Social media engagement algorithms and the illusion of choice

January 11, 2023

So many of us, over the last couple years, have been rethinking our relationship with social media and the internet at large.

My own wonderings about technology have seen me dabbling into using only open source operating systems and software. But I've recently realized that while I appreciate open source and like the idea of all technology being open source, I am not an open source purist or absolutist. Like so many digital citizens, I have concerns about privacy and security. But in these areas, again, I am not a purist.

If I feel this way about technology at large, it only makes sense that I have similar concerns about social media. And I imagine I'm not alone.

So if I'm correct, then it makes sense to ask:

Social media used to be a way to stay in touch with friends and family and random weirdos you found in various corners of the internet. But now it feels like this thing we do out of habit, even though it drives us crazy.

The problem with social media overwhelmingly seems to be the engagement algorithms. The efforts to keep us coming back for more to boost ad revenue, even if that increased engagement results in angrier users.

Angry users—a side effect of these engagement algorithms—is definitely a problem. But I think there's another factor that doesn't get enough attention: These engagement algorithms lead to an illusion of choice.

Users put effort into finding and following relevant and interesting voices, only to have another factor—the abstract yet opaque engagement algorithms—determine their experience on a social network.

When we don't know how these algorithms work—or when they're working—how can we be confident in our ability to build or curate own digital experience? Where is the line between being responsible for our own experience and being manipulated by mysterious forces we're ill-equipped to fight?

As Cal Newport has pointed out, social networks like Facebook, Twitter, and Instagram built their services through network effects. People joined these sites to connect with others they knew or were interested in. And the fact everyone they knew was already on these established sites was enough to keep them there. Leaving behind your connections and starting over was too costly.

But these social media giants gave up the advantage of network effects when they started using engagement algorithms. They threw away the main reason people used their services. And in so many ways, the move from network effects to engagement algorithms felt like a bait and switch.

Why aren't engagement algorithms on TikTok a problem?

While concerns about security and privacy on TikTok appear valid, this post will ignore, but not discount, those concerns to stay on one point.

While users may have concerns about the types of content the TikTok algorithms serve, most users are not bothered by the presence of algorithms themselves on the service. Unlike the case with Facebook, Twitter, and Instagram, engagement algorithms are part of the appeal of TikTok. On TikTok, the algorithms are a feature, not a bug.

TikTok can get by with using algorithms because TikTok doesn't give users the illusion of control. TikTok doesn't pretend to deliver content based on whom you follow. It's common knowledge that TikTok's engagement algorithms, aided with data from your views, likes, comments, and shares, decide what you see in your main feed. You have to choose the Following feed for the hope of seeing content from those you follow. So while keeping up with those you follow is an option, it is not the default. TikTok is made for finding engaging content, not for keeping up with those you already know. TikTok kind of gives you the option to follow individuals, but it doesn't put much effort into that angle.

While TikTok may be worthy of criticisms in some areas, the service deserves credit in terms of algorithms. While parent company ByteDance may not be transparent about how TikTok's engagement algorithms work, it has at least been transparent in the fact TikTok operates through engagement algorithms.

Users are aware of the presence of algorithms when they sign up for TikTok. They know what they're getting into. And they're mostly fine with that because they're not signing up to keep in touch with friends and family as they did on other sites.

The network effect seems to persist only as long as social media services steer clear of engagement algorithms. A couple such examples include Mastodon and micro.blog.

Networks on LinkedIn once had value because the connections were likely to be genuine, in that you either knew the person you connected to, or you had an interest in that person. But now, many connections are made only for the intent of increasing who sees a member's content, AKA engagement.

And once users figure what gets engagement, it's only natural that many would start creating content in the tried-and-true formula. So users see the same types of content over and over. Originality exists on these platforms. But it goes unseen, unrecognized, unappreciated. And so mainstream social media becomes the digital suburbs, full of cookie cutter houses lined with the same bushes and political signs.

While web3 promises to solve all our digital woes, I find the solution to be a simple and old technology: RSS.

Watch the video below if you're unfamiliar with the wonders of RSS:

🗒️ Note: This video is 15 years old, so it doesn't address that Google Reader is now dead. At only $15 a year, Miniflux a great alternative. Or check out Reeder 5 or NetNewsWire if you're on Mac/iOS/iPad OS.