Operators sitting on large data "gold mine" volume too big disaster excavation

The big data will become the operator "to pipeline" the sharp weapon, recently around this topic, the domestic operator personage discussion is enthusiastic. From the impact of OTT, to "go to the telecommunications" and other thinking, and then to the large data of the operators in the hands of the natural gold mine into a theoretical breakthrough, this logic is logical.

Large data is not the exclusive concept of operators, it has become a common concern of the entire Internet industry. Then the carrier in the hands of gold, what is the gold content, how to dig? Dare to dig? Can you dig better than others? This is a common topic for global operators.

Japan

The privacy issue has docomo a headache for NTT.

Tao told reporters that NTT, Japan's largest mobile telecoms operator, began planning for large data use before 2010 Docomo. DoCoMo not only focus on collecting the user's own age, sex, address and other information, but also the production of fine forms, require users to fill out more detailed information business.

"I have seen such forms, including the number of people in the family, the level of education, communication needs, etc. will be involved." And their advantage is to be able to master the data of the national users, and not like our operators zoned provincial and governance, data collection, the integration of different ways. DoCoMo's CRM system and knowledge base has been greatly enhanced, and there are few silly situations in business transactions, such as a user who has just complained about selling similar products. ”

Despite the high level of information integrity, but because the Japanese society attaches great importance to personal privacy, DoCoMo over the years in large data operations still stay in the planning stage, on how to cross the privacy issue of business or more headaches. Tao said that DoCoMo had developed three stages for the future of large data commercialization: first, the establishment of a database, followed by the establishment of flexible mechanisms, and finally the implementation of flexible, and currently only in the first phase.

European

Operators cautiously open Geographic information services

In Germany, Deutsche Telekom and Vodafone, the mainstream operator, have taken a step toward business models, using large data to serve their own business, Fanghonggang said. The main attempt is through the open API, to the data mining company and other partners to provide some of the user anonymous location data to master the crowd travel rules, effectively with some LBS application services docking. The business model of Telefónica is similar to that of the "smart footprint" data service, which was launched last year, which also provides a reference for retail customers to open shop sites and promotions through mobile network data.

"European operators are also cautious about opening large data APIs, sometimes with strategic intent." For example, the data opening at the same time with the charges, and their cooperation with the start-up of the enterprise, the more lenient policy to support, and for the competitive relationship of OTT enterprises, and so on, the demand is higher. In short, operators are unlikely to provide unencrypted user data directly to third parties. "Fanghonggang said.

States

Verizon aggressively "peddling" data to specific businesses

One of the most aggressive steps by Verizon, a US operator, has surprised Tan Ming and others: Verizon has started to sell its user data directly to third parties through a service called Precision marketinsights. Tan Ming A detailed understanding of the business, Verizon's service is mainly for shopping malls, stadiums and other specific public venues, the collection of mobile phone users background information for the third party.

In the United States, the baseball and basketball game is the audience gathered, the most popular marketing occasions, in the Super Bowl and the NBA, Verizon to the audience from the source of the accurate data analysis, the team to understand the audience's preference for sponsors. In sports competitions, the audience randomly buy tickets, there is no link to the identity information input, through any other way to collect data costs will be higher, the operator's large data in this service is irreplaceable.

It is difficult to dig out the largest amount of "gold ore"

In the communication with the IT times, experts agree that operators are far more complete than any internet company in their grasp of large data resources. "User's attributes, the entire communication consumption data, GPS walking trajectory, login site preferences, frequency and other operators have data, so complete and detailed user behavior data single Web site can not master." Fanghonggang concludes that, because of the operator's "plumbing" characteristics, any detail in the pipeline escapes the operator's eyes. At this level, operators have the largest amount of gold in the hands of large data, such as in the large data business is a natural advantage.

Because the amount of data in the operator's hands is too large, the cost of integrating it into valuable information is also higher. How to exploit the "gold mine" into light products with different attributes is the subject of global operators.

"The reason the current big data applications stay at the retail level, not more complex functions, is because the operators in the hands of data in the end how to use, or depends on the most practical needs of partners." Second, complex requirements are technically or difficult to achieve, or cost too high. "When it comes to technical limitations, Fanghonggang says he has personally conducted research on large data projects and found that the time to extract effective data is very high."

"The Internet data collection relatively simple, such as users in Baidu search a keyword, Baidu backstage at a glance." But the data in the communication network has the physical layer, logical layer, application layer and other hierarchical features, each layer is a mapping method, so the problem arises, the data exist in the logical layer, need to deconstruct them from the logical layer to the application layer, data analysis becomes the protocol analysis. Fanghonggang said he had been on the computer only for a small block of specific areas within an hour of data collection, the results ran for two hours.

"Today is an era of information flooding, when the cost is very high, and the data source is often dispersed in various departments, the operators need to do a lot of work in order to carry out the effective depth of the essence of data mining, so if not from the partners of the very clear demand, and the demand can bring matching business value, Business development could be stalled. "Fanghonggang explained. To avoid internal structural problems that hinder the development of large data businesses, both Telefónica and Verizon set up a large data unit last year, separate from the traditional system, which is one of the reasons for the relatively rapid growth of the two big data businesses.

Gu Hongwen bluntly, based on the above factors, although foreign operators have some ground-breaking application cases, but purely individual, and the initial stage traces obvious: "Strictly speaking, the global operators in the large data commercialization of the mining is at a shallow level stage." The concept is currently overheating, the actual operation is good or not depends on the data holder's operational capabilities. ”