And You Will Know Us by the Company We Keep — Remains of the Day

I refer to this as the problem of graph design: When designing an app that shapes its user experience off of a social graph, how do you ensure the user ends up with the optimal graph to get the most value out of your product/service?

For all the debate over whether our current social networks are good for society, I prefer to focus on the potential we've yet to realize. We have the miracle of Wikipedia, yes, but aren't there more types of mass scale collaboration to be enabled?

Recently, Instagram announced it would start showing its users posts from accounts they don't follow. In many ways, this is as close to a concession as we'll see from Instagram to the superiority of TikTok's architecture for pure entertainment.

The term "follow" is fitting. Who we follow can become a self-fulfilling prophecy. First you build your graph, then your graph builds you.

This doesn't stop social apps from trying to fix the problem. Reduced traffic to the feed is existential for many social apps. Instead of fixing the root problem of the graph design, however, most apps opt instead to patch the problem. The most popular method is to switch to an algorithmic, rather than chronological, feed. The algorithm is tasked w... See more

It feels as if we're at the tail end of the first era of social media in the West. Looking back at the companies that have survived, certain application architectural choices are ubiquitous. By now, we're all familiar with the infinite vertical scrolling feed of content units, the likes, the follows, the comments, the profile photos and usernames, ... See more

Twitter, unlike Facebook with its predominant two-way friending, is built on a graph assembled from one-way follows. In theory, this should reduce its exposure to graph design problems. However, it suffers from the same flaw that any interest graph has when built on a social graph. You may be interested in some of a person's interests but not their... See more

The way China has built out its social infrastructure is, in at least this respect, more logical. WeChat owns the dominant social graph, and it acts as an underlying social infrastructure to the rest of the Chinese internet... Rather than duplicate the social graph of everyone, which WeChat owns, other apps can focus on what they do best, which mig... See more

A classic example, though I don't know if this still persists, is how Pinterest skewed heavily towards female users at launch, losing lots of potential male users in the process. This was a function of building their feed off of each user's social graph. Men would see a flood of pins from the females in their network as women were some of the stron... See more