Muse's 10 Million Downloads Are a Retail Warning Shot

Muse's 10 Million Downloads Are a Retail Warning Shot

Ten million downloads in 40 days

Meta's personal agent Muse has become one of the fastest-adopted apps in history. Sensor Tower data puts total downloads at 10.9 million since the September 8 launch, with 10.2 million in the United States. Only 11 other apps have ever crossed 10 million U.S. downloads in their first 40 days, the most recent being Meta's Threads in July 2023, and most of the rest were mobile games.

Retention beats every rival chatbot

The stickiness numbers are arguably more important than the downloads. Sensor Tower estimates Muse reached 2.5 million daily active users for the first time on October 16, up 15% week over week. It retained 30% of users after seven days and 23% after 14 days, against averages of 19% and 16% for ChatGPT, Claude and Grok. Over the same stretch, U.S. downloads of ChatGPT fell 25% from the prior 30-day period.

Zuckerberg confirms 10 million agents on earnings call

CEO Mark Zuckerberg confirmed the scale on the October 21 third-quarter earnings call, saying more than 10 million people in the United States and Canada had set up Muse agents since September 8. About 25% of those who set it up now use it daily, which he said is helping retention. Meta reported third-quarter revenue of $64.6 billion, up 26% year over year and ahead of consensus near $63.3 billion.

Shopping is the killer app, at over $200 a basket

Shopping emerged as one of Muse's top activities. Zuckerberg said more than a quarter of engaged users who shop through the agent had an average purchase size above $200. Financial-technology firm Plaid, whose connector links the agent to buyers' bank accounts, told Meta it had never seen activity like this on any other service's connector before.

Demand blew past Stripe's limits in week one

Volumes overshot every internal forecast. Zuckerberg said Meta had to work with payments processor Stripe to raise credit limits in the first week, because purchase volumes were above what anyone expected. Using Muse is free for consumers unless they want more tokens or more advanced models; Meta plans to charge businesses a fee when the agent completes a sale for them.

Why it worked: errands, not chat

Muse won because it runs errands instead of chatting. It books travel, fills in forms, monitors home cameras and shops across sites with user approval, paying through one-time cards via Stripe's Link service. The app sat atop Apple's U.S. free chart, a run that sent Meta shares up about 11% in a single session. Amazon then cut Muse off from buying on Amazon.com, calling it an unauthorized agent, a blockade that only confirmed retailers see agents as a fight over the customer relationship.

Wall Street splits between fear and vindication

The market reaction cuts both ways. CNBC commentator Jim Cramer described a "consumer inertia" sell-off, the theory that agents which compare prices and cancel subscriptions on command threaten businesses built on customers not bothering. Asset manager Invesco drew the opposite lesson, arguing Muse's early success shows the structural growth story behind the bull market remains intact even as oil prices and interest rates climb.

A hit product, a falling stock

The paradox: a genuine consumer hit met a falling share price. Meta narrowed its 2026 capital-expenditure outlook to $135 billion to $145 billion, reported Reality Labs revenue of $443 million against a $4.13 billion operating loss, and guided to $18 billion to $19 billion of Reality Labs losses for 2026, narrowing toward about $15 billion in 2027. Investors staring at that spending plan sent the shares down roughly 6% to 7% after earnings, even with fourth-quarter revenue guidance of $73 billion to $76 billion beating the Street.

What comes next: monetization and access

Two tests decide whether this demand shock turns into a durable business. The first is monetization: business fees on agent-completed sales, plus checkout rails through partners such as Walmart, Best Buy, Gap, Sephora and Wayfair. The second is access: Amazon's door is shut over authorization, identity and credential concerns, and every retailer's choice of camp will set the boundaries of agentic shopping. The demand question looks answered; the business-model question is next.