Meta to launch in-feed shopping Platform
4 September 2026
Paul Adams – Chief Growth Officer
Meta’s agentic shopping agent, codenamed Hatch, is reportedly weeks from launch. ‘Hatch’ is an agent designed to complete the entire purchase journey inside the Instagram/Facebook feed, with no click-through required. A consumer tags a product in Reels or messages the agent directly, and from there Hatch takes over: comparing options, reading reviews, applying preferences it has learned, and transacting, all without the user ever opening a browser or a third-party app.
Set against the field, Hatch’s position is distinctive rather than dominant. Amazon’s retail agents still win on fulfilment certainty, review depth and one-click logistics, but they only ever see Amazon’s own shelf. ChatGPT’s shopping layer is platform-agnostic and increasingly comparison-led, yet it has none of Meta’s owned inventory of native product tags, creator content and billions of daily social sessions to draw on. What Hatch has instead is context: the ability to act on a purchase intent the moment it forms, inside content the user already trusts. What it lacks is Instagram Shopping’s long-standing weakness — thinner merchant density than Amazon, and a commercial interest of its own that a neutral assistant like ChatGPT doesn’t carry.
Then there’s the price of admission. Premium subscription tiers up to $199.99 a month, a higher number than any current AI assistant subscription, is under consideration. This is a clear signal that Meta expects the agent to earn its keep through completed transactions and saved time, not chat access alone.
Similar to other shopping agents, discoverability in Hatch is partly determined by product data e.g. clean titles, structured attributes and contextual descriptions. However, because this agent is embedded in a social platform, creator-tagged content is a trust signal. Therefore, creator partnerships with a strong engagement history that are systematically tagged, catalogued and tied to a creator’s recommendation post are a valuable data asset.