The Maze: Meta, the company behind Facebook and Instagram, says more than 10 million people in the United States and Canada have set up Muse, its task-completing artificial intelligence assistant. Its October 21 earnings disclosures show early shopping activity. They do not show how many of those people became buyers, or what merchants earned from them.
Source: Meta. September launch demonstration: Muse presents a stroller product detail and a purchase button; this is not evidence of October transaction volume.
The Details:
A setup is the beginning of the measurement. Chief executive Mark Zuckerberg said about 25% of people who set up Muse use it on a given day. That measures returning use across tasks, including shopping and personal productivity. It does not measure daily purchases. Meta also discussed purchase values above $200 within an engaged shopping cohort. The public figures do not supply a complete buyer denominator, order count or measurement period. The commerce discussion includes travel, so the dollar figure should not be read as a typical retail basket. Multiplying the headline audience by a shopping percentage would manufacture a sales estimate the company has not supplied.
The assistant can do work between discovery and payment. A shopper messages Muse a task rather than opening every store and comparing options manually. Meta's product explanation says it can research, open a browser, fill forms and work through connected services. It returns for the person's approval before a purchase. One payment route described at launch is Link, Stripe's checkout service, which can provide a one-use card number. The distinction matters commercially: researching a product, connecting another service and completing an approved purchase are separate actions. A useful recommendation may influence an order even when the assistant does not complete the checkout itself.
The transaction business is still being built. In Meta's official follow-up call, finance chief Susan Li said subscriptions are currently Muse's only monetization. Charging for facilitated transactions requires more users and more connected business partners first. Meta wants to participate more deeply when someone already intends to buy, rather than only showing an advertisement that creates interest. But it has not disclosed a transaction-fee rate, a start date or merchant returns. Li also said Muse remains available only in the US and Canada; serving more countries sustainably depends partly on reducing its computing costs.
MarketMaze Take: Muse has earned a place on merchants' measurement plans. It has not yet earned a sales forecast based on every person who signed up. Meta's own distinction between present subscription income and future transaction fees puts the commercial opportunity in perspective: distribution is arriving before a fully priced distribution business.
The early customer mix can flatter the opportunity. Li said US Muse adopters skew toward Meta's highest-revenue users and do not represent the wider US or worldwide population. That may make the first customers attractive. It also means their shopping behavior cannot establish what a broader audience will spend.
Merchants elsewhere can learn from the mechanism without assuming local access. An assistant that researches and prepares purchases could become another intermediary between a seller and its customer. The useful test is completed orders and profit after any future fee, compared with business the merchant would have received anyway. Meta has not yet supplied that comparison. The audience is evidence of reach; the merchant economics still need their own receipt.
