The algorithm hates me.
Many of my connections on Mastodon celebrated their third anniversary on the service this month, having joined in the great migration from X, the Everything App™ (née Twitter) in November 2022. I had joined (and left) Mastodon a few years before, but came back in October of 2022—the month before the big wave, and sixteen years after I originally joined Twitter.
I was early to Twitter. October 24, 2006 early. My user number was 10,231. I know in later years, Twitter user numbers increased monotonically (a user’s number will always be greater than that of all users who joined before them) but didn't correspond directly to the number of accounts on the platform. I suspect in the earlier days, they were a sequential count, but either way, I was at most the 10,231st user to join. For a service that would eventually boast hundreds of millions of users, I don’t think it’s boastful to say I was early.
One might assume being so early to a massive platform like Twitter would translate into . . . something. Anything? But when I left Twitter at the end of 2022, I had just over 400 followers. My peak was somewhat higher, and came in early 2020, when I posted a viral tweet that was retweeted 14,000 times and seen by three and a half million people. But gravity ultimately won out and dragged me back down.
Sixteen years.
Tens of thousands of tweets.
Just 400 followers.
Which brings us back to the present, and Mastodon. Aside from a brief flirtation in 2019, I was basically a newbie when I came back in 2022. I had zero followers. And now, three years later, and on a platform with a tiny fraction of Twitter’s user base, I have ten times as many followers as I had on Twitter.
I know it’s gauche to talk about follower counts. But I am gauche. And the point I’m making isn’t that I’m awesome, but that there is a quantifiable difference in my performance between the platforms. Let’s be super conservative and say Twitter is ten times the size of Mastodon. I gained ten times as many followers out of a pool one-tenth the size of Twitter, in one-fifth of the time I had spent on Twitter.
Do that math: it is not an exaggeration to say that I am FIVE HUNDRED TIMES AS POPULAR ON MASTODON AS I WAS ON TWITTER.
What accounts for the difference?
I don’t think it’s me. It’s fair to say that I am not the me that I was on Twitter. I have a stronger voice on Mastodon. I take more chances and share more of my foibles and vulnerabilities. My jokes and serious posts both connect more often. But am I 500X better than I was on Twitter?
No, I believe the real reason I succeed on Mastodon but sucked on Twitter is that the algorithm hated me.
What is the first thing you hear when you receive the Good News about Our Lord and Savior, John Mastodon? “There is no algorithm.” Setting aside that this isn’t strictly true—“show me the posts from people I follow in reverse chronological order” is technically an algorithm, and the “trending” lists are algorithmic—what people mean by this phrase is both real and important: there is no corporate interest controlling what you see on Mastodon.
On Twitter, that wasn’t the case. You are shown a timeline of new content designed to increase engagement, to keep you coming back for more.
One thing that often frustrated me on Twitter was seeing accounts that started later than mine take off while mine stood still. I would post what I thought was a real banger and get little to no engagement. My follower counts stayed level for years. A close relative racked up 40,000 followers and actually made Twitter her job. I would have occasional breakout tweets with hundreds or even thousands of retweets, but I remained stuck, a noncombatant on the battlefield of memes and ideas.
I came to suspect I was in some kind of algorithmic jail. I wasn’t aware of the concept of “shadowbanning” at the time, but it felt like I had run afoul of some sort of rule, and that my account was marked for obscurity. (I did call Twitter founder and then-CEO Jack Dorsey a Nazi once, but that was years after my troubles began.)
This may sound paranoid, or like I was (am) entertaining a very elevated idea of the quality of my output, but again: am I really five hundred times as good at Mastodon as I was at Twitter?
No, I think the real story was that I was making an error about the very idea of what the Twitter algorithm was for. I assumed, naively, that the algorithm was designed to show me things that I would like, things that I would have an affinity for, or at the very least things that were good (or, good’s proxy, popular). And if we assume that the algorithm is there to show us good stuff that we would like, it’s fair to assume it will show our best stuff to people who would like us.
But it doesn’t. What Twitter cared about was engagement. Making you laugh, making you mad, making you happy, making you sad, making you horny, making you anxious—it’s all ultimately the same, as long as it gets you to engage, to interact, to keep opening the app and scrolling.
For whatever reason, Max-on-Twitter never quite fit that bill. And that became somewhat easier to take once I recognized that what I was doing wasn’t rejected because it wasn’t good enough; it was rejected because it didn’t fit the agenda of a manipulative, addictive system.
And frankly, my content was good enough, because as soon as I got away from the algorithm, I did find my people.
