ReFiBuy 上线 Meta Muse 支持，购物成 Muse 首要用途

ReFiBuy, the Agentic Commerce Optimization (ACO) platform, today announced support for Meta Muse, now in early access, bringing its closed-loop, agentic product enrichment to brands and retailers selling through Meta’s new personal agent. The platform helps commerce teams find the product data gaps that make their products hard for Muse to recommend, approve enrichments that close those gaps, and monitor how their products show up in Muse.

During its third-quarter earnings call on the 21st of October, Meta said more than 10 million people in the U.S. and Canada set up Muse within six weeks of launch and that shopping is the most common use. Meta added that more than a quarter of weekly users shop with Muse and that the average purchase, travel included, is over $200.

Muse has deep context on the shopper’s plans and preferences from past conversations and connected apps. Most product catalogs have not kept up. What a shopper wants to know (who a product is for, the occasion it suits, how it fits) usually sits with merchandisers and in reviews, and little of it reaches the product data Muse reads. When the data cannot answer, Muse is more likely to recommend a competing product whose data can.

In the first 60 days, Muse has captured the imagination of not only consumers, but brands and retailers alike. Heading into the holidays, there are millions of shoppers leaning into their personal agents. We’re excited to launch comprehensive support for Muse to help merchants meet those customers where they are.

– Scot Wingo, co-founder and CEO, ReFiBuy

ReFiBuy’s research, published by Scot Wingo on Retailgentic on the 13th of October, found that Muse pulls from Meta’s own product catalog, Shopify stores, and product pages across the web, and that products in Meta AI’s top three results also appeared in Muse 89% of the time.

In ReFiBuy’s 5 Levels of Product Data Enrichment, product data built for search and marketplace feeds is Level 1, structured basics. A personal agent matching a product to a specific person needs at least Level 2, conversational attributes, and ideally Level 3, shopper context.

ReFiBuy’s Commerce Intelligence Engine closes that gap SKU by SKU with specialist agents. A content agent drafts descriptions and product Q&A, and an attribute agent fills in the structured details shopping agents match on, down to each variant. A reviewer agent checks every suggestion before a person sees it, and each one arrives with its evidence. Commerce teams set the review rules that decide what goes live, and the agents remember what was accepted and rejected.

For Shopify brands, approved enrichments sync to the store, so the same details can appear on product pages and in their structured data, where agents browsing the web look.

Monitoring closes the loop. ReFiBuy monitors Muse results both directly and through Meta AI, whose product catalog Muse heavily relies on. It tracks whether a brand’s products appear for the requests shoppers make, how they rank against competitors, and whether the details shown are accurate, then turns the misses into the next round of enrichment for the team to approve. The same loop already runs for ChatGPT, Microsoft Copilot, Google AI Mode, and Alexa for Shopping on Amazon.

Black Friday, on the 27th of November, will be Muse’s first, and ReFiBuy recommends bringing the top 5% to 10% of SKUs to shopper-context enrichment before then.

Early access is open now to brands and retailers selling in the U.S. and Canada, where Muse is available. Participants work with ReFiBuy to review gaps, approve the first enrichments, and set up Muse monitoring for their priority products. Brands and retailers can request early access.