The eyeball factory

It seems silly to state, but it is the job of every business to produce goods. And the thing that social media companies produce is users. They are a business that takes raw material — your time and attention – and turns it into a refined good: engagement. To illustrate:

What did news publications produce? Content – the news. They monopolized the supply side of the market and sold access to that supply to readers.

So, what do social media companies produce? It can’t be content; users and publications make the content. It’s not really technology. Almost all social media can be easily replicated by other companies1. One way to put it is that Big Social sells ad spots — that’s the industry-standard way of saying it, at least. Another way to put it is that Big Social sells your attention. You know that saying, if you’re not paying for the product, you are the product?” That is what that saying means: Big Social’s only job is to produce you. They literally monopolize your attention and sell access to that supply to advertisers. You are being churned out of the sentient eyeball factory!!!

That is why, as an aggregator, when you run out of user growth, you die: you have failed to produce your product! You didn’t do your job! It’s over for you!

The End of Closed Social

TikTok, notable crab

Look, I made you some content! Daddy made you your favorite, open wide.

Social platforms have a single euphemistic metric they maximize: engagement. Every platform competes in a perfect market for our attention, because it is the prerogative of Big Social to keep you hooked. They need you to spend the precious free minutes of your day on their platform, because without your eyeballs, they’ve got nothing to sell to advertisers. So if someone invents a more addictive cigarette, the old platforms can either copy them or risk irrelevance. That’s how you get a chum box. That’s how you get the CEO of Netflix saying that they compete with sleep”.

I’ve long argued that TikTok represents the perfection of the phone-based social platform. Short-form video is highly engaging, and supercharging content discovery with a UI designed more for AI than humanity is exactly what data mining has promised to do for us (and for surveillance capitalists) since Facebook started serving us content in the News Feed. Short video + algorithmic feed = the perfect engagement mechanism. So Facebook…and YouTube…and Instagram…and Snapchat…and THE GUITAR TAB APP I USE have no choice but to design for the lowest common denominator: customized, hyper-targeted, user-generated, micro-video feeds. These CHUM2 feeds are the platonic ideal for engagement, maximized.

By the time TikTok showed up, its biggest competitors were already getting old. When it started stealing users and their attention, the apps simply copied it, as is industry custom, without hesitation or shame. Instagram, Snapchat, and Facebook were redesigned to look, feel, and in some cases recommend content more like TikTok’s. YouTube nagged creators to post Shorts” and tried to get users to watch them. Now, the first thing Twitter users see is algorithmically scavenged content in a feed called For You.” — John Herrman, Why Every App Now Feels Like TikTok, But Worse

You know how, given a crustacean, nature evolves a crab? That’s what is maximizing engagement does to social. At some point, you’ve got to wring value out of all that engagement your platform harnesses: given a closed, ad-funded social platform, nature evolves a CHUM feed.

Alright, then. Let’s start a list. It’s clear that:

BUT — in becoming counterfeit TikTok, these platforms have to sacrifice their soul. The platforms cease to be the medium they once were: Instagram is no longer a place to post photos with cute filters, it’s a CHUM feed. Facebook isn’t a place to visit your friend’s wall and post a message, it’s a CHUM feed. Snapchat isn’t a place to share ephemeral photo messages with friends, it’s a CHUM feed. Twitter, even, isn’t a place to post deranged stream-of-consciousness thoughts, it’s a CHUM feed.

The trade-off was clear: In exchange for more of a certain kind of engagement from your users, you make their experiences subtly worse, or at least less personal, under the guise of personalization” (there is perhaps no more impersonal form of content than the second-tier For You” recommendation). — John Herrman, Why Every App Now Feels Like TikTok, But Worse

This convergent evolution is the penultimate step in the demise of Big Social. They don’t realize it yet, but they’ve walked themselves into a corner that no capitalist wants to find themselves in: the commoditization corner. In general, businesses do not want to be commoditized because it means that their products or services are seen as interchangeable with those of their competitors. Say, for example, one CHUM feed for another. Being commoditized can also mean a loss of brand identity and differentiation, which can make it harder to build customer loyalty and maintain a competitive advantage. Commoditization is scary for companies because it means they are in the business of widget making, which is the business of low margins, efficiency, and sad CEOs.

So it’s all kind of dizzying trying to understand why social platforms would risk turning themselves into the same product as all of their competitors. To understand, you should know that nature evolves a crab given a crustacean, not any old sea creature. In the same way, nature evolves a CHUM feed given a certain type of company.

Big Social, notable crustacean

Could I interest you in everything all of the time?

Remember, every social looks the same because every social’s incentive structure is the same: maximize engagement. But why does every company need to maximize engagement? Change the incentive, change the product! Right‽

Before the Web, traditional publishers were the dominant aggregators” of content: they were successful businesses because they bundled the printing press with editorial with paper boys.

When the Web arrived, it lowered the barrier to both the publication and the distribution of content — this new technology commoditized the traditional publisher’s product. Anyone with an internet connection and a dream could have their own voice online. At first, this was ok for publishers. The Web was growing, and growing quickly. A rising tide lifted all the great boats.

But the Web’s greatest gift created a new problem: with the democratization of publishing came a lot of content. For new users of the Web, the problem was no longer obtaining good content, but knowing where to start and where to find it. This created an opportunity for new players to emerge — true Web-native companies were built solely to bundle and taxonomize the best content of the Web. These curated aggregators” took the shape of companies like Yahoo, a massively valuable 90s-era company that made nothing but a list of links.

Through the 90s, this process would repeat itself: the web would grow, a company would find a new way to bundle all of the content together in one user interface, and that company would become the new king of the hill. Yahoo was successful because they put a huge portion of the web on one page. Google beat out Yahoo by aggregating all of the web into one text box. Amazon got big because it put all of the books in the world in one store.

This is the essence of Ben Thompson’s aggregation theory: the dominant players on the Web are often those that can aggregate and control access to a large number of users, suppliers, or both. This allows them to create value by leveraging network effects and economies of scale.

And what came after Yahoo and Google? Perfect aggregators. Social networks. Content platforms. Facebook and YouTube, Twitter and TikTok. Aggregators so complete in their commoditization of content that they placed the New York Times’ articles right next to your decrepit Facebook happy birthday” posts. These 2000s-era Web companies were living the dream of the 90s: get big and profit. Every last one used the same playbook, the aggregator’s playbook, and as a result, every last one ended up with the same incentive structure: grow engagement.

/history lesson

So, an updated list:

If you give a mouse a cookie…

The Aggregator’s Paradox

Here’s the thing about aggregators: they are using one of the most successful business models ever invented. You can’t really blame them3. But, in general, there are only two options to keep growth going.

Plan A is the best option: grow your user base. Remember, this is the playbook of aggregators: provide a good service, which attracts people to your service, which attracts suppliers to your service, and you’ve suddenly got a virtuous cycle going.

But, inevitably, there comes a point where user growth stops. Maybe you run out of people interested in microblogging. Maybe you run out of humans to make a Facebook account. Maybe the next generation stops coming to your platform. It doesn’t matter how, really. User growth always stops.

To keep growing your revenue, you’ve always got Plan B: get your users to use your app more! This option is still pretty good. Sure, you’re making your users more dependent on your app. You might get some flack for making your service too addictive, but it’s a price you’re willing to pay to keep your investors and publishers happy. After all, the more time your users spend in your walled garden, the more you can advertise to them. This usually starts out as a shift from a chronological feed to an algorithmic one.

But, wait. We know this story – this plan suffers from the carcinization problem. If there’s another app out there that does a better job taking up your users’ attention, they’ll just leave you in the dust. In that case, you’ve got no choice other than to steal ideas from that competitor, subsidize content to fill your app’s new feature, and then pray your user base doesn’t abandon you. This is, in every way, a corruption of the medium you established when you started your social network. It’s a step towards commoditizing your service. But it’s a price you’ll have to pay to keep growing.

But there must be a Plan C? Maybe charge users for access to your app? Hell no. Maybe start charging publishers for access to your existing customer base? In the short term, it solves your revenue problem, but that doesn’t solve your growth problem. In the past, you might solve your growth problem by buying another social network popular with the next generation of users. But that won’t happen in today’s regulatory environment. I guess Plan C is to pray that Congress bans TikTok to buy you some time.

This impossible set of choices is what I call the aggregator’s paradox. When you run out of user growth, your march toward death is inevitable.

I forget…

Aggregators gonna aggregate. But given sufficient technological innovation, aggregators also gonna get aggregated. What happened Yahoo — and what happened to traditional publishers before them – will someday happen to Big Social. I believe that we are in the midst of that shift now, as exemplified by whatever it is that’s happening on Twitter.

After the dot-com bubble kicked off the new millennium, people stopped believing in the promise of the Web. But Google and Amazon – and memories of Netscape – carried the torch from Web 1.0” into a new era. Here, with a proven business model and an open Web, social media giants like Facebook, Twitter, and YouTube emerged. Suddenly, it wasn’t enough to have your own website - you had to be part of a larger social graph. Facebook, Twitter, and their ilk became the new gatekeepers, controlling who saw what and when. This has been our Web 2.0” experience, defined by user-generated content, algorithmic feeds, and the creation, abuse, and destruction of social graphs.

A final list:

So now, a new hope. The decentralization movement represents an opportunity to commoditize the key remaining asset of Big Social: the social graph. Our eyeballs. In a way, they’ve already given up on social graphs…it’s too late, they’re all TikTok now. Big Social has literally taken social connection and mined it for all it’s worth. And that’s ok! We’ll just know that big tech aggregators are always ad platforms, first and foremost. The Internet was always bound to have an ecosystem of billboards, I guess.

Let’s take back our connection to others – distributed social networks like Mastodon and protocols like ActivityPub enable new forms of peer-to-peer communication and collaboration where we own our relationships online. We’ve got the chance to build our social graphs into the structure of the Internet itself, which may sound radical, but feels better than building them into the structure of private aggregators, who will always — always — abuse and abandon them.

Some have opined that web 3 is all about crypto or the metaverse” — sure, maybe4! I take a more conservative stance: web 3 will begin with decentralized social. The social web. Maybe more decentralization will follow; as has always been the case with the Web, the story is all about commoditization and aggregation. This time, though, social media is legacy media. And it’s honestly their turn to be commoditized.

Now, we can build our social graph without the need for centralized intermediaries, the same way the Web built structures for publishing and distribution without the need for centralized intermediaries. We’ve got the power to do that! But I forget, does anyone know if the Web was successful?

  1. e.g. Stories.↩︎

  2. 🫡 I might be angry at myself for this terrible hyphen-ridden backronym if I wasn’t so angry about the state of social.↩︎

  3. You can.↩︎

  4. If you know someone who can convince me that crypto is an actual substitute for money, @ me –↩︎

April 20, 2023 · tech · post

Previous:Harbingers (Part 1)
Next:The Studio Metaphor (A Harbingers Interlude)