When China’s latest AI arms race landed on bubble tea, the consequences played out in familiar, chaotic fashion: sudden order surges, temporary store closures, delivery delays and reels of order slips strewn across shop floors. On February 6 Alibaba’s Qianwen app launched a headline‑grabbing promotion — a RMB 3 billion (30亿元) “Spring Festival” subsidy that offered free or heavily discounted drinks to attract new users — and its effects were immediate and tangible for small retailers and couriers.
Journalists who visited tea shops in multiple commercial districts found staff overwhelmed. One store reported more than 100 low‑price orders routed through Taobao Flash Sale in a single day; another completed nearly 150 orders at an effective price of RMB 1–2. A staffer at a different outlet said printed tickets had “circled twice” on the floor at the height of the crush; some shops temporarily delisted themselves from flash‑sale channels to catch up.
The user experience was mixed. Qianwen topped Apple’s app‑store charts after the promotion went live, and Alibaba says the campaign drove more than 10 million AI orders within nine hours and prompted over 30 million “help me buy” commands. But the surge also caused system slowdowns and payment failures for some users, while others reported waits of several hours before their drinks arrived or were prepared at all.
For merchants the economics were unclear at first. Staff told reporters they thought platforms carried the bulk of the subsidy burden; Alibaba later confirmed that Qianwen itself funded the discounts and that merchants were not required to contribute extra. Still, small retailers felt the operational strain: low margins on subsidized orders made the volume a logistical headache rather than a profitable windfall.
Technically, Qianwen’s product already shows signs of refinement. The interface allows users to swipe across a broader set of brand and product options and appears to incorporate users’ past preferences when recommending outlets via Taobao Flash Sale. That suggests closer integration between the AI front end and Alibaba’s commerce ecosystems, enabling personalised, speedy fulfilment — at least when systems and shop capacity hold.
But the AI didn’t always match the more exploratory instincts of customers. Several users said Qianwen favoured familiar, long‑standing bestsellers over newly launched products, and one user who pressed for novel recommendations received irrelevant suggestions before the app “collapsed” into unrelated categories. For shoppers who like to browse and discover, the automated flow felt slower or less satisfying than manual search.
The bubble‑tea episode neatly distills the broader strategic calculus of China’s tech giants. Cheap, social goods like drinks are ideal vehicles for mass promotions: they carry low per‑unit cost, trigger viral sharing and are a controlled way to onboard users into a new AI experience. Alibaba’s splashy subsidy both tests Qianwen’s technical capacity to transact at scale and pressures rivals such as Meituan, Tencent and ByteDance to respond in kind, re‑splitting customer attention and industry margins.
Longer term, the event raises several questions. Can AI‑led discovery be monetised without recurring deep subsidies? Will merchants demand compensatory fees, stricter delivery windows, or regulatory relief to prevent such operational shocks? And how comfortably can platforms stitch together personalised data flows across shopping and delivery services without running into privacy or competition concerns? For now, the blitz has planted AI shopping firmly in many consumers’ first‑time experiences, but whether it will translate into durable change for commerce and delivery ecosystems remains uncertain.