Traditional retail stores have many pain points but still play a crucial role in modern megacities. With the development of AI technology, computer vision offers some solutions for more efficient shopping experiences without lines or checkout. In the meantime, labor costs could also be saved, which have been considered as one major part of convenient stores' operating expense.
However, unmanned stores have also faced challenges around such areas as infrastructure, flexibility and risk of loss. Capital and market's enthusiasm about unmanned retail stores is cooling down gradually. Some stores have already left the game, and some are still entering the market.
Sandstar (视达), founded in 2017 and previously named YI Tunnel, is an unmanned convenient store solution provider, based on AI, facial recognition technology and big data. Through real-time monitoring of sales data, future advances in supply chain optimization are also expected from Sandstar.
One of Sandstar' s technologies that attracted investment was its dynamic recognition technology. Compared with traditional radio frequency identification, it can increase the accuracy rate to 99.7%, with a recognition time of 0.02 second. In addition, the infrastructure is easy to build up through installing cameras or upgrading previous ones. Accordingly, the one-time-spending rate is relatively low. The company claimed that the intelligence shelf only costs 40% of the price of a traditional vending machine; the operating cost will also be 5% of Amazon Go’s.
Amazon unveiled its first unmanned convenient store in Seattle in 2016. According to Bloomberg’s article in September 2018, Amazon Go will consider opening up to 3,000 cashierless stores by 2021. A global investment bank, RBC Capital Markets, also published a report. RBC estimated that the average number of customers in Amazon Go per day ranges between 400 and 700 shoppers, with an average $10 per transaction. Under this simulation, Amazon Go stores could generate sales revenues of $4.5 billion by 2021.
Not only Amazon Go, domestic retail giants also take this as an opportunity to transform. Suning smart unmanned store has been developed last month, based on its big data platform, AI technology, and the Internet of Things (IoT) to implement quick payment and risk control platform. Compared with Suning's previous digital 24-hour store Biu, it has further upgraded its capabilities in big data selection, precision marketing, and procurement management.
The traditional retail industry has been based on empirical decisions, even today, offline retail processes, such as procurement, placement, pricing, still tend to rely on experience. It is believed that the future retail industry will be data-driven rather than experience-based. From the market demand point of view, the size of the domestic unmanned retail market transaction has been close to CNY 20 billion in 2017. It is expected to exceed CNY 65 billion in 2020 and attain a compound annual growth rate of 50% in three years. There is still huge room for growth in the Chinese market.
The aim of Sandstar is to help retailers to understand consumer behaviors, make more logical decisions, and further reduce costs. “Our mission is to let retailers make more money and save more money,” said Wu Yili (吴一黎), CEO of Sandstar.