A store no longer earns its floor space simply by processing a sale. Its value lies in helping customers handle unfamiliar products, understand how they work, and decide whether to trust them.
At the “Retail Reimagined Asia vs Europe” panel at BEYOND Expo in Macao, speakers followed that role from factory data and automation to product demonstrations, customer discovery, and cross-border deployment. Their discussion showed how AI could bring shoppers back into stores, while making the technology behind the shopfloor increasingly important.
The offline revival starts in the factory
Carolyn Yim, Principal at Ply-Knits, described how her family’s three-generation garment manufacturing business is bringing decades of operational information into a form that AI systems can use. That information is spread across SQL databases, Microsoft Access, email, and Excel spreadsheets. That accumulated data must first be made legible to natural-language tools and, eventually, agentic systems.
Yim drew a distinction between knowledge AI and industrial AI. Manufacturing requires deterministic results: A factory needs the same specifications, yield, and quality every time. That makes some of the most useful applications decidedly unglamorous.
“Size scaling, material sampling, customs duty, inventory: boring stuff. AI is fantastic at that,” she remarked.
The technology can also help preserve operational knowledge as experienced workers leave and younger people choose careers outside traditional manufacturing centres. Training systems on years of emails and production records could retain skills and process knowledge that would otherwise disappear from the business.
Physical production still presents limits. Garment automation has long struggled with deformable materials because fabric doesn’t behave like a rigid object. Yim pointed to advances in computer vision, robotics, and machine learning that are improving how machines manipulate fabric. She also described smart textiles that embed sensing capabilities during knitting or weaving, turning the material itself into another source of data.
The implications extend beyond factory efficiency. Better product data, more flexible production, and new materials all affect what retailers can demonstrate and explain to customers once those products reach a store.
Stores are becoming places to explain products
Zhijie Liang, CEO of Stardust & Company, argued that physical retail is becoming more important as products become harder to understand from a listing or advertisement alone.
He cited retailers that give emerging technology brands space to demonstrate products with the help of specialists, interactive interfaces, and AI kiosks. This model differs from conventional wholesale distribution, in which manufacturers can move products without necessarily building a brand, explaining the technology, or protecting their margins.
Liang also pointed to physical retail’s role in demystifying specialised tools such as 3D printers, laser cutters, and computer numerical control machines. Consumers may recognise these categories, he said, without having used the equipment or seen how digital files become physical objects. He cited maker hubs in Shenzhen that bring several types of machinery together in one store, where prospective users can see how they work.
“The trend is definitely moving from online to offline because the experience is the crucial part for the future,” Liang said.
This doesn’t make online retail irrelevant. Yim described digital and physical channels as a funnel: Consumers can discover products online, then try them in person. The store becomes the place where the brand is made tangible.
That distinction matters because online discovery has its own weaknesses. Algorithmic feeds may be efficient at serving advertisements, but they can also make discovery feel repetitive and heavily managed.
Yim contrasted that experience with finding an unexpected shop on a side street in Venice. For retailers, the opportunity is to recreate some of that surprise and pleasure without turning the store into a technology demonstration for its own sake.
AI can’t manufacture meaning
Fabien Pacory, Executive Vice President of the French Chamber of Commerce and Industry in China, noted that many Chinese companies already have high-quality products and competitive advantages as they expand overseas. Their difficulty is often building the narrative and experience around those products.
Immersive retail, gamification, loyalty programs, and other forms of engagement can help, but Pacory argued that loyalty needs to have meaning. He described an emerging form of “brand democracy” in which customers, employees, community managers, and other participants gain a greater sense of involvement in the brand.
AI may help retailers personalise or orchestrate that involvement, but the panel repeatedly returned to a human test. Yim said technology-enabled products still need to connect with people through craftsmanship, heritage, and utility. In luxury, she observed renewed attention to embroidery, textiles, and other tactile forms of craft as a counterweight to screen-dominated experiences.
Her simplest product principle was borrowed from entrepreneur Paul Graham: “Make something people want.”
If a product doesn’t resonate with customers, adding more technology won’t create demand. The same applies to the store. Sensors, digital twins, and AI interfaces are useful only when they make discovery, understanding, or service better.
Asia and Europe face different constraints
The comparison between Asian and European retail was less about choosing a winner than understanding why the regions may adopt AI differently.
Pacory pointed to the Greater Bay Area’s network of factories and smaller workshops as an environment that supports rapid experimentation. New materials, sensors, and retail concepts can be tested across a dense manufacturing ecosystem. Europe, meanwhile, presents questions around privacy, data storage, intellectual property, and differing technical and regulatory standards.
Those constraints don’t prevent collaboration, but they affect how technology developed in one market can be deployed in another. A retail system that depends on collecting and processing large volumes of customer or sensor data may need to be redesigned before it can operate under European requirements.
Liang identified localisation, cultural differences, and trust as persistent challenges for Chinese brands entering global markets. Businesses can build local teams, work with experienced partners, or seek outside expertise, but they still need what he called “cultural decoding” between markets.
The offline revival, therefore, can’t be exported as a standard store format. Retailers need to decide which data can move, how an experience should be localised, and what customers in each market need to see before they trust a product.
AI may be pushing retail back towards physical space, but not by restoring the store’s old role. The emerging store is part showroom, part interface, and part discovery engine. Its value depends on systems that begin well before the customer arrives, and on something the technology can’t supply by itself: a reason for the customer to care.
