Zhou Fanglei: How to achieve greater value by fusing data

2014 Zhongguancun Large Data day on December 11, 2014 in Zhongguancun, the General Assembly to "aggregate data assets, promote industrial innovation" as the theme, to explore data asset management and transformation, large data depth technology and industry data application innovation and ecological system construction and so on key issues. The Conference also carries on the question of the demand and practice of the departments in charge of the government, the finance, the operators and so on to realize the path of transformation and industry innovation through the management and operation of data assets.

In the afternoon operator @big Data Forum, Zhou Fanglei general manager from the Asia-Letters Marketing Services Department brought the "Cross-border data fusion and trading large data cloud platform, Mixdata to help you revitalize enterprise data assets", explains how asiainfo through the integration of data to achieve greater value.

The following is the full text of the speech:

Zhou Fanglei: Everyone leadership colleagues good, I am Zhou Fanglei, just heard Unicom and mobile two have mentioned the data open, but the data open and really play its business value, we understand that there are still a lot of road to go. This 2014 is my ten years of joining the Asian letter such a day, I have been particularly impressed today, when I first entered the letter ten years ago, there was a particularly large picture posted in the propaganda column of the letter, which is similar to the one below, when our slogan is called the carrier frigate, we regard the operator as an aircraft carrier, We are escorts for aircraft carriers escort. Of course, today we will find that of course, with the carrier business transformation, the operator in addition to the need for a frigate, but also need to have a lot of innovation and breakthroughs, this time I understand why in this year Yashintian under the leadership of the general industry Internet, I understand this is also in line with the operator business transformation of the business transformation of the thinking We realize such a transformation from frigate to delivery machine.

So in the past almost two or three years, in fact, I understand from the industrial Internet slogan before, in fact, we have been doing some of the work, this work I summed up, there are several aspects, the first aspect I call it called technological innovation, the second aspect we call product innovation, the third is called business innovation. I am personally fortunate enough to be involved in many of these tasks. In fact, I still very impressive, is 2009, I think 2009 in the operator should be almost no one to mention Hadoop, when in 2009, in fact, the first operator of Hadoop input production system was built by the letter, was a very small system, is in the total part of the telecom business platform on the virtual platform we set up a Hadoop cluster, we are running business analysis, the volume of data is not very large, of course, more is to do business-divided projects. But in 2009 we found that Asiainfo was already doing the work, and 2009 I was the first to participate in the China Hadoop Assembly, the sixth session of the Hadoop assembly or the seventh session, tomorrow we have colleagues to participate and share. Although said Asiainfo in various public occasions propaganda is not much, but we actually have been in fact with operators to explore. Last year we also helped operators build the first one that could be said to be the largest cluster in the Enterprise Hadoop cluster scale, 100 machines. So just listen to fan always said Unicom already has 1200 machines, we do cluster 1000 range, of course, there is no more than fan total of 1200 machines. I hope that the sub-trust Hadoop will provide more help to operators.

In addition our product innovation Woy purchase, this is we and the Asia letter Guangdong Joint Establishment of E-commerce, this is the vertical field of trading platform. This platform in fact its entire system architecture business architecture and technology architecture is exactly the same as the Internet company's architecture, and its overall system transaction size is also very large, because the whole of its planning is the entire Unicom channel terminal sales, the entire trading platform on this platform to do, So this platform brings together channels, distributors, including Unicom companies, including various agencies, can help agents to provide financial services. So the entire business model business system and the entire technical framework, is entirely the same as the ordinary electric platform is the same.

So we're going to see asiainfo in this field doing an exploration and an innovation by the operators.

The third is doing business innovation, in the second half of last year we and China Telecom Group and Beijing Unicom we set up a large data innovation laboratory in Beijing, this innovation laboratory it is not to build a system and research a technology, it is more to say from the business model of the progress of the new business model around large data innovation, With business partners from all walks of life to achieve cross-border cooperation. This we are also very fortunate to participate in such a project, so participate in this project will have today I led a line of business, big data marketing such a line of business. So we will see that Asiainfo particularly wants to change more aggressively in the days to come, to explore innovation and to help the operators to transform. So today we will also see the past two operators share inside see, just fan the last page of the film Inside said hope that our 2015 focus on cooperation is open to the outside, welcome partners and our common exploration of business innovation. He is also in the last page of the film out of such a share of cooperation and win. So in a day like this, we'll find that operators have never been as open as they are today, or want to be able to merge and cooperate across the boundary, I think this is also the large data itself has the natural attributes, that is, large data must be able to play 1+1 greater than 2 of the chemical reaction, its superposition is not simply added.

In the framework of such a large data lab that was launched at the end of last year, we went to work with our partners, including our carrier colleagues, to analyze the pattern of the whole big data operation. In fact, after the analysis of a circle, we actually feel that the beginning of the feeling is particularly confused state, to a particularly high expectations, and finally our entire rational analysis, whether we reflect on the operators in the past in the field of data analysis, or the current to bat as the representative of the Internet business model, in fact, there is no outside of these ways , the first is the analysis of products, we are familiar with the data cube Taobao, or this morning Baidu's big data colleagues also talked about Baidu's Sinan. The other is precision marketing. Because we know that the whole Internet in the past more than 10 years of the development of the most direct way to change is advertising. Advertising is the way traffic business, with large data has been able to achieve precision marketing. In the most representative of this is the end of 2012, Google's RGB business model, with traffic ads to RGB as the core business model. We have an interactive and easy to exchange in the country, in a moment there are easy to exchange partners for us to share their new development in the field of large data marketing.

The other is the public service, ah just what always talk about very much, how in the tourism, transportation and other aspects of public services, in large numbers of our public services have a very broad space for development. In addition is the credit product, now very hot Internet finance, uses the data to analyze the user's behavior, finally analyzes the user performance the situation. After this analysis you will find that in fact these four things in the past more than 10 years the operator has actually been done, but it is not open to the entire industry. You can see what's on the left? I remember when I first arrived in Assyria, when our entire BIM product line was set up years time, in the past we have not stopped is the assessment Report report, we put the data to make a table, form a report. This is our customer service.

Another is customer group analysis, user interaction Circle Analysis, we know is the students or teachers, or canteen staff, and then we will through the students to do precision marketing, the whole marketing method and closed loop has not fully formed, but it is the whole idea is a complete automatic marketing analysis. Likewise we look at public services, so many years to do operators know that all the public security, or some of the NSA to catch criminals, are indispensable to the operators to seek such data support, such as the call record, adjust his account information, as well as the way to help obtain services. In fact, we can think of it as a public service.

At the same time, credit products, credit products in the past operators have done it? As an operator IT system knows that in the operator Bo SS system, when he has no expense, I want to give him downtime, all based on his past consumption record to do, or say I go to remind you, or say I stop, or say I give you some special amount. This will find that the entire credit control in the operator's entire boss system to do a very systematic, but this system is not open to the outside world.

It is simply too easy for operators to make big data, but is it just a matter of opening up these abilities? But it's not as simple as we thought. Although the whole model is not innovative, but in the overall system architecture and business structure is very different, this is our precision marketing as a point of entry we specifically analyze its challenges in the end where, and the opportunities it faces where. This is what we can see just now, in the past so many years operators have been holding their own data to do precision marketing work, this work how to do, is to data-driven traffic, more channels, SMS channels, offline channels, business hall channels, it can through the portal channel, Of course, the Internet community may call it traffic, now maybe we call it Operation class, this is in fact the entire internet everyone called traffic, to my tools to pass the flow of product data, this is the whole process of precision marketing. Let's see if there's a problem with this inside. If the ability to open up, the challenges he faces. We first look at this opportunity, there is no need to open. We look at two intersection points, the first intersection is from the Internet advertising we call the flow-driven burst management to digital business transformation, and another we see operators from data management to open business transformation. Such a transformation can drive the entire pattern of data evolution.

We believe that the entire data-driven precision marketing model in the first phase we can do is to say that the entire system of precision marketing into third-party data, precision marketing, as well as advertising, is the realization of its own channels. In fact, we can completely through the external data supplement and external traffic to achieve our product launch, as well as advertising budget expenditure. Actually realized such a step we can be included in the third party products, after the third party products we are completely open platform architecture, we can achieve data through, flow through and product through, we formed a flow platform, data platform and commodity trading platform. But in order to achieve this, I think it is very simple in terms of the whole model, but through our team over the past year of analysis and practice, we will find that there are really very big, very many challenges. First we go to the challenges of the business, but in the past when we talked about business terminology, I learned that within the operator, I stayed for October years, no one told me about CDC, Cdm,cds and CDA, no one told me that more people are talking about my marketing case, how do I evaluate it, What my clients do with their photos, I find out in this assessment that he's doing architecture design in a way that flows.

The second is the system architecture, which used to mean that operators build an operational architecture platform to achieve marketing control and marketing analysis, but when we implement such an open system, we cannot run our entire system according to our past architectures, when we find DMP, The structure of DAP comes out and I have to absorb the product in this way. To do business and system architecture adjustments. So just see two operator leaders said, we in the entire infrastructure above our operators do a lot of research, accumulation and practice. But from a marketing point of view, the cost is more than technology driven, more business-driven issues. We look at the Internet advertising trading platform is basically in milliseconds to calculate its advertising effect, in the calculation of its expenses, when computing its entire RY technical framework, we can only rely on the RY technology framework to achieve such a business goal. So I just talked about the data to get through, the flow through, we have to solve the technology, systems and business three challenges.

One of the most fundamental challenges behind these three challenges is the challenge of the data, is it that when we put the data open here, it quickly becomes the good business model we expect? We actually found that it can be said that this time last year, at this time today, I feel very different. After such a year, I will find that the challenge is too big, it is not what we think it is very simple to find a partner, immediately can be realized data assets. In the past we called the way and the private way to determine the data. We used to talk to operators about data, speak A, B, B field, M field, completely understand, and many people do not know the operator's data distribution in 31 provinces, and the province of the data is placed in different departments, may be placed on different ports, completely do not understand. They may feel that the carrier data is not the place.

Also when we communicate with our carrier colleagues, we will find someone else marketing, outside some advertisers he is looking at the data, they will say I need what kind of crowd, his purchase tendency is what, he personally have what concern, his geographical distribution is what, this we use pattern to map, We will find this difficult to show, the gap between the very large. How to solve this problem? So I think in the open way, if we open up, we set up an open direction, we are in the marketing of the Internet to view the pipeline large data, we look at this way will find that the first solution is the user identification problem. This morning a lot of leaders also talked about how users can identify how to circumvent privacy issues. In here, the letter to provide a solution, called Mxidea solution, this is our entire proposed framework, this mxidea can not correspond to a person, but through the virtual people, to provide him with the information preferences, through the marketing dimension to the virtual person tag, In this we look at the entire data and data fusion, we will find that the whole system is very complex because of the pipeline data, and because of the cost of such a system is very very large, how we solve the problem of user identification, but also to circumvent user privacy issues.

In fact, once the technology architecture of Miidea's virtual ID is realized, we will find that we have built up a basic bridge of communication, and we can realize the next step after this basic bridge. When we get to this point, we start to analyze the data in this ad, we will find that there is a popular word is called High latitude civilization, in the past there is no third party data in large numbers through cookie data, but the cookie is a valid area problem and cross-domain problem, A user behavior cannot be fully opened in all areas. So we can act as a large data architecture for pipeline data and MXID, and we've built a higher dimension, that's the dimension of ID. When our partner is easy to communicate with Chiang Kai-shek will specifically talk about the problem of crossing the screen. When we find miid, how easy it is to solve the problem of crossing the screen. The elimination of you has nothing to do with you, when you are still struggling to solve the problem, there is another very simple thing to solve this problem. If we have this architecture, we will find that the value of operators is just beginning. In the past, we knew that in the whole big data or in the whole internet marketing, people's grouping can only be graded according to People's preference, there are only a few dimensions, but we will find that when we go back to the entire trajectory of human life, we find that no one is an individual person, He will have a close relationship with his group, and this group has a very close relationship with your position. When we have such a communication data base, you will find that when we go to carve individuals and carve out his social behavior, we will provide a closer to the actual Internet products. For example, the family, such as the business district, the park, the school, the hospital, in the past in the Internet marketing to mention very little, but this concept in the mobile internet era to mention a little more, but has not mentioned internet marketing. When we have this source data, we will find that we will increase the overall marketing process. We will find that the marketing approach is entirely in the form of scenes to drive marketing action, at this time we said we can think that we in the Internet marketing it completely into a completely high level to do, which contains a very large space for imagination.

Here I take a small example to look at the completely different changes it brings. This project is our and WGV a chronic Disease prevention project team, this research project, they are expected through in the operator comprehensive data inside we according to family such a dimension to come and analyze a family inside, long time, For example, a year or two years of such a span. How do we analyze a family's chronic disease the whole evolution of it, this is standing in a province or a city in the dimensions of the project analysis, and then we will release the analysis report.

Why do I cite this example? Because this example has two key attributes, in the past Internet marketing can not get the attributes it is easy to show, one is the family attribute, one is that he for a very long period of time for such a person or a family within a specific analysis. In this example we will see a value of this new data source. So Asiainfo is also in order to be based on the asiainfo of third party data open, we put forward for large data transaction large data platform, Mxdata, we are now in the direction of data marketing, to large data marketing as a starting point, and gradually realize the carrier data based on the integration of data areas and innovative service platform.

and through the development of the past year, we in the data marketing or in the industry partners, we have accumulated a very good partner, we also help partners and operators to establish communication bridges. Last year, when we were communicating with our partners in the company with a client, we also have a bit of doubt asiainfo in the role in the end is what role, in fact, through the development of a year we will find Asiainfo in the bridge and the role of the link is very large, we help external enterprises understand the operator's data , we help him to play such a bridge and the link of such a role, we will find that there is actually a huge communication and cultural gap, Asiainfo hope to undertake such a messenger to achieve such a bridge and link. Well, thank you.

(Responsible editor: Mengyishan)