Last year, on the eve of the traditional American shopping festival "Black Friday", The New York Times published an article entitled: Chasing Amazon, retailers are in an endless arms race. In the face of Amazon's overwhelming advantage, established chain stores are catching up, but it seems difficult to reverse the situation.
Just when they seemed to be in trouble, Amazon once again raised the threshold, not only set up its own aircraft express team, but also applied Amazon's own cloud computing business AWS and artificial intelligence technology, and the pressure on traditional industries doubled. It seems that the competition between traditional industries and Amazon has never been fair competition.
But from another perspective, Amazon is actually innovating traditional industries, constantly stimulating the modernization of the industry through intelligent technology, and accelerating the innovative application of artificial intelligence and machine learning technology in different regions, fields and industries around the world.
In China, Amazon did not simply copy American experience, but focused on localization and continuous innovation practice, which not only made China's cross-border e-commerce industry jump, but also built a global innovation cornerstone based on local demand innovation.
At the "Amazon Innovation Day 2020" held in Amazon China, Li Yanchuan, vice president of Amazon China, summarized the connotation of "China Formula": localization of technological innovation, customization of customer experience and lightweight of business model.
Technology is the foundation and everything is intelligent.
Since its establishment, Amazon has developed from e-commerce platform to e-book Kindle, cloud computing AWS, Prime Air drone, intelligent voice assistant Alexa, and unmanned supermarket Amazon Go. As of 20 19, Amazon has more than120,000 technology patents in the world. Amazon constantly breaks boundaries and becomes a maverick.
Needless to say, the unmanned store Amazon Go was officially opened to the public two years ago. By March 3rd, 2020, Amazon has opened 25 Amazon Go stores in the United States, which can automatically monitor the removal or return of goods from the shelves through Just Walk Out technology and track them in the virtual shopping cart. Consumers can leave the store directly after shopping. Subsequently, Amazon will settle accounts with users through Amazon accounts and provide invoices. Amazon Go uses the same types of technologies as driverless cars: computer vision, sensors and deep learning.
In the field of smart devices, Amazon Echo Dot uses human-computer dialogue to let Amazon shape future life scenes. This smart device can be called "Star Trek Computer at Home". The echo point identifies the consumption environment of the product and ensures the purity and continuity of voice interaction. It seems to be just a toy, but it communicates with users through Alexa voice assistant and becomes an essential expert in smart home.
Now, tens of millions of users are using Alexa intelligent voice assistant every month, and the number of interactions between users and Alexa reaches billions every week. In addition, Alexa has hundreds of thousands of partners around the world, including China. Alexa can now provide more than 65438+ million applications, covering smart homes, cars, mobile devices, music, voice shopping, taxi software and so on.
Technically, artificial intelligence and innovative technologies such as big data and deep learning continue to support the upgrade of Echo. Speech recognition and natural language understanding are one of the most challenging problems in computer science, which require complex deep learning algorithms and a lot of data and infrastructure for training.
In the field of logistics, Amazon has taken the lead in the transformation of intelligent logistics in the world, and the most subversive measures include intelligent warehousing and intelligent distribution. For example, through the cross-border pre-warehouse in Hong Kong to achieve optimal distribution, through big data analysis, select goods with high repeat purchase rate and store the goods in the cross-border pre-warehouse in advance, thus eliminating the link of international logistics transit in the delivery process.
These innovative examples reflect the technical strength of Amazon, especially the innovation and practice of Amazon Cloud Service (AWS) in big data, artificial intelligence and deep learning.
At present, AWS infrastructure covers 77 available areas in 24 geographical regions around the world. Since the infrastructure investment, AWS has been continuously researching and landing products and technologies to the extreme, providing more than 65,438+075 full-featured services, covering computing, storage, database, network, analysis, robotics, machine learning and artificial intelligence, Internet of Things, mobile, security, hybrid cloud, virtual reality and augmented reality, media and application development, deployment and management.
Constantly optimizing hardware around software applications and creating user value not only provide technical support for a series of revolutionary innovations of Amazon, but also polish product technology around user applications with cloud services as the carrier, and finally realize large-volume and fast operation.
Faster and smarter, all customer-oriented.
At the forefront of innovation, Amazon adheres to excellent operation and long-term thinking, not only constantly explores and applies cutting-edge technologies and trends, but also cares about the experience of end users and conducts technology research and development based on user needs. Because, in Amazon, the purpose of research and development is not limited to the technology itself, but to create the best user experience.
For example, last year's "Black Five" promotion, because of the time difference, consumers in China could start snapping up the stocks of Japanese promotional products on 1 1 the night before, and the promotion was held at 8 o'clock in the UK and 9 o'clock in Germany the next morning. It was already 3 o'clock in the afternoon in China the day after the promotion in the United States. In other words, when foreign consumers are still asleep, China consumers can "get ahead of them" to snap up the explosions.
Products and offers connecting four overseas sites need the support of global promotion synchronization technology across borders, time zones and brands. Let China consumers enjoy more discounts and promotions, as well as the long cross-border shopping carnival, behind which is Amazon's industry-leading global promotion synchronization technology, which is also a model of Amazon China's localized technological innovation.
According to Wang Yi, technical director of Amazon's overseas purchase in China, Amazon applies advanced statistical analysis methods of big data such as artificial intelligence, machine learning and neural network to a large number of intelligent application scenarios such as automatic page generation and typesetting, daily background operation, personalized product recommendation, smart size recommendation and accurate information push.
In the field of e-commerce, "thousands of people and thousands of faces" are common. At present, the way to realize personalized recommendation in the industry is generally the mode that operators label goods or consumers. There are too many human intervention factors in this model. Amazon calculates two vectors of user behavior and product characteristics through technical means, including comprehensive analysis of user browsing history, user search habits, user purchase history, product color and size, and introduces personalized content to customers. Then intelligently create a real "thousands of people" customer experience.
For page layout, Amazon automatically selects channels from the channel pool according to the current customer characteristics and business data with the help of big data system; In terms of accurate product recommendation, Amazon China has built its own unique information flow and waterfall flow recommendation. Inspired by the most advanced computer vision technology, Amazon uses deep convolution neural network to accurately quantify the visual similarity of goods, so as to effectively perceive customers' preferences for clothing products and accurately recommend goods; In terms of recommending promotional information, Amazon's "waterfall stream" will recommend different types of discount information and shopping themes such as baby and spike according to consumers' interests, and at the same time increase the diversity of recommended content to create a richer online shopping experience.
In the background operation, Amazon uses a powerful background system to build an automated operation system, and realizes automatic and intelligent page generation by analyzing the data and models of users, goods and operations.
This year, Amazon China's business process automation project, combined with RoboticProcess automation technology, used smart forms and automation scripts. The manual operation processes involved in the operation of Amazon's overseas purchase system in China have evolved into standardization and automation, and it is estimated that1.8000 man-hours will be saved every year, with an average saving of 1.5% of operation time.
It can be said that Amazon uses global resources to apply innovative technologies to the China market and promote them, while the technological research and development achievements of Amazon China will in turn be applied to the global cross-border online shopping field. For example, the "Smart Size Assistant" was developed by the China team and iterated rapidly, thus playing a leading and promoting role in Amazon's global size recommendation function.
Lightweight operation allows demand to be quickly transformed into services.
Based on the global operation network, Amazon China's overseas purchase has formed a unique lightweight operation mode, which provides a very useful reference for other Amazon countries and leads the rapid development of the global cross-border e-commerce industry.
Specifically, lightweight operations are embodied in two levels:
What others have, Amazon will do better, others have nothing, and what customers want, Amazon has the ability to innovate and land. Under the epidemic situation, when the global operation presses the pause button, cross-border e-commerce has become one of the important graspers to promote economic recovery, and artificial intelligence innovation technology has pushed the world to a new peak of the digital revolution. As Li Yanchuan said: "Internet technology and customer needs are constantly changing, but the only constant is Amazon's pace of innovation."