Big data cases 5-10: China Mobile, Twitter and Tesco

05 China Mobile's data operation

Through big data analysis, China Mobile can conduct targeted monitoring, early warning and tracking of the entire business operations of the company. The big data system can automatically capture market changes in the first time, and then push to the designated person in the quickest way, so that he can know the market in the shortest time.

Customer churn warning: A customer uses the latest Nokia mobile phone, pays monthly on time, calls customer service 3 times a year, and uses WEP and MMS services. According to traditional data analysis, it may be a customer with very high customer satisfaction and very low probability of loss. In fact, after collecting customer data including new sources such as Weibo and social networks, the customer's actual situation may be like this: The mobile phone purchased by the customer abroad, some functions in the mobile phone are not available in China. Mobile phones are often disconnected at a fixed location, and MMS is not available - his experience is extremely poor and he is at risk of losing. This is the application scenario of China Mobile's big data analysis. Through comprehensive access to business information, it is possible to subvert the conclusions of conventional analysis, break the boundaries of traditional data sources, focus on new data sources such as social media, and obtain as much customer feedback as possible from various sources. Tap more value.

Data value-added applications: For operators, data analysis has great prospects in the government service market. Operators can also make big data technologies play a bigger role in transportation, emergency response, and stability. Operators are in the position of a data exchange center and have an innate advantage in mastering user behavior. As another revolution in information technology, the emergence of big data is bringing new directions to technological progress and social development, and whoever masters this direction may succeed. For operators, in data processing and analysis, it is not only the skills and legal issues that need to be transformed, but also the need to change the way of thinking and think about big data marketing from a commercial perspective.

06 Interests and emotions in Twitter

Twitter interest clustering: Twitter filters a series of customized customer data streams by filtering user attributions, posting locations and related keywords. For example, by filtering the movie title, location, and emotional tags, you can see which of the most popular movies in Los Angeles, New York, and London are. According to the user's personal behavior description, you can even search for Japanese tourists who are skiing in Canada. From this perspective, Twitter's interest map is more efficient than Facebook's social graph. The potential value of Twitter's user data is equally amazing. As social media sites are collecting more and more data, they may be able to find better ways to monetize the data and make it a prime way to increase revenue. The real value of these social networking sites may lie in the data itself. I believe that in the near future, if you find that you can make full use of user data and reasonably avoid threats to user privacy, the enormous energy of social data will be completely opened.

Twitter sentiment analysis: Twitter does not run every data product itself, but it licenses data to data service companies like DataSift. Many companies use Twitter social data to make a variety of amazing applications from social Monitoring medical applications, and even tracking the outbreak of flu, the social media monitoring platform DataSift has also created a financial data product. One of the daily work of Paul Horting, CEO of Wall Street Capital Market, is to use computer programs to analyze the message of the world's 340 million Weibo accounts, and then judge the public's sentiment, and then "1" to "50" Score. Based on the results of the scoring, Horting decided to deal with the millions of dollars in stock. Hunting’s principle of judgment is simple: if everyone seems happy, buy it; if everyone’s anxiety rises, sell it. Some media companies package viewer ratings data into products and resell them to channel producers and content creators.

Once the precise data is combined with social media data, the predictions for the future will be very accurate.

07 Tesco's precise orientation

Smart merchants build models by analyzing their purchase history, tailoring their future shopping lists, and designing promotions and personalized services to keep them paying for them. Tesco is the second-largest retailer in the world, and the British supermarket giant has gained tremendous benefits from user behavior analysis. From the user's purchase record of its membership card, Tesco can know who a user is in the “category” of the category, such as fast-food eaters, singles, families with children, and so on. Such a classification can provide a large market return. For example, a promotion sent to a user by mail or mail can be very personalized, and the promotion in the store can be more targeted according to the preferences of the surrounding people and the time of consumption. Thereby increasing the circulation of goods. This approach has paid off a lot for Tesco, and only one promotion in the market can help Tesco save 350 million pounds a year.

Tesco coupons: Tesco will customize 6 coupons for customers each season. Four of them are items that customers frequently purchase, while the other two are based on the customer's past consumer behavior data analysis, which is very likely to be purchased in the future. In 1999 alone, Tesco sent 145,000 shopping guide magazines and coupon combinations for different segments of the customer base. Even better, such low prices do not compromise the overall profitability of the company. By tracking the return rate of these short-term coupons and understanding the customer's spending in all stores, Tesco can also accurately calculate the return on investment. It’s already very old-fashioned to issue coupons to attract customers, and many of the promotions are actually just to rob the company’s future sales. However, Tesco, which relies on solid data analysis to target coupons, can sustain sales growth of more than £100 million a year.

Tesco also has a membership database that, through existing data, can find those price-sensitive customers and then set a minimum price for the products that such customers tend to buy at the lowest cost level that the company can accept. One of the benefits is to attract this part of the customer, and the second is that you don't have to waste money on other products to cut prices.

Tesco's precision operations: The supermarket chain collects data on 7 million refrigerators in its data warehouse. Through the analysis of these data, more comprehensive monitoring and active maintenance to reduce overall energy consumption.

08 Facebook's friend recommendation

Facebook is a social network giant, but there is not much way to explore the value of big data. It is worth mentioning that friends recommend it. Facebook uses big data to track the behavior of users on their networks, by identifying your friends on its network, giving new recommendations for friends. The more friends you have, the more viscous they are with Facebook. . More friends means users share more photos, post more status updates, and play more games.

09 LinkedIn's headhunting value

The LinkedIn website uses big data to create a link between job seekers and job openings. With LinkedIn, headhunters no longer have to make unfamiliar calls to potential candidates to try their luck, but they can find potential candidates and contact them through a simple search. Similarly, job seekers can naturally sell themselves to potential employers by contacting other people on the site. There are two examples that can vividly represent the value of LinkedIn's data: A few years ago, LinkedIn suddenly discovered that the number of Lehman brothers' visitors had increased recently. At that time, it did not attract attention. Soon after, Lehman Brothers announced its collapse; and announced at Google. In the month before I left China, I found some of the most common Google product managers online at LinkedIn, which is the same. If LinkedIn is targeted to analyze a home

10 Wal-Mart's Data Gene

As early as 1969, Wal-Mart began using computers to track inventory. In 1974, its distribution centers and stores were using computers for inventory control. In 1983, all Wal-Mart stores began to use barcode scanning systems. In 1987, Wal-Mart completed the installation of the company's internal satellite system, which enables real-time, two-way data and voice transmission between headquarters, distribution centers and various malls. The use of these information technologies, which were still niche and advanced, to collect operational data has laid a solid foundation for Wal-Mart's rise in the last 20 years, and has found a link between “beer and diaper”.

Today, Wal-Mart has the world's largest data warehouse, storing a detailed record of every sale in Wal-Mart's thousands of stores in 65 weeks, which allows business people to better understand their customers by analyzing purchases. Through this data, the salesperson can analyze the customer's purchase behavior to supply the best sales service. Wal-Mart has been working to improve its data collection technology, from barcode scanning to installing satellite systems for two-way data transmission, and the entire company is full of data genes. In April 2012, Wal-Mart acquired Kosmix, a company that studies social networking genes, and added social genes based on data genes.