Meta’s Muse Could Upend All Kinds of Businesses. It Hasn’t Yet

Executives at roughly 50 companies described commerce through AI agents as negligible so far: DoorDash cited orders in the “hundreds,” Hilton called bookings “insignificant,” and Eternal estimated activity as “close to zero.” Meta said more than 10 million people in the US and Canada had activated Muse, with over 25% using it for shopping; Chipotle reported Muse’s debut lifted agentic orders by a single-digit percentage. Companies are beginning to optimize for agents, while weighing transaction fees, security risks, and possible pressure on margins; Amazon and Insurify have also shifted their approaches to Muse.

Analysis

The near-term signal is not a demand shock; it is a bargaining-power contest over discovery, checkout, and customer data. Meta’s upside depends on agents producing incremental purchases that can support transaction fees without cannibalizing higher-value ad inventory. If agents mainly reroute existing demand, Meta could gain a toll booth while retailers lose attribution and pricing leverage—a transfer of economics, not necessarily market growth.

Retailers with strong first-party loyalty, differentiated products, or controlled checkout have more leverage than merchants whose offers are easily compared. Loyalty rules, inventory accuracy, and authentication are practical barriers to agent conversion; fixing them may improve ordinary digital conversion too, but could also make merchants more dependent on a small set of agent platforms. Amazon’s willingness to accommodate outside agents, if sustained, would be a rational defense of purchase intent and fulfillment relevance, not evidence that its own agent has lost.

Over days to weeks, expect narrative volatility rather than measurable earnings impact. Over 1–3 months, watch retailer disclosures for agent-attributed conversion, incremental versus shifted orders, and any fees or data-sharing terms. Over 6–18 months, the structural risk is platform concentration and lower merchant control of customer relationships. Security incidents, inaccurate offers, or low repeat usage could stall adoption; evidence of rising agent-originated sales without higher acquisition costs would falsify the cautious view.

Key Decisions for Investors

• Avoid buying META solely on the agent-commerce narrative until management provides evidence of repeat usage, incremental transaction volume, and fee economics. Track whether commerce monetization adds to or substitutes for ad revenue.
• No broad long/short retail trade is warranted from current evidence. Favor merchants with differentiated assortment, loyalty, and direct customer relationships over easily substitutable offerings if agent-mediated comparison scales.
• Treat AMZN’s posture toward third-party agents as a strategic watch item: continued access could protect its role in fulfillment and conversion, while renewed blocking or loss of purchase intent would weaken that defense. Reassess on concrete policy changes or disclosed traffic trends.
• Monitor agent-attributed orders, conversion versus baseline, merchant fees, repeat usage, and security-related checkout failures. Rising gross orders without proof of incrementality or acceptable economics is not a bullish catalyst.