Yu-Yi Chen, ITRI's Industrial Technology International Strategy Development Institute

​

New consumption patterns are causing a major transformation in the global retail industry

Due to external factors such as changes in global consumption patterns, the rapid growth of e-commerce and social platforms, and labor shortages, as well as the fact that most traditional retail and service industries have not had enough time to transform digitally, retail and service industries around the world have experienced a wave of closures since 2015, with a large number of physical stores closing down, setting new records every year. According to statistics, about 8,100 stores closed in the United States in 2017, about 5,800 in 2018, and it is estimated that 12,000 stores will close in 2019. In 2018, 20,000 retail stores in the United Kingdom closed down, laying off nearly 150,000 employees. Japanese stores are also facing the same crisis. In 2018, 465 century-old Japanese companies announced bankruptcy, suspension of business, and dissolution, breaking the record since 2000. In March 2019, 7-11 Japan announced that some of its directly-operated stores would no longer operate 24 hours a day, but would shorten their hours to 16 hours.

The retail service industry usually starts with local consumer life needs, so the retail application services developed in various regional markets have a high degree of localization, and at the same time, they also shape the ecological system of the retail service industry in various regional markets. Especially at this point in time, various new terms such as "new retail", "super retail", "technological retail", and "unmanned retail" have been introduced, reflecting the urgency of the global retail service industry facing digital transformation. There are two reasons: First, from the end of 2017 to the present, the global retail service industry has faced various digital shocks. For example, the rapid development of various technologies such as the Internet of Things, user big data, and artificial intelligence are changing the lifestyle and consumption behavior of the general consumer public. The outbreak of bankruptcies in the retail service industry has also sounded the warning of "if you don't change, you will be eliminated." Second, the emergence of various disruptive business models (such as Uber, Airbnb, etc.), small-volume, diversified, rapid customization, shortening of the global value chain and other driving factors have also led to changes in the retail supply chain, from the past linear model to the digital collaborative network model (Digital Supply Network, DSN). Retail service providers must be able to respond quickly and dynamically handle all current situations and demands of consumers, partners and upstream and downstream related manufacturers. This has prompted retail service providers to introduce new technologies and then create new business models and new retail application service chains to continue operations or seek new business opportunities.

Taiwan has the natural and acquired advantages to develop a new technology service industry with advantages

Taiwan's innate and acquired conditions are conducive to the development of the retail service industry. In terms of population consumption characteristics, Taiwan is small but densely populated. Its population density ranks tenth in the world, about 651 people per square kilometer. Taipei City is even higher, at 9,844 per square kilometer. There are an average of 6 convenience stores per square kilometer, so Taiwan's convenience store density ranks second in the world, with an average of one convenience store for every 2,211 people. Taiwanese people also like to go to convenience stores. In 2017, the total number of visitors to the four major convenience stores reached 2.88 billion times, with an average of 122 visits per person per year, and a visit every less than 3 days. The average annual consumption per person is nearly 10,000 yuan. Currently, the four major convenience stores in Taiwan are actively developing retail service technology, and each has its own technology projects and partners. FamilyMart and Hi-Life prioritize saving store staff labor, while OK and 7-11 focus on smart vending machines. In addition to convenience stores, hypermarkets, supermarkets, hypermarkets, and beauty chain stores are also well developed in Taiwan. It is worth noting that these retail stores are constantly changing store formats, developing new customer groups, or integrating different businesses to keep old customers coming back or create new customer traffic. Therefore, the boundaries between hypermarkets, supermarkets, and convenience stores are also beginning to blur. For example: (1) RT-Mart opened a "24-hour convenience store" with a business area of 70 square meters, operating in a convenience store format with a large number of products. (1) The company reduced its sales volume to 10% of that of hypermarkets, selecting 4,000 items for regular sale; (2) Convenience stores, supermarkets and hypermarkets all used fresh and cooked food as their selling point, and introduced a variety of nutritious foods, fresh food and handmade cooked food to meet the dietary habits and health requirements of the elderly and young people; (3) In order to further integrate into the local living circle, the company also tried to combine with different industries such as pharmacies, baking, fitness centers, laundry services, etc. to carry out complex operations.

According to a study conducted by the International Institute of Obstetrics and Gynecology, Taiwan has also developed new technology service industries in the retail industry, which can be summarized into four categories, as shown in the figure below:

ITRI uses AIoT integration technology to solve Taiwan Hi-Life's long-term inventory and manpower allocation pain points

ITRI Service Center, Advantech, Innotrans, and other domestic companies have assisted Hi-Life, a local convenience store brand established in Taiwan for 30 years, to develop domestically developed smart store technology and successfully apply it to the Hi-Life "Smart Technology Store" at National Chiao Tung University in Hsinchu. ITRI's technical solution is mainly aimed at the pain points of the industry: due to limited storage space, it is impossible to quickly process a large number of e-commerce orders, which often leads to misplacement of goods, affecting the delivery time (an average delay of 3 days), causing excessive labor load and other problems. Therefore, ITRI provides four main functions to solve these pain points: (1) Precision storage and retrieval: Greatly improve the efficiency of warehouse sorting of items, develop AI demand forecasting and warehouse decision-making technology, assist Hi-Life's convenience store system in responding to the bottleneck of virtual and real order fulfillment, introduce AI high-density dynamic storage and sorting decision technology into warehouse operations, and integrate related automation hardware and software systems to support Hi-Life's large-scale expansion of 500 stores in three years. The physical channel order fulfillment needs are achieved, and the order fulfillment speed is 100,000 pieces (originally 50,000 pieces) per day. The sorting processing volume reduces 1/2 of the employees, increases product sales revenue, and creates package store pickup revenue. (2) Self-driving inventory carts: Robots are used to quickly count shelf merchandise, which reduces the burden on employees and reduces human inventory errors. (3) Electronic labels: Electronic labels managed by WiFi connections replace traditional paper price tags, which can update correct product information and more flexible promotional programs in real time, effectively reducing the tedious work items of store clerks and the risk of price errors. (4) Talent identification: Talent identification is developed in response to the needs of e-commerce delivery, allowing consumers to measure the size of their packages by themselves, improve the accuracy of freight pricing, improve the efficiency of package delivery, and also help small e-commerce companies to deliver goods more efficiently. The entire technical solution integrates Taiwan's advantages in smart hardware technology to solve common problems in Taiwan's general convenience stores, such as high rents, labor shortages, and insufficient storage space. It is expected that it can be expanded to other physical stores in the future to accelerate the automation efficiency of Taiwan's stores.

Using AI and data analysis, Mobile Bagel helps retailers improve marketing and reduce storage costs

Traditional retail industries at home and abroad are also facing transformation, and usually face five major problems: First, it is difficult to integrate data and information within the enterprise. From the past to the present, the internal digitalization of enterprises is mostly completed in a project-based manner. Whether it is completed by the internal information team of the enterprise or outsourced projects, it is usually difficult to integrate various data or databases; second, the degree of digitalization of various upstream suppliers is inconsistent, and it is impossible to grasp the upstream supply chain in real time. Third, when opening new stores, costs often rise due to poor inventory management; fourth, marketing and inventory forecasting still use traditional manual experience to forecast, and are highly dependent on experienced employees to forecast inventory and then formulate marketing plans, which makes it difficult to ensure accuracy and quality; fifth, consumer tastes change, and promotional activities or new shopping experiences must be held from time to time to retain old customers and develop new customer groups. The above five major problems are all problems of supply chain management in the retail industry. The most fundamental problem is that various data on the entire supply chain are mostly "data islands", lacking integrity, and making it impossible for operators to mine gold mines in these data. 96% of Taiwan's retail industry is small and medium-sized enterprises, which have difficulties in using the latest technology to transform and innovate: first, lack of experience in integrating and making good use of data; second, the cost of establishing an internal technical team is too high; third, the high complexity of optimizing data or systems.

Mobile Bagel, a Taiwanese startup, initially targeted common problems and needs faced by retailers, such as integrating all data from upstream and downstream supply chains, data generated by various terminal devices, and online and offline customer data. It then systematically analyzed these big data and generated meaningful data to accelerate the automation of the entire retail supply chain service.

Mobile Bagel has developed its own AI prediction tool, the Decanter machine learning engine, which combines 136 algorithms to build AI modules that meet the needs of enterprises. The current benefits of this technology are: (1) Improving the marketing of physical retail stores: increasing the monthly revenue growth of a single store by 15%; (2) Improving the efficiency of data analysis: data analysis that originally took 3 to 6 months has been shortened to 2 weeks, such as marketing effectiveness prediction, inventory prediction, and production line machine life prediction; (3) Reducing inventory costs: predicting the inventory of tens of thousands of items in chain restaurants in the next seven days, with a prediction accuracy rate of 90%, which is 1.46 times higher than the accuracy of other systems. Mobile Bagel's successful model is to grasp the technological gaps in the transformation of traditional industries, such as big data analysis, introduction of AI tools, establishment of AI models, etc., and then provide customized AI technology services, and further expand into other industries, including telecommunications, finance, insurance, manufacturing, etc.

Improving customer experience and reducing operating costs: Smart retail drives new industry value

Whether it is retailers or new technology service providers, they are facing rapid changes in the international economic situation, consumption patterns and labor structure. With the development of AI, deployment of IoT and growth of big data, smart retail will continue to innovate. In addition to using cool experience to create short-term marketing benefits, it is recommended to accelerate overall operational efficiency, save labor burden, improve employee value and focus on customer experience. Taiwan has the advantages of smart hardware industry and has the niche to develop the emerging retail technology service industry that integrates software and hardware. Taiwan retailers and consumers will test and demonstrate technology applications and then export them to the entire Asia-Pacific market.

​

IEKView

2019 is regarded as the first year of technology retail. At this point in time, various new terms such as "new retail", "super retail", "technological retail", and "unmanned retail" are constantly being introduced, reflecting the urgency and necessity of the global retail service industry facing digital transformation, hoping to drive new business models through new technologies and obtain new business opportunities. Taiwan has advantages such as consumer population characteristics for the development of retail service applications, the world's second largest convenience store density, and smart hardware industry technology. Taiwan's high-tech companies can connect with retail service companies with application fields to conduct field testing to build flexible manufacturing, assembly and verification of AIoT smart retail technology solutions, shaping Taiwan's AIoT smart retail technology service industry chain. On the one hand, it can solve the transformation needs of Taiwan's retail service industry, and on the other hand, it can export Taiwan's local AIoT smart retail technology service solutions or innovations to other countries and markets in the Asia-Pacific region.