This article is based on my thinking and induction on the practical business operation problems in the process of overall informatization architecture design and informatization development strategy planning and design that I have met in the world’s top 500 enterprises in recent eight years. In this article, firstly, I summarized a series of scenarios demands and the characteristics of operation capacity under the corresponding scenarios; Secondly, I concluded the hierarchical ways of promoting enterprise operation capacity by introducing data-driven models.

For traditional enterprises, we often face the problem scenarios of how to deal with resource sharing, cost reduction, benefits realization, scientific decision supporting and risk management to business entities in overall informatization architecture design process. For internet enterprises, we pay more attention to the problem scenarios of how to deal with content layout, experience optimization, revenue realization of traffic and risk identification to online operation entities. Therefore, we found that there is a general demand of improving data operation capacity in the aspects of operational decision support, operational risk prevention and capacity of refining operational service of the products and services.

As a result, enhancing enterprise management capacity by improving data management capacity gradually becomes the core concern of sustainable development of modern enterprises. The key words of digital enterprises have been mentioned in the 13th National Five Years Plan of China, which highlights the urgency and universality of accomplishing the task. However, it is not easy to make data strategy and data transformation, which needs continuous exploration and improvement. In this article, we think that we must build data operation capacity by taking the route of multi-level improvement.

Through the research and summary of data operation of typical enterprises in series of industries, including national financial enterprises, communication service operators, power operation enterprises, construction enterprises and internet companies, we found that data operation capacity concludes six layers, which are data resource layer, data asset layer, data state layer, data diagnosis layer, data prediction layer and data transformation layer. “Digital enterprises” is formed by the iterative construction of high-end digital buildings and the in-depth integration of enterprise operation models based on business-driven to identify basic data resources and business architecture to identify public data assets.

Figure 1 The Six-layer Pagoda of Building Data Operation Capacity

Definition of Each Layer

The first layer is data resource layer. Data resource refers to data repository that supports business operation through the continuous accumulation of Informatization building. Inchoate business operation support is the Informatization service represented by isolated information system building in respective business domain, which has the problems of business data islands, incapability of sharing data and realizing data transformation. Discrete data repository is an important production in this period.

The second layer is data asset layer. Data asset refers to the enterprise-level data repository gradually formed in the process of data management like data standardization and master data management in various fields under the guidance of enterprise architecture concepts. Informatization in this period is focus on enterprises rather than business. With informatization building, data management like public data structuring and standardization of enterprises can be introduced to achieve data transformation and be updated to be data assets. The basis of digital enterprises is formed in this layer.

The third layer is data state layer. Data state refers to the visual display of data operation processes and results needed in activities of enterprise management and operation. In this layer, there are some high-end data problems in enterprise management level and executive level, such as the problems in the following aspects: the source and classification of various accounts, the distribution of new and traditional business human resources, capital channels and its rationality, the source and controllability of management risk and the ways of measuring return on investment. After the diversified and scaled operation, in particular, the management and control of enterprise operation cannot be implemented just in traditional enterprise leadership and decision-making mechanism. In this case, various complex scenarios cannot be deal with by using any personal experience or capacity. Only through building systematized data operation system, we can gain the practical and real operation state in each stage and help senior executives to make decisions effectively. We can form overall data service views through building data operation management architecture which is consistent with the rules of business operation to achieve the visual display of operation processes and results of classified data assets.Making the current state of enterprises visually, which can form a series of data state services, is essential for the development of digital enterprises.

The fourth layer is data diagnosis layer. Data diagnosis refers to providing credible capabilities of data analysis for business recovery and adjustments of business strategies based on the causal analysis according to the exception and deviation in the process of operation management and control as well as actual business operation. In the process, we can analyze the reasons and scenarios of events in a reverse way and rapidly detect faults or variable attributes position according to management standards of data operation. When data assets entered into the stage of massive scale, it is obviously important to build the capacity of data analysis, which can ensure the analysis and prediction to related events efficiently.A series of data diagnosis services can be formed in enterprises.

The fifth layer is data prediction layer. Data prediction refers to finding the phenomenon and developing trend which are conducive to future reference in the process of data mining according to massive data assets and events. In this layer, ML (machine learning) which is based on artificial intelligence is gradually applied and formed the high-end operation data services, which include the dynamic detection of operation rules and dynamic trend analysis to future operation, by using big data cognitive machine to learn technology in business operation.

The sixth layer is data transformation layer. Data transformation refers to gradually transforming the data service of three layers including data state layer, data diagnosis layer and data prediction layer into data productivity. Data productivity is reflected in improving the capacities in the decision-making support, operation risk prevention and operation service refinement under data service driven. Meanwhile, it is also reflected in helping enterprises to achieve efficiency improvement and structure transformation.

Conclusion and Expectation

In the digital times, the healthy and steady development of China’s enterprises operation must be based on good development of data operation. According to the informatization strategy of China 13th National Five Year Plan, in the future, informatization in China will enter into the stage of digital bonus sharing. With the rapid development of informatization, enterprises must be focus on building data operation capacity to achieve the integration of informatization and industrialization. IT GreenFuture Team is willing to cooperate with you and create a better future together.