With the rapid development of cloud computing, big data, artificial intelligence and other IT technologies and the rapid integration with traditional industries, an industrial transformation brought about by digital and intelligent transformation is being conceived.
With the continuous expansion of enterprise scale and business diversification, the service platform of China and Taiwan came into being. The early stage of "China-Taiwan" evolved from the combat system of the US military. Technically, "China-Taiwan" mainly refers to learning this efficient, flexible, and powerful command and combat system. Alibaba released the "Dual Central Station + ET" digital transformation methodology this year. "Dual Central Station" refers to the digital center and the business center.
What is the data center
The data center is a set of sustainable "make the company's data use" mechanism ( it is not a product or project ), a strategic choice and organizational form, based on the company's unique business model and organizational structure, through tangible Supported by product and implementation methodology, build a set of mechanisms that continuously turn data into assets and serve the business.
Data center refers to the collection, calculation, storage, and processing of massive data through data technology, while unifying standards and calibers. After the data center unifies the data, it will form standard data, and then store it to form a big data asset layer, and then provide efficient services to customers. These services have a strong correlation with the business of the enterprise. This service is unique and can be reused. It is the precipitation of business and data of the enterprise. It can not only reduce the cost of repeated construction and chimney-style collaboration, but also the difference. Of competitive advantage.
The data center needs to have four core capabilities: data aggregation and integration, data purification and processing, data service visualization, and data value realization, so that employees, customers, and partners can easily apply data.
The broad data center includes data technology, such as a series of technology collection, calculation, storage, and processing of massive data. The data center discussed today includes data models, algorithm services, data products, data management, etc. It has a strong correlation with the business of the enterprise and is unique and reusable by the enterprise. For example, the company has built 2000 basic models, 300 fusion models, and 50,000 tags. It is the precipitation of enterprise business and data. It can not only reduce repeated construction and reduce the cost of chimney-style collaboration, but also a differentiated competitive advantage.
Reasons for establishing a data center
Compared with the data center, the data center may face a more complicated situation. Reasons for establishing a data center:
-
Big data can tell decision makers some potential laws, and use data to prove or judge decisions. In the past we used data to prove that our decisions were right or wrong. Now we use data to guide us to make the right decisions. In the era of big data, the sample is the whole, and big data can prevent forgery and deviation.
-
Data spawns artificial intelligence. Data is the foundation of artificial intelligence, and can be fused to form new data. Data gives us unlimited innovation, let us keep trying.
-
Data is the robot's instruction, and we form the data service thinking. Data is constantly changing, making machine intelligence a decision-making link, and operations can be intelligent.
The goal of the middle office is to improve efficiency, data operation, and better support business development and innovation. It is responsible for collaboration in multiple fields, multiple BUs, and multiple systems. The platform is a natural evolution of platformization. This evolution brings a "decentralized" organization model, highlighting the ability to reuse capabilities, coordinate control, and differentiate the ability to build business innovation. There are four reasons why data centrality is so important:
1. Return to the essence of service-data reuse
Zhejiang Mobile has taken 2000 basic models as the basis for the development of all data services. These basic models have achieved "the same text and the same car". No matter how complex the data model is, it can always be traced to 2000 basic tables Laid the foundation for data verification and cognition, to the greatest extent avoiding the “wasted costs caused by repeated data extraction and maintenance.”
In the past, there were multiple data extractions for enterprises, one for reports, one for data warehouses, and one for local markets. Regardless of the extraction pressure, maintenance difficulty, and data consistency requirements, they were all very high. At the same time, the unified basic model has made a good aggregation of data in related business fields and solved the demand for data interoperability. This is of great significance. Everyone knows the meaning of data 1 + 1> 2.
2. The data center needs continuous business nourishment
In the enterprise, whether it is a topic, report or access, the current basic chimney-type data production model or project-based construction method will inevitably lead to the lack of precipitation and continuous development of data knowledge, resulting in the model can not really become a reusable component , Unable to support the rapid response and innovation of data analysis. In fact, the most important thing for the business is the stability of the model. If a data model is to pursue stability without change, it will be self-defeating to a certain extent. Such an approach will inevitably lead to the creation of other new similar data models.
The data model does not need to be "stable", but needs continuous nourishment. Only in the nourishment can the initial field be gradually grown into the most valuable model asset of the enterprise.
Taking reports as an example, the reasons for thousands of corporate reports are often not caused by precipitation. For a business report, due to the different perspectives proposed by different business personnel, hundreds of thousands of reports will be transformed. If there is a report in the middle You can put forward some principles of benchmark reports, such as one report per business. Some business reports are only allowed to be modified but not allowed to be added. Naturally, old reports will continue to be improved due to new requirements, which can evolve into The basic report catalog of the enterprise, otherwise it is a pile of reports, the subsequent data consistency problems are endless, management costs increase sharply, and more and more human resources are invested. Such things happen in every enterprise.
3. The data center is the ground to cultivate business innovation
Enterprise data innovation must stand on the shoulders of giants, that is, starting from the data center, it can not always start from the foundation, data center is the guarantee of data innovation efficiency. Those who have studied machine learning know that there is no good regular data, and the process of data preparation is extremely lengthy. This is also a core value of the data warehouse model. For example, the operator needs to obtain 3 months of ARPU data. If there is no fusion model Support, you have to collect and correlate from a single layer of account, and the speed can be imagined.
In today's Internet era, companies are all striving for transformation. The key to transformation is to have the same rapid innovation capabilities as Internet companies. Big data is one of the core driving forces, but having big data is not enough. The ability of data center Often the speed is ultimately determined, and speed of ownership means that the cost of trial and error is very low, which means that you can do it again.
4. The data center is the cradle of talent growth
It turns out that new employees need to grow for employment, one is to bring people, the other is to find someone to ask, and the third is to log in to various systems to see the source code. This kind of learning is fragmented. In fact, it is difficult to understand the whole picture, and it is impossible to know what is for the enterprise. Is the most important, and the documentation obtained is often outdated.
Now that there is a data center, many growth problems can be solved. With the basic model, newcomers can systematically learn what basic data capabilities the company has. The increase in O-domain data gives it a broader perspective and has a fusion model. , Newcomers can know which subject areas, from the subject area to the overall understanding of the company's business concepts, with the tag library, newcomers can get all the wisdom of the predecessors, with the data management platform, newcomers can clearly trace data, The ins and outs of tags and applications, all knowledge is online, the latest, which means a high starting point for newcomers.
More importantly, the data center allows newcomers to get rid of the transitional dependence on mentors at the initial stage, quickly integrate into the team, and innovate on the basis of their predecessors. The natural unity and integration of data in Taiwan may allow newcomers to break the shackles of the dotted line, quickly build their own knowledge system, and become experts in the field of enterprise data.
Of course, the establishment of the data center is not a one-off process. Every enterprise should build its own unique center capabilities based on the actual situation. In this process, some principles need to be followed:
First of all, the organizational structure and mechanism of enterprises need to change with the trend. For example, the department or team responsible for data in the past often lacks the right to speak, and it is often a passively accepted role in the face of business needs. This makes all the data center thoughts into a bubble and requires authorization for the data center team.
Second, we must change the way we work. Nowadays, the main work content of many enterprise data teams is project management, demand management, etc. When a project is completed, it is invested in the next project. After a demand is completed, it is responsible for the next demand. This kind of work is really very training. Organization and coordination ability, but the improvement of this ability and the length of working time are not linearly increasing. Although the experience of project and demand management is increased, it is not possible to get the knowledge and experience in a certain professional field. With the passage of time, more and more people will lose their initial work enthusiasm and creativity. In fact, data personnel only have the deepest research on business, data and models, end-to-end practice to create a data center, is the greatest value Only by creating can continuous innovation become possible.
Third, the data center team should gradually change from the traditional supporting role to the operating role. Not only in the data, but also in the business, we must strive to catch up with the business personnel. The middle office personnel should gradually establish the right to speak to the business, not only to accept the role of demand, but also to be able to make reasonable suggestions and bring to the business New growth points, such as precision marketing.
Finally, the middle office is suitable for the company's characteristics. The most suitable middle stage is when you have a deep understanding of the business, products, systems, and organizations, and not only where you are today, but also how the past evolved and how it will evolve in the future. Only after knowing all the things can we make a better design of the middle platform architecture.
Summary: The data center is to transform business production materials into data productivity, and at the same time, data productivity feeds back to the business, continuously iterating the closed-loop process of the cycle-data-driven decision-making and operation
