Judging myself by likes

How we measure success online (and off)

Welcome to my personal newsletter. I’m publishing essays on digital technology and culture in the run-up to my January 2024 book FILTERWORLD: How Algorithms Flattened Culture. Pre-order it now! Subscribe to this newsletter or read the archive here.

---

When I launched my Filterworld book cover last Monday, it was a great opportunity to test how social media works at the moment and see how algorithmic feeds pick up content. Every platform behaves differently, but they all measure engagement: how many users decide to click the like button on a particular post. My Substack newsletter itself got 34 likes — a fine amount, since likes matter much less than email opens. (It’s not clear what a Substack like really means.) On Twitter, my tweet of the cover got 130 likes, which is not very good considering it was a big personal announcement and I theoretically have 26,500 followers! (I feel like I’m negging myself here.) And yet I was okay with it, because Twitter is actively falling apart and few of my tweets get many likes these days, whereas a few years ago 130 was a common occurrence. (Negging myself again.) Twitter’s feed is much more algorithmic now and I can’t compete with monomaniacs like Menswear Guy. But on Instagram, where I very rarely post on the grid and my median like count is maybe 30, my cover post got 551 likes, with 4,246 followers. 551! It’s quite possibly my highest ever.

Many factors go into this success. The cover is a visual artifact, of course, so it suits the Instagram feed. It was also reposted by other people — friends, colleagues, and collaborators. That helped it get more exposure. But watching who liked the post and when, it was clear to me that Instagram was serving it up to more and more of my followers over time as the algorithmic recommender system saw that the post was succeeding. The first people to see and like it were my friends, who I interact with on Instagram all the time. Then the likers grew farther flung and less known to me; the post reached a swath of people who, though they do follow me, never see my posts because the algorithm doesn’t think they’re that good (which is to say, likely to inspire likes).

We’ve been encouraged to judge ourselves by likes ever since Facebook added the thumbs-up button in 2008. Likes are a way to keep score, to compare content that is actually incomparable — an engagement announcement versus a vacation photo, say. I’ve certainly judged my articles on how many likes they get. But as feeds have become more algorithmic, it’s harder to tell when a post gets fewer likes because the algorithm ignores it, or because your followers just don’t like it. The calculus has changed. It feels like the middle ground has fallen out: a post either gets zero attention or tons of it. Which factors determine its fate — likes, comments, replies, shares? — we never know, we just try to guess and game the system in our constant state of algorithmic anxiety.

So I’m sad that I got too few likes on Twitter and happy that I got more likes than I expected on Instagram. That doesn’t say anything about the actual quality of the book, none of which is public yet, only the cover and the marketing copy. Two articles this week reminded me of the stranglehold likes, and the pursuit of online engagement that they represent, have on our lives:

• The Connecticut senator Chris Murphy wrote a NYT op-ed titled “Algorithms Are Making Kids Desperately Unhappy.” “Algorithmic recommendations now do the work of discovering and pursuing interests, finding community and learning about the world,” Murphy writes. The way that algorithms figure out what you will pay attention to is by your likes: Whatever you hit like on, you’ll get more of it. But that process robs us of some of the wandering, ambiguity, and challenge that happens when we figure out what we like on our own. Murphy writes: Why should users “take the risk to explore something new when their phones will just send them never-ending content related to the things that already interest them?”

• At Vox, Rebecca Jennings wrote another fun jeremiad against tourism culture. TikTok-era, post-pandemic tourists are crowding the same small set of places (cough Lake Como cough) and making them more miserable both for themselves and the locals. Do the “best” places get the most likes? The piece is against optimization, the idea that one needs to hit the best spots, follow the best-of lists, sort your experiences by the highest number of likes. (Note that Vox Media publication The Strategist is very guilty of perpetuating this kind of life-hacking consumerism.) Stay tuned for next week, when Jess and I and the gang head to lesser-known areas of Provence and watch out for other Americans in the village boulangerie.

The age of likes might be slowly ending, but not in a good way. On TikTok, you don’t need to hit a like button for your attention and engagement to be measured and factored into the recommendation algorithm. It’s enough to just keep watching a video, search a term, or click in to an account. It’s not that likes aren’t there; it’s that they’re mostly invisible, but they’re still influencing everything regardless, reinforcing the tyranny of data, engagement, and likability.

---

My next book Filterworld: How Algorithms Flattened Culture comes out January 16, 2024 from Doubleday! Request it at your local bookstore or pre-order here.

---

Notes

• Jess wrote a great piece on how extremely online the Republican party has become. Ron DeSantis is courting the Internet far-right with incomprehensible references and meme-y videos, but Trump is still winning by far on his home digital turf.

• Bloomberg says that mini-martinis are in!!! I’ve long been a fan of making these at home, where drinking a full martini before dinner has a tendency to make me a worse cook. But the other night at a hotel bar in DC, at a friend’s urging, I ordered a round of mini-martini shots, which came in little crystal glasses. These were a resounding success. The bartender even customized each one, dirty, with a twist, or clean with an olive (my standard order). Try it at your local cocktail bar!

  

  Le Clou in DC