Recent news from China regarding Alibaba.com’s Qwen AI model has often been accompanied by sensationalist headlines claiming “economic chaos,” glitches giving away free food, crazy restaurant queues, and a system crash that supposedly marked the failure of the Chinese giant’s initiative.
Analyzing the facts, the reality is quite different: this was not an error, but one of the most aggressive Customer Acquisition operations ever recorded in Food Retail. It is an event that anticipates dynamics destined to impact the P&L of the sector in Europe as well.
The Transition: From Chatbot to “Buyer”
Between February 6 and 8, 2026, during the Lunar New Year, we witnessed a technological paradigm shift: AI stopped being a conversational tool and became a transactional agent. This is no longer a series of sporadic episodes, but a clear signal of how heavily companies are pushing for AI to replace the buyer’s choice—not just the user’s interaction.
In this specific launch, upon receiving a user’s voice command (“Order me a coffee from Luckin Coffee”), the system did not provide suggestions; it executed the selection, checkout, and payment. The elimination of decision-making friction produced impressive volumes: according to the Chinese press, over 10 million orders were recorded in just 9 hours.
Coupon Engineering: Why It Wasn’t “Free”
It is crucial to dismantle the narrative of the “free meal” to understand the economic sustainability of the operation. The mechanism was based on a “Free Order Card” with strict rules: 1 Coupon = 1 Order (no infinite freebies); a maximum cap of 25 RMB (about €3.20) per order; a symbolic payment of 0.01 RMB to close the fiscal transaction; and a viral loop requiring users to invite others to accumulate more credit.
Alibaba was not doing charity; it invested CAPEX to force the adoption of a new purchasing behavior, involving high-frequency players like Luckin Coffee, Mixue Bingcheng, and Bawang Chaji.
Cost Analysis: Strategic Budget vs. Burn Rate
To properly size the event, we must distinguish between strategic plans and actual out-of-pocket costs. Sources indicate an overall incentive plan of 30 billion RMB (approximately €3.8 billion) for the long term. However, isolating the 9 hours of the event and applying simple arithmetic to the 10 million declared orders, we get an estimated burn rate of about €32 million consumed in half a day. This figure was covered by the platform’s promotional budget to validate the AI model as the sole intermediary and acquire a raft of new clients through existing users.
The Italian Scenario: Data Beyond the Hype
It would be a strategic mistake to dismiss these events as distant Asian dynamics. According to the B2C eCommerce Observatory (Politecnico di Milano), the groundwork in Italy is already primed for a similar transformation. In 2025, the online Food & Grocery sector in Italy reached approximately €4.9 billion. The crucial data point is penetration: Food Delivery accounts for 46% of this total and now reaches 76% of the Italian population. The logistical infrastructure is laid down and capillary. The real question is: how is the business model running on this network changing?
The End of Pure Dark Stores and Structural Integration
A careful analysis reveals that the failure of “pure Dark Store” business models (such as Gorillas or Getir) has not erased the demand for convenience. That demand has shifted to robust platforms evolving into integrated retail ecosystems.
- Glovo and Mass Retail: The operational partnership launched in February 2026 with Unes Supermercati is a fundamental evolution. Managing urban hubs with over 3,800 SKUs means transforming delivery into a distributed supermarket.
- Deliveroo: Consolidating hybrid models by working with retail partners to cover rapid delivery without fully shouldering the real estate risk of warehouses—a decidedly more sustainable model for the P&L.
- Just Eat Italy: The introduction of AI assistants for decision support and expansion into high-margin categories like pharmacy indicate a desire to become the app for daily necessities.
Forecasts: When AI Becomes the P&L Gatekeeper
The Chinese Qwen case allows us to draw a clear operational forecast for the European QSR market. Today, Europe features an AI that suggests; China has showcased an AI that executes. When this technological “layer” arrives here, the consequences will be profound:
- Brand Legibility: If the user interface disappears and an algorithm is doing the buying, visual marketing loses relevance to structured data. A restaurant’s menu must be optimized to be the “best answer” to an AI’s query. Those without structured data will become invisible.
- The Operational Stress-Test: This is a major critical issue. AI can generate violent, concentrated order spikes to maximize economic advantage. If an AI funnels 500 orders in 15 minutes to a QSR outlet to exploit a promo, current kitchens will crash. Layouts will need to be re-engineered to handle automated order “bursts”.
- Delivery as an Operating System: We are transitioning from delivery as an accessory sales channel to an Urban Operating System. Platforms will control not only logistics but the initial purchasing intent.
The risk for Italian retail is not technological innovation itself, but the speed of adaptation. If the end customer delegates the purchase to software, the real customer to convince is no longer the person eating the burger, but the algorithm deciding where to buy it.
— Michele Ardoni
https://www.linkedin.com/pulse/lai-conversazionale-di-alibaba-sposta-10-milioni-ordini-ardoni-f1fdf