore not only signaling distribution (and amplification) networks – they also allow users to prove their signaling messages.

In essence, Julian defined what a social network is from first principles:

  1. Showcase what you have done (proof of work). This can come under many shapes and forms, like Tweets, Instagram Posts, a Clubhouse room, a TikTok video.
  2. While an audience consumes and pays for it (the social capital). The payment is done via a like, a favourite, a follow, an endorsement. Capital is measurable.

An example is liking someone’s post:

Tommy Rivs is a Ultra Running Legend

Tommy Rivs is a Ultra Running Legend

Different companies will employ multiple mechanisms as both proof of work and social capital:

What I have mention before defines the atomic characteristics that any entity needs to have, in order to be a social network. Starting from these, it becomes easier to categorize these systems.

For instance, one can categorise StackOverflow has being one, even tho there’s a thread disputing it. StackOverflow, and the whole StackExchange network, has a clearly defined proof of work and social capital with answers, reputation and badges. Quora follows the same pattern as StackOverflow, via likes and followers.

My profile on StackOverflow

My profile on StackOverflow

Not having followers has the condition from which we assert that a system is a Social Network or no, is a naive view of the problem.

If we look at WhatsApp from these lens, it has a clearly defined proof of work: texting, replying, interacting with others. The social capital is disputable however. One could make the case that having a great conversation with someone can constitute social capital. The more people know that you are an awesome person to talk with, the bigger the social capital. The problem is that you cannot quantify said capital. You can say you had three awesome conversations last week and that people think you are a nice person, but those feelings do not take shape in the digital world.

Facebook is able to do so via badges within it’s Groups product. The individual, is rewarded by the value of what they have written. A striking difference between StackOverflow and Facebook is that the former is focused on the utilitarian value, while Facebook takes a more holistic approach on the subject:

Facebook’s badges

Facebook’s badges

Facebook knows how impactful these badges are in instigating good, healthy conversations, that might not have any serviceable value, but can have a positive net effect:

  1. Incentivise newcomers to participate and feel welcome.
  2. Encourage people to participate more in conversations. Potentially breaking Nielsen’s 90-9-1 rule.
  3. Promote healthier, more civilised conversations with multiple points of view.
  4. Create closer relationships.

I do not believe that these attributes, or mechanisms, would work particularly well in places where the Interest in the Topic is much higher than the Interest in the Group Members. An example of this would be a subreddit like wallstreetbets: people are there because of the topic, not because of its members. This is one of the reasons that this mechanism, in Facebook, is only accessible to groups and communities. In particular (emphasis mine):

We know an important part of building a community is getting to know other group members. New members have expressed that they would like an easy way to identify leaders within a group. We’ve also heard that admins and moderators want new ways to welcome members and recognize current members for outstanding group contributions.

The goal is simple: build closer relationships between members. Encourage people to express themselves without fear of being criticized. Create a feeling that being emotionally vulnerable is the best way to create trust. All of this by sharing ideas and experiences.

Going back to WhatsApp, it doesn’t provide such mechanisms, because it wasn’t conceived to be a social network and has no pretence to be one. Calling WhatsApp a social network is the equivalent of saying that email is one. The proof of work exists, but there’s no social capital to be obtained.

One could however imagine a service where by introducing these incentives, better, closer, more respectful conversations could emerge. Facebook, from a first principles that Julian established nails both the proof of work and the social capital. However, in particular with social capital, it’s difficult to transfer said capital across other communities. If you are a nice person to talk with in a cat group, you likely are as well in a group about board games. It’s the equivalent of being rich in England but not being able to show-off your wealth in France. The proof of work, as is, it’s also simple, not much more commenting/being active in the group. The social capital signalling is weak and bound to a small label next to your name. HackerNews status signalling is equality weak, but I fear it’s either an after thought, or not meaningful component of the platform.

A better service would be able to showcase your “work”. This would likely be the ability to have a rich profile, from where people that don’t know you could see your badges, reputation, or maybe a list of conversations where people endorsed your words and thoughts.

The solution space through these lens is not small.

Feel free to get in touch about this topic here.