The Clone in the Wings

On July 5, 2023, Meta shipped Threads — a Twitter clone so unabashed it didn't bother pretending to be anything else. Same column of text. Same reply chains. Same dopamine architecture, minus the blue-check drama. Adam Mosseri had the thing sitting in the wings for months. What he needed wasn't a better product. He needed Elon Musk to trip over his own feet — and that weekend, right on cue, Musk throttled Twitter with "rate limits," locking users out of the timeline they were addicted to. The exit door swung open. Meta was already standing in it.

The numbers were obscene. Thirty million signups in a day — enough to snap ChatGPT's record. A hundred million in five. The fastest-growing platform in the history of the species, and it happened precisely because the platform already existed. Threads didn't ask you to sign up. It asked Instagram for your username, your face, and your followers, and ported the whole social graph over while you blinked. Two billion Instagram accounts, one tap from a new feed. That's not growth. That's a wire transfer.

And then — of course — it emptied out.

By late July, daily active users had cratered more than 80%. Average session time fell from twenty minutes on launch day to under three by August 1. The fastest-growing platform in history became one of the fastest-evaporating, and Meta relearned the lesson every clone eventually learns: you can copy the interface, you can copy the graph, you can copy the exact pixel width of the reply button. You cannot copy the reason anyone showed up.

Here's the part the growth charts can't measure. A platform isn't its features, and it isn't its headcount. It's the pattern of attention running between the people on it — the running jokes, the beefs, the accounts you check first, the low hum that something is happening here and you'd better be in the room. That pattern is not a database you can migrate. Twitter, for all its rot, was a place. Threads was a very well-attended lobby where a hundred million people stood around asking what they were supposed to do next.

Meta mistook the map for the territory. They held every Instagram address in the world, so they mailed everyone a key to a house nobody had a reason to live in. The crowd walked through the door, looked around, and left the lights on for someone else.

To be fair — and it costs me something to be fair — Threads didn't die. It bolted on a following feed, wired itself into the fediverse, and clawed back a real user base over the two years since. Credit where it's due: they shipped, and they stayed. But notice how they got the second chance. A startup that sheds 80% of its users in three weeks is a corpse. Meta just shrugged and kept walking — because it owned the distribution. Two billion Instagram accounts meant it could fail in public, eat the humiliation, and keep mailing keys until a few of them finally turned in a lock. That's the cold cut the launch numbers and the obituaries both miss: the moat was never the hundred million. It was the power to survive losing them.

Which reframes the whole trick. Presence didn't fail to port — it failed to port instantly, because presence is the one asset on a platform you cannot wire over. You grow it, the way you grow anything, in soil, over time. That's what the slow recovery actually proves. Distribution power buys the land and the second chance; it does not buy the reason anyone stays. Meta had to earn that the long way, one running joke and one checked-first account at a time — the way a place is built rather than transferred.

The clone was always going to launch. The only open question was when the original would stumble, and Musk answered it faster than anyone dared hope. Somewhere right now there's another clone in the wings — of something you use, something you'd swear couldn't be copied — waiting for the same weekend to come around again. It always does. It'll pull the same crowd through the same door, because the crowd was always portable. The reason they'd stay never is.