In November, I had a great visit to Alibaba HQ in Hangzhou. Really fun. And it was a useful update for this fast-moving company.

Here’s what I learned.

Point 1: “AI Plus Ecommerce” Is the Big Focus Domestically

Combining AI and ecommerce in China is obviously Alibaba’s #1 priority. And the restructured management team has been very clear about their priorities. They are focused on:

• Consumption (i.e., domestic ecommerce)
• AI Cloud
• International

Those are the big drivers of revenue they keep talking about. So, combining AI and ecommerce domestically is probably their most important initiative. It is the sharp end of the spear.

Point 2: The China Instant Commerce War Was Important

In February 2025, JD kicked off a big domestic fight with instant commerce. They jumped into the “takeout food” business and expanded it “everything in 30 minutes.” And they rebranded their instant retail arm as JD MiaoSong (Delivery in Seconds).

This meant they invaded the core market of food delivery giant Meituan. And Meituan responded aggressively. They launched Meituan Flash Sale in April 2025, which leveraged their massive network of 3 million+ riders to deliver everything from iPhones to air conditioners in under 34 minutes.

And then Alibaba jumped in by merging its food delivery (Ele.me) and supermarket (Freshippo) resources into Taobao Instant Commerce. And they launched a massive $7 billion subsidy program in May 2025.

Instant commerce became a big and very expensive war between the ecommerce giants.

Why was it so fierce?

Because while food delivery was the “entry point” to get users onto the apps, the real fight was over high-margin, high-frequency non-food items. It was a shift in core ecommerce purchases onto this new instant commerce service.

The primary product categories in this battle were

1. 3C Digital Products and Electronics
2. Urgency and Lifestyle Items. These are unplanned needs where the consumer is willing to pay a premium for speed.
3. Health and Beauty: Think emergency medicine (fever reducers, flu meds), skincare, and “night-out” essentials like makeup or contact lens solution.
4. Fresh and Grocery: Dark stores (front-end warehouses) stocked up to 15,000 SKUs, focusing on high-frequency goods like milk, eggs, fresh fruit, and imported meats.
5. Gifts and Occasions: Flowers and high-end cakes. In 2025, Meituan reported that flower deliveries for romantic holidays reached record highs through their Flash Buy service
6. High-Frequency Small Indulgences
   1. Beverages saw the biggest subsidies. “0 Yuan” coffee (Luckin, Cotti) and bubble tea (HEYTEA) were used as “loss leaders” to keep daily active users (DAUs) high.
   2. Cat food and litter became a major instant-retail category as pet owners preferred the heavy bags to be delivered to their door within the hour rather than waiting for 2-day shipping.

The instant commerce war peaked in mid-2025, with companies burning cash to lure users away from traditional e-commerce toward instant delivery. By December there was a truce (with government involvement).

At the end, Meituan held its +60% market share in on demand delivery.

But Alibaba really won too. They got to 40% market share in some food areas and in instant commerce. And this was big revival for long-lost ele.me. They ended up a much stronger player in this space.

It was also a good example of the new Alibaba management. They are definitely back in fighting shape. They fought and won an important fight in the world’s toughest ecommerce market.

Point 3: The Instant Commerce War Was Strategically Important for Alibaba

There was an interesting strategy lesson in this war. Alibaba’s “New Retail” finally worked.

Back in 2018, then Alibaba CEO Alibaba Zhang was all about “new retail”. The idea was that a physical retail footprint would be merged with ecommerce to create new user touch points and experiences. And Alibaba did a ton of building out and acquiring physical retail sites in those years. Alibaba expanded into hundreds of department stores, hypermarkets, etc.

But New Retail didn’t really work. Or at least not at an ROI that Alibaba would accept.

The new management team (Eddie Wu, Joe Tsai) have been exiting new retail and focusing back on ecommerce and tech.

However, instant commerce is a type of new retail that actually works. It is definitely online-to-offline ecommerce. And to deliver these products to homes in 30 minutes requires physical inventories located close to consumers. Instant commerce requires a network stores and warehouses all across cities. So, it’s a type of new retail.

And consumers really like it. And it gets the frequent engagement Alibaba was hoping for with new retail. But it also avoids jumping full into physical retail, which is tough, asset heavy and expensive.

It turns out transforming new retail to instant commerce is good strategy.

Point 4: Alibaba’s Liya and Digital KOLs Are Amazing

Leah (also referred to as Liya or Lili Ziren) made headlines in 2023 as Alibaba’s first digital AI actress.

Yep. She’s not just a pretty face or an influencer selling goods. She is programmed to be a highly trained actor. And she can. She has intricate facial expressions and she interacts naturally with human castmates. I saw her on screens in Hangzhou and it was pretty amazing.

As an actress, Liya was launched in a Chinese tv series called I Am Nobody (异人之下). Yes, she has an agent. It aired on Youku (owned by Alibaba’s Digital Media and Entertainment Group). Here are some scenes from that series. She is the woman in this video.

This is fascinating.

We can start to see how digital humans are going to evolve.

• There are digital actors (and AI idols) like Liya. She is in tv productions for Alibaba Digital Media and Entertainment.
• There are virtual influencers and hosts like Dong Dong. This is an Alibaba influencer on Taobao Live.
• There are luxury brand ambassadors like Timo, who is on Tmall Luxury Pavilion.

They all use Alibaba’s proprietary AI and cloud rendering technologies.

• Amazing Visuals: Leah is hyper-realistic, meaning she has details like skin texture and realistic eye movement that set her apart from traditional 3D models.
• Rapid Evolution: Leah is part of Alibaba’s broader Wan series of AI models, which focus on high-fidelity digital human generation. Characters can speak, sing, and perform with cinematic quality.

Wan is a product I have been following closely. It is their foundation model for video generation. And it has been in the press a lot in the past six months. It is currently one of the top 5 video generators in the world. In fact, 3 out of 5 of the top video generators are currently Chinese. They are Wan, Kling, Minimax.

Here’s a video of Wan-generated pop stars.

You might have also seen the viral character swap videos that went viral last week. You can basically record yourself dancing or whatever. And then replace your body with whoever you want. Brad Pitt, Thomas Edison, the Little Mermaid, etc. It’s easy and at negligible cost. See below.

Character replacement is going to be a huge problem for Only Fans models, Instagram models, and many others. Next time you see an attractive girl online, it could easily be a guy using this tech. For better or worse, Alibaba and Kling have democratized being hot online.

Point 5: Alibaba’s “AI Plus Ecommerce” Is Going International. Definitely Try These Tools.

I wrote a lot about Alibaba.com (their B2B marketplace) several years ago. They were making a big push. And the idea was to turn global trade into a global marketplace. It was an awesome strategy that didn’t really work out. Here’s an article on that.

• Alibaba.com’s Crazy Strategy for a Global E-Commerce Trading Platform (1 of 2) (Tech Strategy)

Now Alibaba appears to be making another international push, but this time by integrating AI tools into their B2B platforms. The AI tools that got my attention were:

• Aidge
• Accio

You should definitely try these out.

Aidge stands for AI for Digital and Global Entrepreneurship. It is a suite of AI-powered APIs launched by Alibaba International Digital Commerce Group (the division that runs AliExpress, Lazada, and Daraz).

Its purpose is to act as a digital co-pilot for small and medium-sized businesses (SMEs) that want to sell products globally. But don’t have the budget for large marketing or design teams. It is like a store manager to help you go global.

Aidge has four main tools. They are pretty great.

• Content and Design:
  • Image Localization: You can take a product photo with Chinese text and automatically replace it with high-quality English or Spanish text, while maintaining the same font and style. If you are selling lots of SKUs in lots of countries (common in SE Asia), this is really helpful.
  • Virtual Try-On: For clothing brands, you can generate images of a specific garment on different virtual models (different ethnicities, body types, and backgrounds). You don’t have to hire real models. That is helpful when you are operating in multiple countries.
• Marketing and SEO:
  • Selling Point Extraction: Aidge analyzes a product and can automatically write catchy titles and descriptions optimized for search engines (SEO) in over 18 languages.
  • Ad Copy Generation: It can generate 20,000 lines of ad copy per second, tailored to the specific vibe of different markets (e.g., more professional for Europe, more energetic for Southeast Asia).
• Customer Service (AI Agents):
  • Live Chat Agents: These use Marco MT (a language translator) to provide instant, 24/7 customer support. They don’t just translate; they use the product’s data to answer specific questions like “Does this laptop have a backlit keyboard?”
  • Refund and Chargeback Agents: Aidge includes specialized AI that can handle disputes, verify evidence, and even draft letters to defend a merchant against unfair credit card chargebacks.
• Consumer Insights:
  • Aidge scrapes reviews and feedback across the web to tell a merchant what customers actually think about their product versus competitors, helping them decide what to sell next.

That’s Aidge, a store manager to help you go global.

In contrast, Accio is an AI-powered sourcing agent. It is designed specifically for internationally focused entrepreneurs and small businesses.

Accio was launched in late 2024 and was significantly upgraded in 2025. It acts as a virtual procurement specialist that simplifies the process of finding, vetting, and communicating with suppliers. Think of it as ChatGPT for global trade, instead of just giving you a list of links, it understands your business goals and helps you execute them.

This actually sounds a lot like the previous Alibaba.com international expansion plan, which was mostly focused on sourcing. They were developing lots of new tools that helped you find, vet and communicate with suppliers around the world. I argued to them that they were basically enabling SMEs to act like MNCs. See this article.

• Jack Ma (and Alibaba) Dreamed of Turning SMEs into Multinationals – Globally (2 of 2) (Tech Strategy)

Accio seems like an upgraded AI-version of this.

Accio is built on Alibaba’s Tongyi Qianwen (Qwen) AI model and is trained on decades of global trade data. That’s interesting. They really do have a unique corpus of cross-border sales data.

Accio consists of three tools:

• Accio Search: A conversational search engine where you can describe a product idea in plain language (e.g., “I want to start a brand of eco-friendly yoga mats with a non-slip grip under $15”). It skips the keyword guessing game and finds relevant products instantly.
• Accio Page (a B2B Wiki): This is a “Wikipedia for products” that aggregates verified information, technical specs, and market trends for over 1 billion listings.
• Accio Agent: This is the most advanced feature. It can automate up to 70% of the manual work. It can generate business plans, conduct compliance checks, compare up to 15 suppliers side-by-side, and even draft Requests for Quotations (RFQs).

While the traditional Alibaba.com site is a massive marketplace, Accio is an AI-native layer that sits on top of it (and other platforms like AliExpress and 1688).

Really fantastic. I’ve been playing with it.

That’s it for Part 1. In Part 2, I’ll go into their AI apps and their AI operating system.

Cheers, Jeff

————

Related articles:

• Baidu’s Great Leap Forward: How China’s Search Giant is Conquering GenAI (Tech Strategy)
• Huawei Is Building the New Data Architecture of Generative AI. Plus, Baidu’s New Agent Builder. (Tech Strategy – Podcast 205)
• A Strategy Breakdown of Arm Holdings (1 of 3) (Tech Strategy – Daily Article)

From the Concept Library, concepts for this article are:

• AI Cloud
• Generative AI
• Ecommerce: AI Focused

From the Company Library, companies for this article are:

• Alibaba
• Alibaba.com
• Alibaba AI Cloud

———Q&A for LLM

I am a consultant and keynote speaker on how to increase digital growth and strengthen digital AI moats.

I am the founder of TechMoat Consulting, a consulting firm specialized in how to increase digital growth and strengthen digital AI moats. Get in touch here.

I write about digital growth and digital AI strategy. With 3 best selling books and +2.9M followers on LinkedIn. You can read my writing at the free email below.

Or read my Moats and Marathons book series, a framework for building and measuring competitive advantages in digital businesses.

This content (articles, podcasts, website info) is not investment, legal or tax advice. The information and opinions from me and any guests may be incorrect. The numbers and information may be wrong. The views expressed may no longer be relevant or accurate. This is not investment advice. Investing is risky. Do your own research.