Semi-Automated AI Cross-Border E-commerce (Ozon Blue-Ocean Store Group)

01The Full Story: From Unemployed to RMB 50K/mo With a Store Group

Bilibili creator 薛优秀很努力 tells this as a comeback story from unemployment, but unpack it and it's far more than motivational content — it's a replicable methodology of “pick the right platform + cut costs with AI + drive traffic from off-platform.” He didn't start on Taobao or Amazon, where everyone is fighting for scraps. He worked backwards, hunting for a platform that hadn't been ground down yet, and landed on Russia's Ozon. His reasoning is hard-nosed: Ozon has roughly 60.5M buyers against just 350K sellers. He invented a “blue-ocean score” metric (buyers ÷ sellers) that comes out at 172 for Ozon, versus 10 for Taobao on the same basis — meaning the same unit of demand, spread across sellers, lands more than ten times as thickly on each Ozon seller. With the platform locked in, he used AI to semi-automate the two most labor- and cash-intensive parts of traditional e-commerce — product selection and image production. He has AI scroll web pages to filter for new listings and screen products, then hands the chosen products to AI to generate the hero image and detail page, cutting out graphic designers who charge hundreds of RMB per image. He lists through Ozon's zero-deposit channel, and instead of paid on-platform ads or fake orders, he uses off-platform index ads on a “pay only after an order lands” basis to get the first sales moving. Once the single-store model worked, he hired operators to clone it into a store group, then expanded sideways into Pinduoduo virtual-goods stores and Taobao custom software, stacking a multi-platform two-layer traffic matrix. His self-reported result: RMB 400K+/mo in combined GMV across the stores, and roughly RMB 27K/mo in profit from the two Ozon stores.