How to use big data analysis to create business value

Rule 15 - The value of big data lies not in its size, but in its ability to mine

Victor Meyer-Schonberg gave various examples in the book "The Era of Big Data" , all to illustrate one truth: when the era of big data has arrived, we must use big data thinking to explore the potential value of big data.

What is big data thinking? Victor Meier-Sch?nberger believes: 1) need a full data sample rather than sampling; 2) focus on efficiency rather than accuracy; 3) focus on correlation rather than causation.

We believe that big data does not lie in "big", but in "useful". Big data thinking first requires the ability to fully understand the value of data and know how to use big data to provide a basis for business decision-making, that is, to create business value through data processing.

The core of big data thinking is to understand the value of data and create business value through data processing

"Harvard Business Week" pointed out: Data scientists are the sexiest profession in the 21st century. After acquiring massive amounts of data, you must consider how to utilize the data. Data scientists are engineers who use scientific methods and data mining tools to find new data insights. The era of big data highlights the importance of data scientists and the need to combine data analysis with business. When the hardware and infrastructure are available to generate massive amounts of data, someone needs to turn a large amount of scattered data into structured data for analysis, integrate and clean it to form a result data set.

Talent radar is a typical example. Based on the network data left by each person on the Internet, which contains personal information such as his life trajectory, social words and deeds, etc., relying on the analysis of these data, we can peel off his interest map, personality portrait, and ability assessment from his online behavior. Talent Radar, a talent recommendation platform based on data mining, helps companies match people with jobs more efficiently and provides headhunting services. In order to evaluate the professional skills of a technical staff, Talent Radar will use data such as the number of posts on professional forums (such as Github, CSDN, Zhihu, Dingxiangyuan, etc.), the number of content citations, and the influence of the citations. This information is modeled to complete the judgment of its professional influence. At the same time, Weibo data is also fully utilized. The social relationships reflected in it are also one of the factors that determine a person's professional ability. Therefore, judging the professional influence of a user's friends on social networks is also a key point in the talent radar recommendation system. At the same time, even if the recommended person's personal abilities are difficult to meet professional needs, if he has a good friend relationship, he can also be a suitable "recommender" to spread the task to the next level. Different users also have different behavioral habits on social networks, such as the time pattern of posting on Weibo and the length of time spent on professional forums. These behavioral patterns can be used to determine their working time patterns and see whether they meet the corresponding job requirements. Through the integration and analysis of various data sources, talent radar can not only help companies improve the efficiency of talent recruitment while saving costs. Compared with traditional headhunting business, its use of group wisdom can screen talents more extensively and objectively, and its passive measurement method can also avoid the false performance of some job seekers during direct interviews to a certain extent. Its current customers include Taobao, Microsoft, Baidu and other well-known companies.

Amazon received a new patent for "anticipatory shipping" in December 2013, which allows the company to start delivering items even before customers click "buy." The technology can reduce delivery times and reduce the number of visits consumers make to physical stores. In the patent filing, Amazon said delays between ordering and receiving "could dampen customers' enthusiasm for purchasing items from e-commerce." Amazon noted that it predicts sales in a specific area based on earlier orders and other factors. Products that customers may purchase but have not yet ordered, and packaging and shipping of these products. According to the patent, these pre-delivered items are stored at the courier company's shipping center or on a truck before a customer places an order. When predicting "expected delivery" items, Amazon may consider a customer's past orders, product searches, wish lists, shopping cart contents, returns, and even how long the customer's mouse cursor was on an item. The patent shows Amazon hopes to leverage the vast troves of customer information it possesses to gain a competitive advantage.

The most essential application of big data is prediction, that is, analyzing certain characteristics from massive data to predict what may happen in the future. When different data streams are integrated into large databases, the breadth and accuracy of predictions will increase massively.