Meta Muse builds apps from a sentence. The prize is your Instagram saves

Need a holiday gift tracker? Used to be: find an app, create an account, start a subscription. With Meta now: one sentence to Muse.

Sounds like tinkering. It isn't.

What Meta launched

On November 6, Meta introduced "Spaces" in its Muse AI app. You describe an idea, Muse builds a small app that stays in Muse. You can then add features, change the look or tweak how it works, all by prompting. Spaces can be pinned to the sidebar or main tabs, hidden, or scheduled to appear at set times, say a weekly planner on workdays that turns into weekend plans on Saturday.

With connected apps, Meta says spaces update in the background. Its examples: a shopping hub pulling together gift ideas, saved Instagram wishlists, and Marketplace listings with buyer messages. An order tracker for small businesses: new orders, customer messages, what ships today. A client hub for freelancers with projects, deadlines and invoices. Sharing with others is being explored, no timeline. Meta is also rolling out specialized models trained on learnings from Muse Wonder.

Spaces is available wherever Muse is. Per reports, the Muse app still hasn't launched in Europe.

The app isn't the product. The wishlist is

What I find interesting: everyone will talk about "vibe coding for everyone." The real move is in the shopping example. Saved Instagram posts, Marketplace and gift ideas land in an interface Meta owns.

Which means purchase intent isn't just collected anymore, it's organized and managed. Inside Meta. Before anyone opens a search engine, before they hit your site. Lower funnel, inside the assistant.

For brands, that changes what an Instagram save is worth. So far a soft engagement metric. Soon, maybe, an entry on a shopping list the assistant maintains.

Small SaaS has a pricing problem

The second point hits software makers. Habit trackers, wedding planners, simple client portals for freelancers: if that's one sentence in a free app, nobody pays $9 a month for it. The interface becomes a commodity. What survives is data, service or community Muse can't rebuild.

To me it's the same pattern as Adobe and SeenThis this week, from the other side. There, specialist software moved into chat. Here, chat builds the specialist software itself.

What I'd do now

• Take saves seriously. Report saves per product post separately, not just likes and reach. That's the number feeding Meta's shopping hub.
• Secure your owned asset. Anyone bookmarking your product should have an alternative with you: wishlist with email, price alert, newsletter. Before the list lives entirely at Meta.
• Stress-test your product. Selling software? Could someone rebuild the core in one sentence in Muse? If so, your value has to sit in data and service, not the UI.

Bottom line: Meta isn't building apps. It's building the place where purchase intent gets managed, and the apps are the bait.