Earlier this month, Alibaba’s Ali Cloud hosted the annual Yunqi Conference in Hangzhou, and Azoya sent representatives to participate in various conferences regarding retailing, e-commerce and supply chain. The Yunqi Conference has a tradition of focus on cloud technology, and this year’s focus is on retail tech. As the very fundamentals of e-commerce, cloud technology has enabled retailers’ digitalization. Development such as artificial intelligence, big-data, IoT (Internet of Things), and smart supply chain are also revolutionizing the retail landscape.
Ali Cloud New Retail Summit
The four-day event presented a lot of new concepts, technologies and demos. We divide this article into 3 parts: new retail transformation, smart supply chain, and artificial intelligence, to help you understand what are changing in China’s retail market.
New Retail Transformation
The ‘New Retail’ concept has been a hot topic in China for nearly a year, and now the retail industry has finally hatched some good cases. With domestic e-commerce growth stagnated, retailers begin to realize that the offline bricks-and-mortar stores may be the next growth point. By simply dividing customers into offline or online group, retailers are missing the picture of what their online customers do when they visit offline stores, or vice versa.
Alibaba Cloud Retail Services introducing New Retail Business Transformation
The so called ‘New Retail’ transformation is to help retailers remove the data boundaries between offline and online retailing, and provide a holistic experience for the customers.
One of the concept under ‘New Retail’ transformation is that every object should have both an online and offline form. For example, a female handbag presented to offline customers will also have pictures, profiles, specs and other digital assets available for e-commerce. For pure e-commerce, having an offline store where customers can experience, trial and receive post-sale service is important.
For offline retailers, the increasing use of IoT devices, sensors and cameras will also help make remedies of the lack of tracking on their customer traffic.
Evolved human-machine interaction through AR (Augmented Reality) technology. Now that everyone has a mobile phone with high-res cameras, improved computing power on mobile devices and on cloud, it is possible for retailers to create better shopping experience. There is a ‘smart dressing mirror’ that can pick up customer height, body curve, facial image information, so that customers can try out new costumes without having to go to the fitting room. For beauty industry, Alibaba had developed a smart mirror that can capture the face of customers, and simultaneously renders the effect of cosmetics (lips, eyes), accessories (glasses, earings) to help customers make decision.
AR technology to help make furnishing decision
Female experiencing virtual mirror at the Ali Cloud exhibition. Imagine that this mirror will be placed in all the apparel retail stores in the future.
App that can help customers try on new lipsticks virtually. Mobile phone camera to capture the facial image of customers, and apply the lipstick on their phone after adjusting environment parameters.
App that can help customers find products in-store (in-store locating). When using the app in an offline supermarket, customers can know the shortest route to pick up products.
Facial recognition technology by Alipay to enable payment by looking at the camera. The technology is already in commercial use at selected KFC restaurants in Hangzhou and Shenzhen (Mix-City Mall).
Movement tracking via sensors and cameras in-store to help analyze customer behavior (e.g. traffic monitoring, customer identification, picked up what kinds of products, abandoned products, most favorite, combining with payment information, decision process, in-store routing). This is similar to the Amazon Go project. With this technology applied in-store, physical retailing would be benefited with an explosion of information that will help them make better merchandising decision and provide better customized experience to customer.
Retail Connect: Software & hardware (“零售通”: “Retail Connect”) to empower mom & pop shops with better insights in the serving parameters of convenient stores/small & medium stores, e.g. where is the customer traffic within the 3KM of the stores, what are the events happening that can boost sales in the parameters, what are the target customers like in the location (e.g. working hours, when are they working overtimes, when are they having vacation, what time do they come back, what are their income, what do they purchase most, etc). The system can also help them manage order from the suppliers. Alibaba expect to launch this system to more than 1 million retailers in China in the next few years.
‘Data Bank’: This is a very powerful marketing tool for brands to enable full tracking, multi-tier management and planning of marketing events based on AIPL (awareness, interest, purchase, loyalty) theory. Traditionally, brands in China have limited ways to track the performance of advertisement on traditional media outlets, or ads out of the eco-system of Alibaba. Global brands such as Unilever and P&G had cut digital ads budget by more than 40%. Now they can track almost everything that are in the eco-system of Alibaba (e.g. Youku for video, Weibo & Momo for social media, Tmall for e-commerce, Alipay for finance). The process includes to integrate Brands’ own database with Alibaba’s, and tag customers with improved accuracy and more information. Now there are more than 300 brands in China that had integrated this databank. Mars the chocolate manufacturer (Snicker, Dove, M&M) highlighted the operation with Alibaba so that they can launch new products with smarter media budget and marketing plans.
Smart Supply Chain in practice
The traditional pain points of retail supply chain include inefficient planning, labor intensive warehousing, and slow response to market changes. Now, increased usage of cloud computing, IoT equipment (sensors, cameras) in the supply chain will serve to improve overall efficiency.
Sensors can capture a lot of information that traditional man-powered warehouses are unavailable or too costly to perform, e.g. geographical data, time, property, recipient’s information etc. While the analytics of data from upstream and downstream supply chain enables retailers to stock smartly, optimize logistics route, allocate resources, make predictions (forecasting), create better services, etc.
Five challenges regarding the supply chain of current day e-commerce businesses were identified: multi-category (products with various attributes), up-stream and down-stream collaboration, complex warehouses network, globalization, and stability.
Deep learning and machine learning algorithm are applied to tackle the challenge. Under the complex network and increasingly diversified consumer demands, retailers will need a smarter supply chain to automate and respond quickly without human intervention. In the operation level, AI can be used to identify tags such as popularity, best seller, high margin, attractor, etc. When a product is contributing less to the platform considering its placement, resource, time, revenue, margin, it will be automatically replaced.
For example, the ERP system will order more mooncakes from suppliers automatically when the Dragon Boat Festival approaches, launch with optimal pricing that balance sales and revenue, and take off the products when the festival ends. This requires that retailers to have end-to-end data to enable smart supply chain. The ultimate goal is to rely less on human experience, and maximize efficiency by data and analytics.
Use of AI in retailing has been a rather new trend. For retailers, AI can be mainly used in these areas: capturing information, process information, make decision.
Firstly, capturing information:
Alibaba is using sensors, IoT devices, cameras to capture massive information, and store them on Cloud. Machine vision, movement tracking, identification, recognition technology are being used in logistics, retailing, payment and many areas.
Ali Cloud AI conference introducing the future of machine imagery: what can be seen can be analysed, what cannot be seen cannot be analysed
Secondly, processing information
Chief scientist of Alibaba said on a conference that ‘new retail’ is not just using data to improve efficiency, but ‘information’. There are a lot of data, and to use the data to conclusion is difficult. They are using deep learning, machine learning and many types of algorithms to decipher the data they have so that they can develop smarter work flow that will save billions.
For example, in some scenarios, e-commerce operation team can leverage AI to make decision without human interaction. They are using machine learning to determine when to launch products, when to remove them, when to price higher, when to make discount, when to order from suppliers. They take into consideration the product functions, product price, margin of retailing, competition in market, supply chain cycles, and many more factors, and they will organize supply chain, logistics so that they will be most efficient/automated. Traditionally, merchandising and supply chain synergy relies on human experience, they are now being replaced/aided by artificial intelligence. With massive amount of data, Alibaba knows categories very well. They know when and where they will need certain products, and of course when and where certain products will be needed.
Thirdly, act on information:
Having the information means smarter decision, and such decision can be made in every interaction with customers. For example, you will be greeted by a shop guide that knows what your preferences are, you will be recommended products when your perfumes, or toothpastes, or shampoos run out.