AI has made Double 11 in 2025 smarter. Now it must make it sustainble

READ FULL ARTICLE ON Campaign Asia

By Minnie Wang

Published November 12, 2025

The Double 11 shopping festival turns 17 this year and what was once a giddy celebration of flash sales, marathon discounts, even emotional fulfilment is starting to feel more like a reality check.

What began as a retail arms race for gross merchandise value (GMV) has now morphed into something much bigger and arguably more complex. AI is everywhere, yes. But so are questions about what growth really means when speed, service, and sustainability are the new currencies of competition.

“Platforms still chase GMV,” says Subramania Bhatt, founder and CEO of China Trading Desk, “but the winning brands are balancing sharp, transparent price points with credible specification, cultural storytelling and service certainty.”



That tension between efficiency and emotional engagement defines this year’s Double 11. As Bhatt notes, “Xiaohongshu's deeper tie-in with Alibaba formalises the journey from lifestyle content to cart, while one-hour 'instant retail' (Taobao Flash vs Meituan) turns speed itself into a premium brand cue in categories like beauty and gifting.”



And the numbers back it up. Taobao’s daily active users climbed from 375 million in April to 437 million in September, a 12% year-on-year jump. Its Flash Sale service generated over 100 million orders from new users, with nearly 20,000 food & beverage brands and 800 non-food labels doubling transactions during the festival.



“Emotional resonance has become a key factor in brand differentiation,” adds Moqian Sun, founder of The Harvest.

Moqian refers to the price wars in China's delivery platforms and the algorithmic recommendations that have flattened much of the competition. So, brands that connect emotionally rather than compete transactionally will cut through. The 2025 Gen Z Double 11 Consumption Report supports this: travel and experience-led spending now lead all “emotional consumption” categories, while purchases of digital services, in-game assets, and cultural collectables that cater to emotional and experiential needs witness a sharp spike.

The shift also explains why platforms like Xiaohongshu have become powerful conversion engines. As Sun notes, younger buyers “no longer see Double 11 as a shopping event, but as a social ritual—a moment to express taste, participate in trends, and reward themselves.”

AI is everywhere. Is it changing marketing or an assistant? 

For the first time in a decade, fair to say that Double 11 has been an open-lab experiment for AI-powered commerce. Beneath the hype, there's a more grounded debate: is AI truly changing e-commerce, or just catching up to consumer behaviour?



“This year’s Double 11 in China looks different, because AI is everywhere,” says Miro Li, founder of Double V. Consulting on LinkedIn. “Shoppers are now turning to AI tools to search, compare, and even calculate the smartest way to shop.

“Taobao introduced new features like AI Universal Search and AI Help Me Choose, allowing users to chat naturally with the app and ask things like ‘Which sneakers are better for winter runs?’ or ‘Can I still use this coupon if I check out after midnight?’ The AI then compares prices, reviews, and specs, offering suggestions in seconds,” she continues. 



Even on social media, AI has become a verb. On Xiaohongshu, the term “AI 凑单” literally means AI cart optimisation. Li talks about how shoppers share screenshots of prompts given to Deepseek or Taobao’s own AI and ask it to build the most cost-efficient combination of products and coupons to unlock full discounts. One Xiaohongshu post showing the 'AI cart hack' and racked up over 3,000 likes.

For years, Double 11 has been criticised for confusing discount rules and complex math. But now Li observes, “AI has quietly become the new shopping assistant, translating this chaos into clarity.” 



Still, not everyone sees this as revolutionary. Yaling Jiang, founder of ApertureChina and the Following the Yuan newsletter, remains unimpressed. “The most significant way that AI is being used by e-commerce this year is through helping consumers make decisions. But it’s not groundbreaking, and they’re merely catching up after Deepseek’s breakout success at the beginning of the year. I think they’re only matching consumer expectations in doing so.”

Alibaba, for its part, insists that the AI push is more than a cosmetic upgrade. The group says 2025 marks its first large-scale integration of large language models (LLMs) into Taobao and Tmall’s search and recommendation systems. Its AI Business Advisor tool has already generated five million analytical reports since October, it cuts merchant workloads by 30% and boosts efficiency by 150%. Meanwhile, its AIGC suite has produced 150 million visual assets and saves an estimated RMB 40 million (US$5.6 million) a day.


On the consumer front, Taobao’s Dianxiaomi assistant has handled 300 million queries, while merchants using AI service tools reported a 30% lift in conversion rates.

“AI-powered customer service, virtual shopping assistants, and recommendation systems are making the shopping experience smoother and more interactive,” says Sun.



Bhatt adds that even user reviews are getting algorithmic makeovers. “Reviews are now auto-clustered into ‘what people like/avoid’ blocks by persona and skin type, climate, device ecosystem. This surfaces credible visuals and pros/cons that deflect pre-purchase chat volume and boost assisted conversion.”



The big question then: is AI driving more value, or just more volume?



Jiang is sceptical. “We are not seeing as many discounts as in previous years, while consumer fatigue about shopping holidays has continued to grow. It’s likely that there will be a lot more volume…But I suspect general GMV will decline compared with the same time window.”



Bhatt argues that success needs a new metric: “Let models decide who gets direct discounts vs bundles vs trade-in offers based on expected 90-day value. Judge success on incremental margin dollars and re-purchase rate, not raw GMV.”



“The biggest opportunity,” he adds, “is to make AI the front door to the brand, not just a backstage efficiency tool.”

For Lucy Zhang, chief relationship officer at Omnicom Media Group China, AI’s true potential lies not in automation, but in connection.



“The greatest unrealised potential of AI in e-commerce marketing is not about improving efficiency, but about redefining the fundamental relationship between brands and consumers,” she says.



Zhang believes AI must evolve into “a centre of emotional intelligence,” breaking silos across browsing, service, and repurchase so that “every interaction extends the brand narrative.”



“The future of e-commerce will no longer be a battlefield of algorithmic pushes,” she adds, “but a real-time alignment between brand identity and consumer context.”

The sustainability imperative

As AI makes commerce faster, more personalised and frictionless, it’s also accelerating consumption at an unprecedented scale. Every year, the world’s biggest shopping festival generates a parallel surge in packaging, logistics, and carbon output—the uncomfortable side-effect of its success.

In 2023, the amount of waste generated from express delivery packaging reached approximately 20 million tons. In major cities such as Beijing, Shanghai, and Guangzhou, this type of waste accounted for over 93% of the increase in household garbage. Currently, the overall recycling rate for express delivery packaging in China is less than 20%, according to an expert.

During the Double 11 Shopping Festival in November 2024, the number of packages delivered on November 11 surged to 701 million. The State Post Bureau projected that parcel volume this year could exceed 190 billion, which would represent a doubling compared to 2020, setting an all-time record. And now, with AI optimising carts, each “instant” sale is multiplying its carbon footprint.

Both Alibaba and JD.com have long positioned themselves as green logistics pioneers. Cainiao introduced reusable packaging, electronic waybills, and smart routing to cut emissions, while JD Logistics’ Green Stream programme promotes large-scale recyclable cold-chain boxes. But for now, these measures are chasing a rising tide.



Bhatt believes visibility, not just intent, is what’s missing. “‘Green’ needs to be surfaced as a consumer choice, with numbers,” he argues. “Double 11 should turn ESG from CSR copy into performance: lower unit costs in logistics, higher preference and more repeat.”



He sketches what that might look like: “Carbon-label baskets at checkout, default consolidated shipping toggles with clear ETAs, ranked low-carbon or refillable SKUs, and pilot ‘low-carbon delivery’ SLAs with small customer incentives. We should be publishing CO₂e per order, total CO₂e savings, percentage of green packaging used, and the share of last-mile delivery done by electric vehicles.”



Sun agrees that sustainability needs to move to a solid commitment. “To prevent ESG initiatives from becoming one-off promotions, platforms need to build long-term mechanisms dedicated to sustainability funds, open environmental data sharing, and continuous resource investment,” she says.



Still, the economic context is hard to ignore.



“During the downturn, consumers tend to deprioritise environmental concerns,” says Yaling Jiang. “Chinese companies intend to make ESG efforts not for consumers, but for the government.”



And for the few who do care, trust is fragile. “The recent Arc'teryx scandal is a cautionary example,” Jiang warns. “It should be a warning sign for all.”

During the recent pre-order period in late October, Arc'teryx did not appear on Tmall’s list of the top 20 fastest-growing sports and outdoor brands. Was it a quiet rebuke for the brand’s ecological misstep in the Himalayas or just a coincidence? Either way, after 17 years of faster, smarter, louder selling, perhaps the smartest move Double 11 could make is to pause and reckon with what all that efficiency is costing.