The Platform Wars

Joshua Citarella

May 10, 2023

TikTok vs. Meta vs. Twitter vs. Substack vs. Bluesky vs. Mastodon vs. Patreon vs. Google vs. Amazon vs. more, LFGGGGG

We’re at an inflection point for social media. Elon Musk has radically transformed Twitter. A TikTok ban looms on the horizon. The Metaverse is stalled indefinitely and Web 3 is hibernating in deep winter. Here’s what happens next.

In recent months, online content creators have observed higher click-through rates when posting images-of-URLs as compared to “click-out” links. Algorithms are serving these images-of-URLs to many more viewers (some of whom will manually tab out and type in the address), over serving content with embedded hyperlinks that may bring their captured audience elsewhere. This seems to be the case across all platforms. Try it for yourself and see.

There has always been a mild algorithmic penalty for click-out content. But Twitter has taken it to new heights in its on-going battle against Substack. Today, tweets that link directly to Mastodon or Substack cannot be reshared. Similarly, “swipe up” links on an Instagram Story now limit your reach. The same goes for YouTube videos whose captions contain links to articles, Patreon or anything else. These content skirmishes prefigure a broader platform war.

The Platform Wars is a stage of social media where platforms viciously compete for data and users by locking out features and the ability to move between these now distinct spheres. While this somewhat existed before, it will soon massively ramp up.

All platforms have deployed nearly identical tools: $5 rolling subscriptions, live-streaming, long form posts for premium users, private chats, vertical video shorts, and more. In the short term, platforms will compete on the level of culture (ie which sites have the best offerings for content). In the long term, they will compete on the level of network effects, towards the end goal of establishing an everything-app monopoly. Whether it be Elon Musk’s new X app, Meta, or something else entirely, eventually a quasi-state American analog to WeChat will arise. Soon, you will post tweets, call rideshares, order seamless and dial telehealth, from an all-in-one app. In this siloed model of the internet, your doctor being “out of network” might mean they are on Meta but not on the Amazon Health x Twitch x Whole Foods stack.

This was the kind of dystopian future scenario myself, Brad Troemel and Mike Pepi used to talk about in art & tech reading groups back in 2011. Right now, I’m writing on Substack. Later I will post the article to Twitter. Then I will screenshot it and post to Instagram Stories. After, I’ll remediate the text to Patreon which I will post in the Discord that will notify everyone when I’m live on Twitch likely reading this very post. (By the way, I’m also now on Bluesky too.) I never fell for the Mastodon meme — but congratulations or I’m sorry that happened. This increased level of administrative friction will slowly disincentivize creators from posting on every platform and lead them to optimize for a select few. Perhaps even just one.

As we exit the duopoly stage of social media, the incumbents still have a distinct advantage. But legacy sites have become overly populated by ads and may shed market share to their alternatives. Facebook and Instagram are now just cable tv for the millennial generation. There are already sharp taste distinctions between people who get their news and culture from mainstream platforms (the mall) vs. that new cool site with weird stuff (the niche record store a few blocks down).

In Platform Capitalism, Nick Srnicek describes a key example where Uber and Google compete to capture data over roadways. (Listen to our podcast on the topic here: Nick Srnicek on Platform Capitalism.) Uber maps the entire continent of Europe every week or so. Whereas Google dispatches its own cars to map the same territory every few months. The data is so valuable that companies refuse to share, or even sell it, to their competitors. Similar dynamics are now taking place on the social parts of the web.

In the late 2010’s, growth incentives aligned for cross-posting between medium-specific platforms, like Spotify to → Instagram. But in the early 2020’s, the macro incentives have shifted and platforms will attempt to pen traffic in. You will still be able to share a product from Facebook Marketplace to → Instagram Stories, because they are both in the same Meta universe. But you won’t be able to share a podcast from Substack to → Twitter, because Twitter already has a competing feature with Spaces. Soon we will see more lockouts and competition with less integration and cooperation.

Here’s the good news: In the short term, this will likely yield a better media ecosystem. The terms of service will vary from platform to platform. Some stacks will lean clearly right or left. There will be fewer instances of context collapse or Poe’s Law. Its probably better to have the Proud Boys on Gab and Antifa on Discord, instead of having them fight each other on the sole option of Twitter. People can choose what they want to see and where they want to spend their time. A more balkanized media landscape, alongside a consensus mainstream, will yield less divisiveness overall and is likely a positive development.

The more dystopian edge of this forecasting is the possibility for competing stacks to make TOS-level decisions in a quasi-state capacity. For example, Red platforms might allow only He and Her pronouns, while Blue platforms could label every user as Ze. In other instances, protected legal speech or keywords associated with unionization might be banned. Once these ideological views are coded in, users will not be able to exit to their preferred political values because they remain materially reliant on other lock-in features of the stack: like cash and health care data that are non-transferable.