The Internet Rhymes: Lessons from the Evolution of Social Media
From single-player to multiplayer
Not long ago, I read The Rhythm of War — Brandon Sanderson’s fourth installment in his fantasy series The Stormlight Archives.
A major theme across the epic is, as the name suggests, rhythms. Without spoiling anything, rhythms underpin the laws of physics within the Sanderson’s Cosmere Universe and fundamentally change how matter interacts. In fact, a race of characters attune vocal rhythms rather than facial expressions to express emotion.
What is a rhythm?
A rhythm is the sound of change.
Technology is accelerating the rate of change in our world today. It is nearly impossible to predict what will change and when. But paradigm shifts, industry disruptions, and technological innovation do not happen in a vacuum — when they occur they make a sound that ripple waves in its wake.
Often, there is a common rhythm to them.
Single-Player Social Media
Social media networks are perhaps the greatest businesses in history. Today, dynamic feeds, interactive content formats and integrated follower graphs capture user attention and carry a Spartan’s worth of demand generation on its shoulders.
Yet, this was not always the case.
Indeed, the first social “networks” were more web hosting platforms than networks. While these platforms powered the long tail of Internet users with the ability to create their own webpages, they did so in a static, single-player fashion.
GeoCities, for example, was consistently one of the most trafficked websites on the Internet during the 1990s, with tens of millions of visitors per day. GeoCities allowed users to create and publish websites for free based on their themes and interests. These websites greatly lowered the cost and degree of difficulty to produce and host digital media.
While not a true network, GeoCities ‘neighborhood’ architecture hinted at organizing content based on proximity to a person’s interests. The Hollywood neighborhood, for example, commonly hosted webpages relating to entertainment and celebrity. This architecture allowed for more collision between individuals with similar interests and enabled more socialization between GeoCities users.
GeoCities, and later MySpace, championed these user-created webpages as the primitive for new ways for people to engage with the nascent web. This media was ‘social’ in the sense that it was user-generated, an expression of identity, an often produced for and consumed by one’s friends.
Yet, distribution of these webpages often remained analog and disparate; there was no follower graph or centralized feed for one to consume the webpages of their friends. Instead, a user had to share their webpage address outside the walls of the social media platform and hope their friends bookmarked their address to stay in touch.
Moreover, while these social media platforms enabled new content creation and expanded the plane by which individuals could creatively express themselves, the paradigm remained monolithic. The atomic unit was the webpage.
To grasp this distinction, let’s do a quick thought experiment. Imagine the webpage as a blank canvas. Like the various physical canvas, it has unique intrinsic properties. Marble, for example, can be etched and carved to create physical contours. Oil canvases are primed, bound and stretched specifically to retain oil-based paint; whereas absorbent canvases are processed in a completely different manner to retain pigments bound in a water-soluble emulsion.
The art, or content, that the artist ultimately produces is constrained by the canvas. Thus, the properties of the canvas influences the content it facilitates.
Creating exceptional works of art, or facilitating content that best meets demand, requires the production of content that maximizes the intrinsic value of the medium. An oil painting atop of a marble slab is not as compelling as a sculpture; an etching of an oil canvas is not as compelling as a portrait; 240 characters on TikTok is not as compelling as a video with music. This is what I call Medium-Product Fit.
To produce and distribute content on a canvas, a publisher is beholden to the canvas’ containers. Containers are the units by which a user can produce and distribute content, media, information, art, etc. over the canvas. The more rigid the container, the more time, technical skill and energy is needed to produce and distribute content. It’s much easier to paint an oil canvas than it is to carve a slab of marble, for example.
While anyone could now create their own personal website, the entire production of the page’s content remained burdened by the publisher. Viewers could consume the website as a whole, but could not post, contribute or collaborate to its contents. Thus, the atomic unit of user creation was a single, monolithic website that was highly static and made interacting or consuming “social media” a read-only, passive experience for viewers.
In this paradigm, the relationship between publisher (producer) and viewer (consumer) was unilateral and one-to-many. Any form of communication was limited to 'a separate ‘email’ or direct message flow that took place outside of the webpage surface area as a one-to-one message elsewhere.
Moreover, the relationship amongst viewers was single-player: multiple viewers could consume the same media but had no means of interacting with each other.
Indeed, for a viewer to produce UGC (“User-Generated Content”), they could not contribute to the same webpage. They had to create a new canvas, in the form of their own webpage (i.e. MySpace profile), which was distributed as a separate monolithic object. This severely limited the net creation of content. It also prevented UGC from being interactive and truly social.
While early social media platform users could create their profiles via micro websites on Geocities, they had to go elsewhere for real multiplayer, many-to-many connection. Human connection, and true community, is not formed from passive consumption of content — or products — both through active participation in a social interaction. The majority of active social interactions take the form of a natural language conversation.
Consequently, users visiting early dial up chat rooms created new connections, new language, and new behavior. Lol, brb, ttyl, gtg emerged on these platforms. Unlike web hosting platforms like MySpace and GeoCities, these places were highly collaborative.
AIM replaced dial up chat rooms in the 2000s and quickly became the de facto communication channel for the social web. While it enabled users to instantly communicate with friends, it’s identity layer was limited to a simple username.
Nonetheless, the Away Message became a popular surface area for users to demonstrate their creativity by tinkering with font types, background colors, hyperlinks and emojis. It allowed users to play with semi permanent expressions of themselves, their interests and their current life moments.
Their friends, in turn, could asynchronously consume this ‘user-generated content’. Many times it prompted their friends to comment on the substance or creativity of the Away Message. This rewarded the individual with social capital and status, much like a ‘like’ does today.
Indeed, the integrated “Away Message” tied to a friend list was perhaps one of the earliest forms of a social feed and inched the direction of social media webpages towards a more composable and collaborative paradigm.
While The Away Message and subsequent responses were not a single, completely integrated container like a post and comment are today, it demonstrated the value of lighter, more dynamic formats of UGC creation that fundamentally tie a bilateral, collaborative means of interaction between producer and consumer.
However, these communication channels were ephemeral and lacked the robustness to create accruing benefits where a user would actively accumulate and redeem status to improve their experience. Users did not have profile pages to produce and distribute media that composed an individuals social identity, and so AIM and other chat forums behaved more like communication channels than social media platforms.
Enter Facebook
The early days of Facebook looked similar to MySpace and Geocities. Facebook acted as a web hosting platform that enabled college students to publish websites in the form of their profile page.
Again, the atomic unit of UGC was a monolithic profile page with a single container of media. Users could update the profile page by adding photos and their interests, all of which would be visible to their friends to view and consume.
In 2007, Facebook launched the initial ‘mini-feed’. While not a multiplayer object, it greatly increased the supply of user-generated content by turning user’s behavior on the platform (such as the core action of friending) into content that other people could consume.
But in 2008 and 2009, something fundamentally changed. Facebook launched The Wall and real time Home Page Feed.
While originally hated by many users and ignored at first as just a feature, these design innovations characterized a fundamentally novel way that social information was distributed on the web.
With the advent of The Wall, the atomic unit for publishers on the platform went from a webpage to a post. Viewers, meanwhile, could consume media in a new read-write dynamic via commenting and liking.
This resulted in a fundamental change in user behavior: publishers became creators, viewers became users.
This form factor change was revolutionary for a number of reasons.
For one, it created more intimate connections between users since every piece of media became a two-sided exchange between creator and user. Users could have a direct interaction with the creator as a means of consuming the UGC itself.
It also enabled a proliferation of content creation as the containers became more atomic. This lowered the ‘production cost’ of producing UGC as it required less time, energy and skilled. The difficulty of posting on someone’s Wall was much lower than creating an entire MySpace page. As a result, more surface area emerged for content production, unlocking the website as a primitive of collaboration.
Furthermore, this allowed for users to benefit from network effects. The more friends they had, the more leverage and free content was contributed to their page via posts, photo shares, likes, comments, etc. This made their profiles more interesting and dynamic, while also also creating mounting losses as a Facebook user had more to lose by leaving the network.
Below is shows a Facebook Profile Page in 2008 with the newly created ‘Wall’. The bed boxes outline the new containers where users could contribute media to the webpage in the form of user-generated content through posting, sharing, commenting or liking, making the page multiplayer.
Consequently, the webpage became a window into a social space where many people could seamlessly, both synchronously and asynchronously, conversate, interact, support, like, share and on and on with many other people in smaller and smaller containers (think of the text box) that enable new content creation. The exchange of media became more efficient and new utility emerged in socialization.
At the same time, the website became more interactive and enabled more personalized distribution.
A new post would show up both on a user’s profile page and in their friend’s home page Feed. When a user posted on another users Wall, it would should up in both places as well.
These new micro containers became the atomic units for a new homepage Feed that aggregated UGC across the Walls of your various friends and combined that media with the feed native content.
More surface areas of content production create more modular containers to distribute, which ultimately help composed the Feed, and which ultimately redefined how advertisement products could be interchanged and distributed in both a seamless and personalized manner. This network effect compounded, not just adding more supply to the pool of content, but better training the budding algorithm on what content to distribute to each individual.
The News Feed caused more people to spend more time on Facebook, a trend that picked up speed during the years that followed. Meanwhile, nearly every other social web service, from Twitter to Instagram to Pinterest, adopted the format.
Many regard it it as the most significant invention in the history of the social web.
The future bends toward Multiplayer
The era of single-player media, or read-only, was also regarded as Web 1.0. While more time spent consuming media was transitioning online, advertising spend, and the ability to monetize such engagement, lagged far behind. Mary Meeker observed the funkiness of this dynamic in her annual Internet Trends report in 2009.
But something changed around this same period. As we’ve discussed, our social media platforms transformed from single-player web hosting platforms to multiplayer networks.
The rhythms of history have taught us that digital transformation is not a story told in linear algebra, but one where an inflection point occurs that leads to nonlinear adoption of a new digitally native paradigm that (1) takes market share at a nonlinear rate and (2) grows the category overall.
This is the dance between native and skeuomorphic innovations that compound to drive more GDP online.
As online social media transformed from single-player web hosting platforms to multiplayer networks, attention and revenue followed at a nonlinear rate.
One can reasonably extrapolate that such transformation is the natural order by which such digitization occurs. Tremendous opportunity exists by recognizing this evolution and identifying industries that are constraining the multiplayer nature of people by failing to leverage their tool kit of digital primitives.
When such innovation doesn’t occur, like in eCommerce, a growing supply cannot best serve the needs of a burgeoning demand and growth flattens amidst the natural upper limit. This is because how supply and demand exchange, via a single-player webpage, has not changed.
eCommerce today looks a lot like Social Media did in 2004.
As a result, the time spent shopping online severely outweighs eCommerce spend as a total percentage of retail, reflecting Mary Meeker’s 2009 observation of digital media on the precipice of its transformation from single-player platforms to multiplayer networks.
In November 2021, I I published this observation eCommerce is Dead, it Just Doesn't Know it Yet. I concluded:
The future of digital commerce will adopt similar design construction that has pervaded across social networks and modern gaming platforms, such as connecting to a meta-game, variable reward mechanics, frequent feedback loops and identity.
[..]
Faster Delivery is a dead end. Digital marketing is a dead end. Even on a macro level, as they stand today, Shopify and Amazon are legacy businesses — they just don’t even know it yet.
When such Native innovation occurs, like it has in social media and social gaming, it causes a nonlinear growth of online GDP and enables new ways to interact with the Internet. Thus, it both pulls GDP from offline to online and creates net new digital services and products, leading to positive sum GDP growth.
That horizon is massive and inevitable. We can take lessons from the evolution of social media to will a new multiplayer future of shopping into existence and change for the floundering paradigm forever.