China's retail revolution: Smart shelves, livestreamed avatars, and clued-in consumers

Huawei's Executive Editor-in-Chief Gavin Allen sits down with the head of China's leading trade association for chain stores, franchisors, and consumer goods companies.

Peng Jianzhen: China's retailers began digitalizing after 2012 as offline traffic declined. Building websites wasn’t enough; integrating online and offline operations became critical. This required systemic restructuring, breaking down silos, and policy-backed investments in infrastructure, accelerating transformation. Now, companies prioritize extracting value from data.

Gavin Allen: What’s the biggest disconnect between consumer expectations and operational capabilities?

Peng Jianzhen: We need to look beyond digitalization now because China’s rapid growth has created an economic shift. Since 2020, annual per capita GDP has exceeded US$10,000. That brought a major change in consumption. Consumers used to be open to many different commodities: a range of dishes that consumers thought were quite good, so long as they were brought to the table. But now, they’ve tried all those dishes. With more money, they’re more selective – they just want to pick ones they like or have never tried. The challenge for Chinese retailers is that they’re still offering the same commodities to consumers who don’t want common or normal goods, they want new dishes.

It happened in the US, too, but the differentiation in the US happened around 1978, before e-commerce emerged. In China, the market had to deal with the challenges brought by e-commerce, and then, entered the stage of differentiation.

Walmart’s Sam's Club is successful in China because it provides commodities that consumers need and enjoy, with a localized supply chain and targeting consumers who are middle class and beyond. Sam's Club offers exclusive merchandise not available in other stores.

Pangdonglai, or DL, is another successful retailer that has developed a range of good commodities. Consumers are increasingly paying attention to healthy products, such as sushi and diet foods. However, across the industry, product homogenization remains the most pressing challenge we face. Digitalization can help to develop the products consumers need and optimize the supply chain.

Gavin Allen: What’s the biggest bottleneck in retail digitalization?

Peng Jianzhen: E-commerce penetrated into every aspect of Chinese life early on. Consumers required a lot from retailers, expecting offline stores to be digital too, with delivery and instant service. As digital consumers seek more than just offline shopping, they expect multiple touchpoints — and retailers have had to respond by creating them. With strong technological capabilities, Sam’s Club has been able to execute its transformation efficiently. The challenge, however, lies in how other retailers can respond to increasingly diverse and differentiated consumer needs — any meaningful change requires substantial investment, and they must carefully evaluate whether such investment will generate adequate returns.

Gavin Allen: What do retailers get from digitalizing?

Peng Jianzhen:  China’s end-to-end methodology is called “people-goods-scenarios.” Years ago, retailers didn’t know who went to their stores. Today, with membership schemes and digital assistance, they know who enters, what they purchased, and even which commodities were picked up but put back on the shelf. Collecting this information gives stores better understanding of consumer behavior, enabling automatic recommendations and improved customer interaction.

Replenishment of stock has also become more intelligent. Moreover, the concept of "scenarios" is no longer limited to restocking shelves — it now extends to delivery and warehousing operations. There are many more touch points, with the whole process much smoother. We can capture the entire consumption journey, as e-commerce does, and a company such as Rainbow can even evaluate customer satisfaction. Supply management has been fully visualized and internal management improved. For instance, through digital applications, stores can flexibly recruit temporary staff via apps during peak hours. This allows companies to manage their workforce more flexibly, while also enabling these workers to earn additional income. Digital technologies address many pain points that could not be addressed before.

Gavin Allen: How will disruptive technologies like Generative AI and DeepSeek reshape retail, and how widespread is AI adoption?

Peng Jianzhen:  GenAI is very strong in data processing and content generation. It can be integrated into marketing, supply chain management, customer service, and many other retail aspects. But the journey is still ongoing. The future holds significant opportunities for further exploration and innovation.

Our members represent diverse business models with distinct needs. Generative AI can be applied in various ways, but its adoption varies depending on the enterprise’s strategic focus. For instance, a retailer can use AI to trigger an alarm if theft is detected during bill payment. For discount stores that sell a large volume of near-expiry food products, AI can help improve the accuracy of sales forecasting, guiding both procurement and pricing decisions—a capability that is critical to their business model.

For example, Bosideng, a down jacket brand, needs to complete the design and production of its winter collection by summer. To anticipate consumer preferences, sensors are installed on summer garments in stores to track which styles are tried on most frequently. By analyzing large volumes of data—such as try-on frequency, sales performance, and customer feedback—AI can identify which colors, styles, and cuts are most popular, predict upcoming fashion trends, and support design decision-making.

Retailers aim to leverage AI to develop products that are more likely to appeal to consumers. Social e-commerce providers and players will want to adopt some AI algorithms and use AI for recommendations. Meanwhile, one of the most frequent applications is the use of digital avatars during livestreaming.

However, it’s generally point by point, and not really at a systematic level. Technology’s iteration is very fast, so maybe at some point there will be a disruptive change. But many enterprises are still quite cautious about adopting AI technologies. Using AI to empower their staff is the consensus.

Gavin Allen: How challenging is it for small and medium-sized retailers to make the necessary investments?

Peng Jianzhen: The balance between investment and return is a big challenge for retailers of all sizes.

In this digitalization phase, although enterprises know what they need, traditional ERP providers often struggle to meet their requirements. To better serve both online and offline consumers, some companies have had to develop their own systems by themselves — a move that requires significant investment.

For example, in the early years, a leading retail enterprise invested over 5 billion RMB in this area. Rainbow has an R&D team of 300 to 400 people. And all the investment is not done at one time, but continuously, as investment is needed every year. For many retailers, their revenue cannot support that investment. So, in China, some retailers have started to turn their digital departments into service providers — offering training and exporting their digital capabilities to other businesses, in an effort to achieve cost recovery through the external sale of digital services.

Gavin Allen: You’ve said “store digitalization remains an industry shortcoming.” How so, and what’s hindering progress?

Peng Jianzhen:  Enterprises know that digitalization is the way forward, but there are still shortcomings at the store level. We have the tools, but the processes and staff mindset are out of sync. Some staff lack a strong educational background and prefer to use traditional approaches: why bother to use digital technologies, they argue, since it could make their work more complicated? That is a big challenge. So, starting last year, we worked with EY (Ernst & Young) and Intel on a project to improve digital technology utilization by staff and released new guidelines. Also, in the past, data collected from stores was transferred back to HQ or the cloud for processing. But now, data collected locally can be processed directly at the edge, making tech utilization more convenient. So, we also provide many tools and methodologies to help staff better use those digital technologies. Staff training is really important.

Gavin Allen: How can AI not only enhance efficiencies, but also deliver the emotional value that customers want?

Peng Jianzhen: This is a key focus, particularly for offline retailers. If we pursue efficiency alone while neglecting the human touch, what value do physical retail stores truly offer? Retailers want tools which can enhance both. And warmth means building trust between the consumer and offline store, by delivering quality services or by providing the most suitable products. For example, Sam's Club and Pangdonglai have very successful experience in developing products, which built trust with their consumers. Then, the retailer can take lots of advantage by using digital and intelligent technologies on it. So, AI may not directly deliver emotional value to consumers, but it can help build trust with them.

Gavin Allen: Which retail segments will face the most disruption from digital intelligence?

Peng Jianzhen:  We should not intentionally pursue disruptive change. It should follow or be based upon satisfying the changing needs of consumers. For instance, in China, Alipay was a disruptive innovation. When Alibaba first launched its e-commerce platform, consumers lacked trust in the system and were hesitant to pay with credit cards. To address this issue, Alibaba introduced Alipay, a third-party payment service. With Alipay, customer payments are held in escrow and only released to the seller after the buyer confirms receipt of the goods. This mechanism helped build trust and protected both buyers and sellers. It was disruptive. But it wasn’t created out of imagination. It was created to solve a trust problem between a platform and its consumers. Many disruptive capabilities in China were created like that. If you identify a consumer pain point that’s still to be solved, and solve it, then a disruptive technology may emerge.

Gavin Allen: Amidst all this industry change, what keeps you awake at night?

Peng Jianzhen:  The biggest challenge for retail digitalization in China, and what keeps me awake at night, is that many companies find the existing systems unable to meet their needs. So, they are forced to develop their own systems, which is very tough with often unfavorable returns on investment. I really hope and believe that a company, maybe Huawei, will one day be able to solve this problem.