So remember last weekend, when half of Twitter spent a Saturday convinced the sky was falling because of a Buzzfeed report that the company might test an algorithmic timeline? #RIPTwitter was trending, people were posting their greatest hits tweets like it was a eulogy, Jack Dorsey had to hop on and personally reassure everyone that nothing was launching imminently. I rolled my eyes through most of it, not because the concern was crazy, but because everyone arguing about chronological vs. algorithmic was missing the actual problem, which is that the user growth numbers have been flat for ages and a sorting change was never going to fix that.
Well, today we got the earnings report, and it turns out I was right to be annoyed at the wrong debate happening. Monthly active users basically didn't move, still parked around 320 million, and the stock got hammered in after-hours trading, down over 12% at one point. This is a company that IPO'd at $26 a share and is now trading somewhere in the $16 range. That's not an algorithm problem. That's a "normal people don't understand how to use this thing and nobody at the company can decide how to fix that" problem, and it's been the problem since roughly 2013.
I've been on Twitter since 2009, back when the whole appeal was that it was the one place breaking news showed up before anywhere else. Earthquake in your area? Twitter knew before the news vans did. And it's still genuinely good at that, better than almost anything else out there. But try explaining Twitter to a normal person who isn't already extremely online and watch their eyes glaze over at "reply, retweet, favorite, quote tweet, and also here's a stream of half-finished conversations from strangers in an order that used to make sense but might not anymore, good luck." I've tried to get my sister onto it three separate times over the years and she bounces every single time within a week. That's not an isolated failure, that's basically the whole growth story in one anecdote.
The thing that actually annoys me about the algorithmic timeline debate is that both sides are treating "chronological" as some kind of sacred design principle instead of what it actually is, which is just the easiest thing to build in 2006. Facebook moved off pure chronological ordering years ago and nobody serious argues Facebook would be better today if it hadn't. I'm not saying Twitter's algorithm, whenever it actually ships, will be good. Knowing Twitter's product decisions over the past few years, I'd bet against it being good on the first try. But the instinct to treat "the feed is in time order" as the one untouchable pillar of the product, while the actual company is hemorrhaging its ability to bring in anyone who isn't already a power user, feels like arguing about the color of the deck chairs.
What I actually want from Twitter, and I doubt I'm getting it, is better onboarding and some kind of real answer to "who do I even follow" that isn't a generic celebrity list. Not a magic sorting algorithm. Not another redesign of the icons (please, no more icon redesigns, the heart-instead-of-star switch a few months back was already more controversy than it was worth). Just make the thing legible to a new user in the first ten minutes, because right now it very much is not, and no amount of feed-ordering cleverness fixes that.
Anyway. Dorsey's back as permanent CEO now instead of interim, which I think is fine, he clearly loves the product in a way that matters, but loving the product and knowing how to grow it past its current ceiling are two different skills. We'll see what Q1 looks like. I'm not holding my breath, though I'll admit I'll still be checking my timeline obsessively regardless of what the stock does, because that's the other thing about Twitter nobody ever mentions in these earnings post-mortems: the people who are on it don't leave. We just complain a lot. Which, fair, I did some of that complaining myself just now.