Meta's new app Muse has, according to TechCrunch, outdownloaded and outused ChatGPT in its first twelve days. More installs, more daily active users, higher App Store rank. Read as a headline, it sounds like Meta has built something better than ChatGPT. I don't think that's what it shows at all.
The article is honest about the mechanics, which is why I trust it more than the headline it produced elsewhere. Over 95 percent of Muse's users are also Facebook users, and 63 percent are on Instagram. Muse works inside WhatsApp too. Meta didn't launch an app into an empty market and win on merit. It launched an app into a market it already owns, through platforms with billions of people already logged in, already trained to tap a notification from Meta. ChatGPT had none of that when it first arrived on mobile. It had to earn every one of its downloads with word of mouth.
Distribution wins the first two weeks, every time
I build much smaller things than Muse, but I've seen the same pattern at that smaller scale, and it taught me not to be impressed by launch curves. For BabyJunctions, I set up AI-made slides and carousels that post to Instagram on a schedule, so the account never goes quiet. That alone changes the numbers a follower sees: consistent posting, a live-looking feed, engagement that compounds because the algorithm rewards accounts that don't stop. None of that tells you whether the content is actually good. It tells you the account didn't go dark for three weeks while someone was busy. Those are different things, and it's easy to present one as the other.
Muse's first twelve days are the app equivalent of that scheduled feed. Meta has the distribution machine, perfected on Threads, which grew past 500 million users largely through cross-promotion inside Instagram and Facebook rather than through Threads being self-evidently the best product in its category. Comparing Muse's download curve to ChatGPT's is comparing a company with a captive audience of billions to a company that had almost none. The comparison the article makes is careful and clearly labelled as such. What people do with the number afterwards usually isn't.
A launch metric tells you about the launch, not the thing itself
I had a project a while back that looked simple from the outside for exactly the same reason a download chart looks simple. It was a printed form filled in by hand, and I assumed it would be one of the easier documents to automate. The layout was fixed. Every field was labelled. Every page had the same shape. The structure told me exactly where to look on the page. What it never told me was what was actually written inside the boxes, and the handwriting turned out to be the entire problem. The form looked solved before I'd solved anything.
Download counts and daily active users are the printed structure of an app, not the handwriting inside it. They tell you people opened it. They don't tell you what happened next: whether people came back weeks later, whether they used it for something they'd have paid attention to anyway, whether it displaced a habit or just added a notification to a phone that already had Instagram, Facebook and WhatsApp open. Apptopia is careful to say these are third-party estimates and can only be read as a general ballpark. That caveat is doing a lot of work that most retellings of this story drop.
What I'd actually want to know before calling Muse a hit
- Retention past the first month, not the first twelve days, since a novelty install curve says nothing about habit
- Usage per session, not just daily active users, since opening an app for a few seconds counts the same as a long session in most DAU figures
- Whether people are choosing Muse for a task or landing on it because a notification from Instagram put it in front of them
- Whether the cross-promotion keeps propping up the numbers once the initial push from Meta's other apps slows down
None of that is answerable from what's public yet, and to its credit the article doesn't pretend otherwise. What bothers me is how quickly a launch comparison like this gets flattened into a verdict on which product is better, when what it actually measures is which company already owned the audience before the app existed. Muse might turn out to be genuinely useful. I have no idea, and neither does anyone quoting these numbers this week. What I do know is that a strong first fortnight built on someone else's existing platforms is not the same evidence as a strong first fortnight built from nothing, and treating them as interchangeable is exactly the kind of shortcut that gets a business to trust the wrong signal.
