JD.COM Big Data User Preference Model of "Thousands of People and Thousands of Faces"

Statistics on the Order Volume and Display Volume of JD.COM Big Data "Thousands of People"

What is JD? COM's "thousands of faces" "

Behind the "thousands of people" is the support of big data.

"Thousands of people and thousands of faces" is the goal that all major e-commerce companies are pursuing after the arrival of the Internet era, especially after big data has prompted businesses to provide personalized customized services. In the past few years, 3C products (computer, communication, consumer electronics) have been updated most frequently, which has earned enough attention for JD.COM, who once focused on 3C products and won a large number of users.

"At present, JD.COM has set up more than 300 labels to define the characteristics of users, covering the basic attributes, purchasing power, behavioral characteristics, social characteristics, psychological characteristics, interest preferences and other aspects of users." Wang Xiao said. "In my understanding,' thousands of people and thousands of faces' is an important way for websites to provide users with personalized and precise marketing under the guidance of big data. JD.COM has always attached great importance to the user experience. In fact, the focus behind this is the user portrait technology. " For users with different behavior habits and hobbies, there has been obvious difference in label orientation, and JD.COM adopts differentiated marketing methods.

Wang Xiao introduced that user portraits make marketing systems such as search, recommendation and advertising more intelligent to serve users. When different users search at different times, the same search term may have completely different shopping intentions. According to the user's attribute characteristics, personality characteristics or behavior habits, combined with the situational analysis of user behavior, displaying or recommending products that meet the user's preferences can also greatly improve the user's purchase conversion rate and repeat purchase rate.

JD.COM Big Data "Thousands of People and Thousands of Faces" Personalized Recommendation Case

Real-time Interest of JD.COM Big Data "Thousands of People and Thousands of Faces"

User Portrait Service of JD.COM Big Data "Thousands of People"

JD.COM Big Data "Thousands of People" Product Portrait

Problems to be Solved in Personalized Recommendation of Big Data in JD.COM

Big data business is the cornerstone of all businesses.

In addition, it is JD.COM's comparative advantage to cover as many commodity categories and e-commerce data in the whole value chain as possible. Wang Xiao said: "This kind of data is scarce in the industry, which makes the application value advantage of big data in JD.COM stand out. The high integration and unification of big data in JD.COM in data, models, technologies and tools has greatly improved the efficiency of the integration and utilization of big data in the whole group and promoted the deep value mining of big data. "

JD。 The layout of COM on the big data platform has never left the idea of "independent research and development". Wang Xiao said, "JD.COM. COM's layout strategy on the big data platform is to continuously increase investment in JD.COM. COM's self-developed big data platform technology and product system ensure the sustainable development of big data business. " Of course, with the further expansion and enrichment of business scale, JD.COM has not ruled out the possibility of foreign data cooperation. "Through business development and data cooperation, we will continuously enrich and expand the data content system of JD.COM Big Data Platform, and gradually enhance the integration value and deep application of big data. This is also one of our strategies. " Wang Xiao added.

Talking about the current challenges of JD.COM data, Wang Xiao said frankly: "At present, JD.COM. The business of COM is developing too fast, and the company is getting bigger and bigger. How to serve the big data needs of the whole group under the current scale is the biggest challenge before us; Secondly, how to continuously and deeply explore the application value of data in the existing data content is another challenge. Basic statistical report analysis can no longer meet the requirements of decision support. We need to continuously tap the value in the data treasure through a series of technologies and analysis methods to help improve the user experience and improve internal operational efficiency. "