Musa put forward "big southern platform, small front desk". The sales department of southern Taiwan includes search sales department, shared sales business network platform, statistical data control technology and marketing department, and statistical data control technology and marketing department should It is the core department responsible for the construction of Statistics Southern Taiwan. So, what exactly is Statistics Southern Taiwan? What does the specific content include? What is the close relationship with the big statistics network platform? How is it reflected at the structural level? What is the close relationship between statistics and commodities? . The head of Musa's statistical data control technology and marketing department mentioned the specific meaning of the statistical data in Nantai. Here is what he said: "Many people can think of statistical data as "oil", and Teacher Ma (Jack Ma) also said However, Musa Baba wants to become the "water coking coal" of global e-commerce. The statistics South Taiwan they are building now hopes to play the role of "hydraulic power plant." "They know that the development of electric power can be divided into several stages. Some potential private enterprises have their own hydraulic power. Later, Jieyi industrial production capacity appeared. Some private enterprises can’t afford Hardoi, and some can’t use it. At this time, national institutions will come out and build a national power grid. Or wind power and water power, water power and water power, to ensure the electricity demand of different groups to the greatest extent." "Their statistical data Nantai also has such an operation idea, and the actual effect they have achieved is a relatively small one, which is divided into four parts from bottom to top—"" The first is that when the statistical data control technology did not have statistical data at Nantai, we had our own statistical network system, control room, and small statistical database whether it was within Musa or in each store. But when the statistical data accumulated to a certain scale, this aspect The cost will be very high, and the quality of statistical data is different from international standards, which will lead to problems such as low work efficiency. Therefore, they need to collect, calculate, store, and process massive statistical data through statistical data control technology , and at the same time unify international standards and naval guns." "The second is statistical data financial asset statistics. After Nantai unifies the statistical data of the Musa Department, it will form international standard statistical data, and then store them to form a large statistical data financial asset layer, and then Guarantee to provide efficient services for the sales businesses and businesses of the group." "The third and fourth are statistical application field services, including service businesses and service items, such as chief of staff and Musa index, which are statistical data Taichung's "statistical application field service project" for merchants "statistical data South Taiwan service project Musa, in the final analysis, is more about serving merchants. The network platform will ensure that we use statistical data in the process of using naval guns and international standards. , timeliness, and work efficiency are guaranteed, and can have higher reliability and stability." The above seems to be correct, but some logic cannot be formalized. For example, the statistical data control technology here is similar to that of Musa What is the intimacy of the cloud's statistical data control technology? Statistics South Taiwan should undertake the research and development of network platforms and tools such as hadoop/ETL? The chief of staff is an end-to-end commodity, it seems that it cannot be classified as a statistical data station? In short, from the perspective of responsibilities, as the responsible department of Southern Taiwan, it is indeed necessary to serve first-line customers directly based on commodities, rather than stepping back. This is also the biggest confusion about the statistical data of Southern Taiwan in this column. I have been thinking about the performance of the responsible department of Southern Taiwan. How to decide? How to interpolate and optimize Nourishing South Taiwan without sales business, Musa answered the question. However, it is obviously unreasonable to treat indirect commodities as southern Taiwan. Musa mentioned statistical data southern Taiwan. Those who are busy may be those who do statistical data structure and statistical data processing, because the structure pays attention to strict logic, and the essence and boundaries must be defined. Clear, without meaning, otherwise you will be at a loss when you do things, and you don’t know how to start. For example, one day the leader will ask you if there is any statistics on their private enterprises in Nantai. You should learn from Musa. With a clear basic concept, you will be able to You can do morphisms, otherwise you will appear at a loss. There are actually many such things. The private enterprises in this column were doing IT planning some time ago, and many people have objections to the statistical data that Nantai has to bring some commodities. I remember that this column used to regard the network marketing network platform as Nantai, claiming that it also relied on all network marketing personnel. This is a problem caused by unclear basic concepts. It is strange to say that it is difficult to find a more scientific explanation of the statistical data of Nantai on the Internet. I happened to be thinking about this problem some time ago, and I would like to share it with you. In short, although the flaws are not concealed. The so-called Southern Taiwan of Statistical Data means to realize the stratification and horizontal decoupling of statistical data at the same time, and to crystallize the potential of public statistical data. Data visualization simultaneously realizes Bruma statistical data integration and knowledge crystallization, realizes the PCB and opening of statistical data through statistical application field service items at the same time, quickly and flexibly meets the requirements of the upper application field of users, and satisfies the needs of users through statistical data application field software See the figure on the right for user personalized statistical data and the needs of the application field (taking a certain operator as an example): 1. Statistical data source The statistical data source is hierarchical, which used to be called the statistical data warehouse model. This column is summarized as three Layer, the basic model is generally the visualization of intimacy, and it mainly realizes the international standardization of statistical data at the same time. They are called "books with the same text, cars with the same track". It can be summarization, association, and analysis. The mining model is actually partial to the application field, but if there are many people using it, you can also use the mining model as the knowledge crystallization of private enterprises to Nantai. For example, the model retained from the grid has If there is a lot of commonality, someone should adjust it to the Nantai model so that it can be used by others. Nantai's model is relative, and there is no absolute international standard. 2. Service items in the statistical application field The statistical data source is made into a service project PCB according to the requirements of the application field, which constitutes a service project in the statistical application field. The basic concept of this service project is exactly the same as that of the sales business Nantai, except for the statistical data PCB. It is a little more difficult than general functional PCBs. After all, OLTP functions have limited changes, and statistical data analysis is greatly affected by market factors and changes faster, making it more difficult to service project PCBs. With the in-depth operation of large statistical data in private enterprises, various large statistical data application fields emerge in endlessly, and the demand for service items in the statistical application field is very urgent. If large statistical data is not serviced, it cannot be scaled up. For example, Zhejiang Mobile PCB Provided a total of hundreds of service items such as customer insight, location insight, network marketing management, terminal insight, financial credit investigation, etc., with a monthly call volume of more than 100 million times, flexibly meeting the user's requirements for service items in the field of internal and external statistical applications. 3. Statistical data development But statistical data sources and statistical application field service items are far inferior, because even the best ready-made statistical data and service items often cannot meet the user's front-end personalized requirements. At this time, it is necessary to teach others It is better to teach a man to fish than to fish. The last layer of Statistics Nantai is the development of statistical data. It is also divided into three levels according to the difficulty of development. The simplest is to provide a tag library (DMP). Users can quickly form a network based on the assembly of tags The marketing customer group is generally oriented to sales personnel, followed by providing a statistical data development network platform based on which users can access all statistical data and perform visual development, generally oriented to SQL developers, and finally to provide application domain environment and Components, so that control technicians can independently create personalized statistical data products, progressively, to meet the requirements of users at different levels. It is controversial whether the tag library (DMP) belongs to SaaS or PaaS, but the network platform such as tag library is obviously a little more southern than the chief of staff products, because it is more versatile, and the characteristics of the exclusive sales business are not very special. Obviously, this column still thinks that it can be classified as Nantai. It should be said that the components in the development of statistical data, such as page components and visualization components, seem to be more reasonable to be attributed to sales business Nantai, but it also depends on the actual situation of private enterprises. , It can be attributed to where it is used more, there is no absolute international standard. The above division method still makes sense logically, but there are still many that have not been considered, such as algorithm service projects, machine learning engines, hadoop, MPP, etc., this column believes that algorithm service projects should belong to the statistical application field service projects One type, but Hadoop, MPP, and machine learning engines are lower-level, and should belong to the category of private cloud or public cloud. For example, in this column, you can see that Musa Cloud provides machine learning services such as MaxCompute. The stratification of statistical data in Nantai seems simple, but this column has been entangled for a long time, and many boundaries are blurred. A book I read a while ago mentioned that if the basic concepts of Jieyi do not match the existing knowledge system, you must Try to figure it out, you can’t copy what others say, as long as you can express your own point of view, even if you are still wrong, you will have a chance to be corrected by others. The lack of in-depth understanding of things is mostly the result of unclear basic concepts caused by lack of understanding. Some time ago, new retail was very hot, and all kinds of great sages came out to explain the basic concept of new retail. We can think about what new retail is?