Delivered in a Flash: Meta Tests AI Shopping
OMR Team12/31/2025
Mark Zuckerberg’s tech giant integrates personalized product search into its chatbot to challenge Google and OpenAI
A few weeks after initial hints were dropped, Meta's AI shopping bot has begun appearing for the first users accessing Meta AI in the USA via web browsers. What does Zuckerberg’s company have to offer against competitors like OpenAI and Google?
Seamless Product Discovery via Chat
Meta has begun testing a new shopping research feature within its Meta AI chatbot. The goal is to integrate product search directly into the chat flow and assist users in their purchasing decisions. This move puts Meta in direct competition with OpenAI (ChatGPT) and Google (Gemini), both of which are increasingly focusing on e-commerce capabilities.
From Vision to Reality: The Personal Superintelligence
Mark Zuckerberg had already prepared the market for this step earlier this year. During an earnings call in January, he spoke of a vision for a "personal superintelligence." He announced that Meta would deliver products offering a "uniquely personal experience" based on users' interests, relationships, and history. His ultimate goal: AI agents that find exactly the right product from Meta’s massive catalog.
A Tailored Shopping Assistant in Action
The tool functions as an interactive shopping consultant. The chatbot displays a gallery of images including details on brand, price, and website. A journalist from Bloomberg News, who was able to test the tool, received a product carousel featuring women’s puffer jackets in New York after searching for "puffer jackets."
Meta utilizes existing data, such as location and inferred gender (derived from the user's name), to provide relevant results. Alongside images, the AI provides brief, bulleted explanations as to why specific products are being recommended. Direct checkout is not yet possible; users are currently redirected to the merchants' online shops via links.
Meta’s Strategic Advantage: The Social Graph
Meta possesses a unique infrastructure. Once the shopping assistant is fully integrated into Facebook, Instagram, Messenger, and WhatsApp, vendors could potentially reach billions of people worldwide. By accessing the vast catalog of businesses already present on Meta’s platforms, the offering is likely to be extensive from day one.
Even though Meta is not the first to launch such a tool, the platform holds significant advantages over Google and OpenAI:
- The Social Graph: Recommendations are based on specific interests and behavior within social networks, allowing for deeper personalization.
- Native Environment: Unlike ChatGPT, Meta AI is embedded directly where people already spend their digital leisure time and communicate.
- Monetization: The tight integration with the product catalogs of existing advertisers should allow for more effective monetization than at Google, where the bot must crawl the general web for matching products.
The Adoption Challenge: Will Users Actually Engage?
The big question remains: Will anyone actually use it? New York e-commerce entrepreneur Juozas Kaziukėnas, who was among the first to discover the new Meta AI feature and report on it via LinkedIn, hit a nerve: he claimed he doesn't know anyone who uses Meta AI. However, he followed up with a "glass-half-full" perspective: "Meta is spending a lot of money to stay relevant in the AI race."