Meta’s Muse shopping figures: what merchants can infer

Meta says more than 10 million people in the US and Canada set up Muse personal agents in roughly six weeks. The figure, disclosed on its 21 October earnings call, measures setups. It does not mean 10 million shoppers or completed purchases.

Keep the shopping figures in context

In the official earnings transcript, Mark Zuckerberg said around 25% of people who set up Muse use it on a given day. CFO Susan Li separately said shopping was used by more than 25% of weekly active users.

Zuckerberg described average purchases above $200 and later clarified that the commerce measure includes travel. That is not a Shopify retail basket benchmark. The figures are Meta’s disclosures, not an independent measurement of merchants’ incremental sales.

Transaction fees are a plan, not a published rate

Zuckerberg said Meta plans commerce tools that earn fees on facilitated transactions, expected to be paid by businesses. He also described subscriptions for heavier usage and business features. The call did not provide a merchant transaction rate or start date.

Our recommendation is to keep prospective fees as an explicit unknown in a channel model. Do not enter an assumed percentage as if it were a contractual rate. When terms become available, assess the total cost alongside fulfilment, returns and customer support, then test whether attributable orders contribute profit.

What these numbers cannot tell your store

Do not multiply the setup count, shopping share and average purchase figure to estimate gross merchandise value. They have different denominators and measurement periods, and a shopping interaction need not be a completed purchase. The travel component also makes the purchase average a poor substitute for a particular store’s expected basket.

A more useful merchant test starts with evidence from your own channel: completed orders, cancellations, returns, fulfilment costs and the source of attribution. Keep observed results separate from platform-wide adoption claims. A product being mentioned by an assistant is a different outcome from a paid order.

For the operator side of the product, see our existing report on Muse for Small Business and its Shopify connection. Using an assistant to help run a store and receiving purchases from shoppers’ agents are separate use cases.

The new disclosure is worth watching because it provides dated evidence of shopping activity and a proposed commercial model. It does not establish a predictable sales lift for an individual merchant.

The transcript is available from Meta’s investor events page, alongside the earnings release and presentation.