The data management platform can be built through reports or BI modules. In the Zhihu column "FanRuan Data Application Research Institute", there are many cases about enterprise data management and BI report platform construction.
From the perspective of project implementation, it can be considered from two aspects: technology and business.
Conduct demand research in advance. A map of construction ideas is listed.
Technically, two main subjects need to be considered, one is the company's data volume, but the specific needs.
Data situation
1. Data sources: What are the business systems that consider the main sources of data? Do you need to build a data warehouse? And preparatory work, such as the preparation of the relevant data dictionary, and even the fetch interface and other issues.
For example, financial data comes from EAS, management data comes from OA, sales data comes from SAP, and database types of various business systems.
2. Data situation: What is the data volume and data increment of the largest single table in the business system. By understanding the data volume and data increment of the largest single table, preliminarily plan the data processing plan, whether to establish an intermediate layer, index, GP, etc., and choose the appropriate analysis platform, whether it is BI or a general reporting platform.
For example, the sales data sub-table has 2000W of data, and the monthly increment is 50W. Then the processing plan can be displayed through the reporting platform FineReport, and the middle layer and index can be established to improve the query efficiency and display performance. Or through the big data BI platform FineBI to establish a front-end analysis, and use the FineIndex method to improve the display performance.
3. Data supplementary record: mainly dimensional data maintenance and fact data supplementary record, the specific work is to fill in the page maintenance. For example, salary analysis in manpower analysis is generally offline financial data, so it is necessary to develop a fill-in and import page for data entry.
This work needs to consider where the supplementary data is recorded: if a data warehouse is established, it is directly filled in the data warehouse; but if the report is developed through direct connection, whether the supplementary data needs to be directly entered into the business system.
4. Data warehouse: Determine whether a data warehouse needs to be established according to the demand subject, analysis indicators and data situation. If financial data and expense data need to be integrated and analyzed, but financial data comes from EAS and expense data comes from OA, then a data warehouse needs to be established to realize integrated analysis. Here, it is necessary to consider whether the corresponding relationship of the relevant dimensional data corresponds to the personnel organization structure effectively after the establishment of the data warehouse.
Demand situation
Demand analysis is mainly about the determination of business indicators, subject modules and analysis scenarios.
1. In the early stage, it is necessary to determine the applicable objects for this platform. Leaders at different levels pay attention to different indicators.
Planning analysis indicators and analysis forms. For example, leaders only pay attention to several major data such as sales and profit margins, while the business layer may also involve drill-through analysis and linkage analysis on the basis of indicators.
2. The main body of demand mainly includes three parts: demand module + module index + analysis form. Through the demand module + module index + analysis form, the demand content is initially framed, the workload is assessed and the quotation is implemented. For example, sales analysis involves analysis indicators such as customer analysis, order analysis, product analysis, and channel analysis. Each indicator may require analysis forms such as hierarchical drill-through analysis and year-on-year analysis.
difficult point
First of all, there must be a good positioning for such a platform - a business-oriented agile corresponding development platform.
Secondly, how to deploy personnel and how to promote the smooth progress of the project? My personal suggestion is that if there are enough personnel, you can try to allow front-line business personnel to participate in or even directly carry out system development, which can not only improve development efficiency, effectively meet business needs, but also reduce the pressure on the IT department. Second, there must be the participation of senior leaders in person to promote implementation from top to bottom.