One thing that many people may not have thought of is that the concept of business intelligence BI has been developed for decades. During this period of development, business intelligence BI has formed a set of mature theories and product systems, and with the addition of modern informatization and digitalization, it has become a mature product for enterprises in various industries.
1. Business Intelligence BI
The definition of business intelligence BI is actually very simple. A brief summary is that business intelligence is a complete set of data-based technical solutions composed of data warehouses, query reports, and data analysis. It can realize the standardization and flow of business processes and business data. It integrates and standardizes different business information systems such as ERP, OA, and CRM, and integrates and summarizes enterprise data.
Business Intelligence BI - Pico Data Business Intelligence BI Visual Analysis Platform
Business intelligence BI is a very comprehensive data-based technology solution. For example, business intelligence BI can produce data visualization reports that satisfy employees of different departments and levels, and can help front-line business personnel to achieve business tracking, forecasting, and review operations; Help the senior management of the enterprise to obtain comprehensive information of the enterprise through the management cockpit of business intelligence BI, core KPI indicators, group Kanban, etc., and assist in decision-making.
To summarize the core content of business intelligence BI, there are roughly three characteristics:
- A complete set of data technology solutions consisting of data warehouse, query report, data analysis, etc.;
- Open up and effectively integrate the data in different systems (ERP, OA) in the enterprise;
- Use appropriate query and analysis tools to quickly and accurately provide reports to provide decision support for enterprises.
Business Intelligence BI - Pico Data Business Intelligence BI Visual Analysis Platform
Business intelligence BI is indeed very important in an enterprise. This is because business intelligence BI plays a connecting role in the enterprise. Looking down, business intelligence BI can open up different business information systems such as ERP, OA, and CRM, and convert the cleaned data Unified storage in the data warehouse; looking up, business intelligence BI can provide data visualization reports of different themes and forms, comprehensively display the development status of the enterprise through data visualization analysis, and assist managers in making decisions.
Business intelligence BI can be divided into three levels according to different stages of the enterprise data life cycle:
The first layer, visual analysis and presentation layer - the demand layer of business intelligence BI, on the one hand, represents the needs of users, what users want to see, what to see, and on the other hand, it also represents what users want to analyze. These are in this layer to show.
The second layer, data model layer - business intelligence BI data warehouse, is mainly responsible for the analysis model of enterprise data, and completes the transformation from business calculation rules to data calculation rules.
The third layer, data source layer - the data layer of business intelligence BI, the business information system of different departments and business lines, the data of the underlying database is extracted into the data warehouse of business intelligence BI through ETL, modeling analysis, etc., the final support To the front-end visual analysis display.
2. The position of business intelligence BI in enterprise IT informatization
As mentioned above, the position of business intelligence BI in an enterprise is mainly a link between the past and the future, and it is an important part of informatization construction. Business intelligence BI has formed a complete set of data value system around data, giving full play to the value of data in enterprises.
When it comes to the position of business intelligence BI in enterprise IT informatization, we must first understand what the enterprise informatization is. Generally speaking, the informatization construction of enterprises is universal, and the IT informatization of most enterprises can be divided into two stages: one is business informatization, and the other is data informatization.
Enterprise informatization - Pike data business intelligence BI visualization analysis platform
These two departments are independent of each other and influence each other, but in general, data is still regarded as the basis. Business informatization generates data and optimizes the business itself through data. Data informatization uses data but can also let data play a role and optimize Business informatization.
Business informatization - ERP, CRM, OA and self-built business information systems used by enterprises are collectively referred to as business informatization. The main function of business informatization is to optimize and adjust the business process of the enterprise. Through standardization, standardization, and online, it can improve the efficiency of business operation, reduce the cost of manpower, time, and energy, and accumulate a large amount of business data. It is the core of the business management idea. It is also the embodiment of modern enterprise management.
Data informatization - Big data, business intelligence BI, data analysis, data mining, etc. that we often hear are collectively referred to as data informatization. Data informatization can help companies fully understand their business management, adjust their business management model from experience-driven to data-driven, reduce subjective influences such as emotions and psychology, form data-based business decision-making support, and improve the accuracy of decision-making This is a higher level of enterprise management.
Enterprise informatization - Pike data business intelligence BI visualization analysis platform
The informatization construction of an enterprise is a complete process. Without the construction of a business system, there will be no data accumulation, and without data accumulation, an enterprise will have no basis for deploying business intelligence BI. This is the two-way effect of business informatization and data informatization, which can enable business systems to promote the deployment of business intelligence BI, and also allow business intelligence BI to improve the effectiveness of business systems.
3. Who are the main users of business intelligence BI?
Business Intelligence BI - Pico Data Business Intelligence BI Visual Analysis Platform
The main users of business informatization - the main users of business informatization are front-line business personnel, so the users of business informatization are mostly from the business perspective, entering data, recording processes, viewing business information, etc. for the business.
The main users of data informatization - the main users of data informatization are management decision makers. In the daily process of enterprise operation and management, decision makers use business intelligence BI and other data technology solutions to locate more from the perspective of management problem, analyze the problem, and finally form a business decision.
4. What exactly does the data island mean?
When an enterprise develops to a certain level, due to the increase in the amount of data and the necessity of informatization construction, the enterprise will build corresponding business information systems for different departments. These business information systems (ERP, OA, CRM) can standardize business processes, form a standardized business model, and automatically accumulate business data through the system database to accumulate data assets for enterprises. There is no doubt that at the moment when the value of data is prominent, it is of course a good thing to be able to accumulate business data.
However, the data in the databases of these different departments and different business information systems are often not interoperable, and can only be stored in their own databases, which cannot be used in a unified manner, and there is no overall perspective for the enterprise as a whole. In this way, the data of each department and each business system are separated from each other, like isolated islands overseas, unable to connect and communicate with each other. This is the data island that is often heard.
Business Intelligence BI - Pico Data Business Intelligence BI Visual Analysis Platform
As a data technology solution, business intelligence BI can solve the "data island" and "information island" problems faced by enterprises through data informatization, data warehouse and data visualization when facing the problem of data islands. Therefore, business intelligence BI needs enterprises to Senior managers plan and mainly provide decision-making information for managers at all levels of the enterprise to assist in decision-making.
Management Cockpit - Pico Data Business Intelligence BI Visual Analysis Platform
Therefore, when introducing business intelligence BI, it is necessary to understand the needs of different personnel. From the perspective of different employees in the enterprise, some people think that there are isolated data islands, which must be resolved. Some people don't think that there are data islands, or even if they exist, they don't have much impact on them, so they don't need to solve them. The root cause is that they don't know the real service objects of business intelligence BI.
5. The way of business intelligence BI to fetch data from the business system
Business intelligence BI retrieves data by accessing and connecting to the data source database of the business system. No matter what type of database it is, business intelligence BI extracts the original table data of the business system through ETL to the data warehouse for processing, and finally Supported to the front-end visual analysis report display.
Business Intelligence BI - Pico Data Business Intelligence BI Visual Analysis Platform
A friend asked this question before: Does the data source layer need to develop an interface?
In fact, generally speaking, it is not necessary. Basically, those who ask questions like this have experienced the interface docking of software systems. The interface docking of software systems is because some business software is developed by JAVA, some are developed by .NET, and some are B/S Architecture, some are C/S architecture.
The interface between software systems requires the participation of development, which is mainly to connect the business processes of different software. This interface needs to move the code. But business intelligence BI has a different interface for obtaining data, which has nothing to do with the business system software itself. It only needs to access and connect to the database behind the business system, and directly fetch data from the database, so no software interface is required, or There is no software interface to access this concept.
Unless in one case, the business system is a public cloud, pure SAAS mode, in this case, the data can only be accessed through the API interface opened to the outside world by the software.
Performance Analysis of Salespersons in a Pharmaceutical Industry - Pieco Data Business Intelligence BI Visual Analysis Platform
Channel Terminal Management Analysis of a Liquor Industry - Pike Data Business Intelligence BI Visual Analysis Platform
6. How should we understand the relationship between data center, business intelligence BI, and big data?
When the system's business intelligence BI encounters large data volumes and unstructured data processing scenarios, the underlying data warehouse is upgraded to a big data data warehouse architecture. This is the business intelligence BI analysis under big data; On the basis of the warehouse architecture, the data collection capability is further expanded to the left. In addition to the data warehouse modeling of the original big data architecture, the concept of data assets, data asset inventory, and data asset management are added in the middle. The capability of data service has been expanded, and the data processed by Taichung in the data center according to certain rules is packaged to provide external services. Therefore, data acquisition, data warehouse modeling, data asset management and data services under the big data architecture constitute the core of the data center.
Data Visualization - Pico Data Business Intelligence BI Visual Analysis Platform
The foundation of the data middle platform is the big data architecture. The data warehouse is a big data upgrade of the traditional business intelligence BI data warehouse, and the business intelligence BI becomes the application layer above the data middle platform, using the data service of the middle platform to obtain data. analysis display.
This is the relationship between business intelligence BI, big data, and data center and the evolution process in different data scenarios and service scenarios. If you understand this process, you should not easily confuse their concepts. As for which business intelligence BI, big data, and data center should choose, in fact, how to choose the appropriate technical route and technical architecture ultimately depends on what the enterprise itself wants to solve, and cannot choose blindly. The result of blind selection is a large investment, but a small output does not meet the expected expectations. We should still focus on the demand itself, and demand is king.
7. Several major misunderstandings about business intelligence BI cognition
Many enterprises use business intelligence BI as a pure reporting tool, and the output form has become a visual chart, but the content displayed in the chart is still the previous departmental business information, which only shows the basic situation of the front-line business department, and managers still need to spend a lot of money. Time and energy to understand the overall development of the enterprise.
Business Intelligence BI - Pico Data Business Intelligence BI Visual Analysis Platform
Here I have summarized some misunderstandings that everyone often encounters in their understanding of business intelligence BI:
1. Business intelligence BI is report visualization, which is a bunch of visual charts, and business intelligence BI is front-end visualization.
2. Business intelligence BI is a drag-and-drop analysis tool product.
3. Business intelligence BI is business intelligence BI and has nothing to do with data warehouses.
4. With business intelligence BI, there is no need for data warehouse modeling, and business personnel can do business intelligence BI analysis by themselves, and they can drag and drop to do business intelligence BI analysis.
5. Business intelligence BI is business-driven and does not require the support of IT personnel. Agile business intelligence BI does not require IT intervention.
6. Isn’t business intelligence BI directly connected? Analysis can be done without directly connecting to the data source, without the need for a data warehouse.
First, briefly correct the understanding of these issues.
1. Business intelligence BI is report visualization, which is a bunch of visual charts, and BI is front-end visualization.
Business intelligence BI is a complete set of data technology solutions consisting of data warehouse, data analysis, data reports, etc. In a BI project, 20% of the time is spent on front-end analysis reports, and 80% of the time is spent on the underlying data Warehouse design, ETL development, access development, etc.
So the visual report is only the final presentation of BI, but not the whole of BI.
2. Business intelligence BI is a drag-and-drop analysis tool product.
To be precise, drag-and-drop visual analysis tools can only solve part of business intelligence BI, that is, visual analysis. But in fact, the scope of technology included in business intelligence BI is relatively wide, involving all aspects from bottom-level data acquisition to front-end display and analysis.
Strictly speaking, the simple drag-and-drop business intelligence BI visual analysis tool can only be positioned at the individual and departmental level, which is very different from the enterprise-level business intelligence BI, so the simple previous business intelligence BI analysis tool cannot play a business intelligence BI The real role of BI cannot replace the position of business intelligence BI.
3. In the past, some people always said that business intelligence BI is business-driven, business intelligence BI is BI, and has nothing to do with data warehouses.
This question is very in-depth, and I thought so before. I always feel that with BI, data warehouse modeling is not needed, and business personnel can do BI analysis by themselves, and they can drag and drop to do BI analysis. No support from IT personnel is required, agile business intelligence BI does not require IT intervention, and does not need to build a data warehouse.
Management cockpit large screen - Pike data business intelligence BI visualization analysis platform
As long as there is any business intelligence BI sales or pre-sales, tell users that your business intelligence BI project does not need to build a data warehouse, and can handle all the analysis in the enterprise directly through the business intelligence BI analysis tool, without the support of IT personnel. , Business personnel can do it by themselves... Those who dare to promise like this, either do not understand business intelligence BI, or are really fooling around.
In the construction of enterprise-level business intelligence BI projects, it is really possible to achieve data visualization analysis by simply dragging and dropping by business personnel. At least in my personal work experience of more than ten years, more than 95% of enterprises have can't do it. The key companies I have served include: SHP (Security Health Plan), Microsoft (China), Microsoft (USA), VWFC (Volkswagen Finance), etc.
What VWFC has done is very good. A few business personnel make a lot of reports by themselves, running thousands of reports online. Why? Because the underlying data warehouse has been built for many years, the underlying data architecture is relatively standardized. Business Driven business-driven, what is its premise?
1) The quality of the underlying data is very standardized, and the data warehouse structure is very complete. Business personnel are not allowed to touch the underlying data. ETL, data fetching, index calculation, etc. are all maintained by the IT department.
2) After training, business personnel must be proficient in the use of business intelligence BI front-end reporting tools, and they must be very familiar with the data analysis model interface released.
3) Business personnel must be very familiar with business and data.
Articles 2) and 3) have no problems for many companies. Article 1) can be solved by directly using a front-end business intelligence BI tool for business personnel to solve? Obviously business people do not have this ability.
This is when it comes to training, business intelligence BI tools are very simple to use, but once it comes to the actual enterprise business intelligence BI project development, it is difficult to move forward. Because during the training, the data tables given are all selected, and they are always high-quality and standardized. You only need a simple left table and right table, such as sales order table and order details table. Naturally, it is easy to give visual reports to Realize it.
Data Visualization - Pico Data Business Intelligence BI Visual Analysis Platform
However, in the analysis of actual enterprise business intelligence BI projects, the calculation rules of analysis indicators cannot be solved by simply associating a few tables. If you don’t believe me, you can challenge an actual indicator calculation logic: challenge a small case of ETL data cleaning in the database There is only one data table, and the data is easy to understand, but many business intelligence BI developers need to waste a lot of energy to do it, let alone business personnel self-service business intelligence BI analysis.
Talking so much is not to blindly deny the role and capabilities of self-service business intelligence BI. Self-service business intelligence BI has its usage scenarios, and it has indeed helped us simplify a lot of BI work, but from a professional point of view, it is particularly disgusting. Business intelligence BI vendors repeatedly reinforce concepts similar to this to the market in an irresponsible way: business intelligence BI is a visual report, business intelligence BI does not require data warehouse modeling, traditional data warehouse modeling is very backward, business intelligence BI It is self-service analysis, business intelligence BI self-service analysis is very simple, and business users can learn it in a few days of training and analyze how they want...
From the perspective of marketing and sales, there is no problem in simplifying the complexity of the product and the difficulty of getting started. The problem is that a wrong and unprofessional explanation eventually misleads the company to accept these incorrect concepts. , and use these incorrect concepts to evaluate and plan the construction of business intelligence BI projects, and did not fully anticipate the challenges and risks that may be encountered during the construction of business intelligence BI projects, which eventually lead to unsuccessful and failed projects and repeated construction .
One of our clients in Beijing spent more than one million yuan on a so-called business intelligence BI project. The project went online for about a year, but in the end it couldn't be pushed at all and failed. Later, we found Pike Data, and we provided them with Pike Data's business intelligence BI analysis platform. We have done this project for several consecutive phases, and the customer even wrote a thank you letter.
Why couldn't push it before, and the project would fail: I didn't pay attention to the planning of the data warehouse. Because their business is continuous and changing, the annual demand needs to be adjusted dynamically, the data continues to increase, and the depth and breadth of analysis are constantly changing. Without a good underlying data structure to support, only relying on SQL to fetch It is impossible to support the changes in business analysis requirements of an enterprise in the next 3-5 years or even longer in the form of data collection and report creation.
8. How did the reporting tool come about?
I have been in the technical field, informatization field, and business intelligence BI industry for more than ten years, and I have never left this circle. Has done JAVA (AWT, SWING, JSP, Hibernate, Spring, ibatis), .NET (ASP, http://ASP.NET , C#.NET), Object-C, JS and other technology development, business software system platform development .
The early front-end technology was very weak, and the implementation of AJAX also required handwriting. To implement click-to-edit and modify data in a form, you need to use JS DOM to operate. Report making is basically JSP and ASP scripting languages nested in HTML at the front end for circular output. The report style is very native and ugly, and the slightly more complicated table report style needs to be adjusted with JS.
The reports used at that time, such as Crystal Report, Runqian Report, etc., had tags in the front-end scripting language that could be directly referenced, and report generation replaced a large number of handwritten codes. The early front-end and back-end technologies were inseparable. http://ASP.NET was slightly better, and the front-end gradually had some integrated controls that could be used directly, but JAVA really did not. When was the stage mentioned above? Around 2005 and 2007, I think it has been widely used. There should be many original report label posts on the old CSDN.
There is also a certain market in China for old reports such as Jinfeng Report Jreport, Star Report StyleReport, etc. As early as 2010, the revenue of some report manufacturers had exceeded 100 million, indicating that the basic report market is still very good.
What was the positioning of the report at that time? It was a pure report report. Through the program, the data aggregation list returned by querying from the background database was bound to the front-end script page to generate various reports, which were actually used in various business software. The report display in the system is far from the level of business intelligence BI analysis.
And there are still a large number of software developers who actually have strong report capabilities, but these report capabilities have not been independently used as report products to operate on the market.
Gradually, with the improvement of front-end technology and front-end framework, from the traditional table technology to the visual display of various histograms, bar charts, and pie charts, at this stage, the boundaries between reports and business intelligence BI become more and more blurred . Why? The report display capability of business intelligence BI is roughly equivalent to that of traditional reports, and the ability of self-service analysis and self-service drag-and-drop to achieve rapid multi-dimensional analysis has not yet appeared.
What I want to say after talking so much is that many business intelligence BI projects we have seen are implemented with report thinking, that is, from SQL to data sets to front-end display. And what should be the real business intelligence BI thinking? Multidimensional thinking and model thinking determine the final direction of a business intelligence BI project, and these points will be discussed later.
9. The essence of business intelligence BI - the implementation of enterprise business management thinking
The Essence of Business Intelligence BI - Pico Data Business Intelligence BI Visual Analysis Platform
What exactly is Business Intelligence BI? technology? product? Or something else? We have raised our understanding of BI to another level: business intelligence BI is the implementation of an enterprise's business and management thinking. How to understand this? To put it simply, the content presented on the visual report is what a company really cares about. There are business analysis indicators that the top management pays attention to, as well as specific departments.
10. What are the differences and connections between BI and Data Warehouse?
It is often encountered that people ask what is the difference between business intelligence BI and data warehouse. In fact, behind this question, it can reflect that some friends' understanding of business intelligence BI is still somewhat inaccurate and biased. This question actually conceptually compares BI and data warehouse. The data warehouse is artificially fragmented. This situation is actually quite normal, because everyone’s first impression of business intelligence BI is all kinds of cool visual charts and reports, and there are many lightweight front-end visual business intelligence BI analysis tools on the market. The cognition of BI stays in the visualization part.
To be precise, business intelligence BI not only includes the capabilities of front-end visual analysis and report presentation, but also includes the construction process of the underlying data warehouse. Gartner has mentioned Business Intelligence in the 1990s, and it believes more: BI is a data-based technical solution that extracts and analyzes valuable data from many different enterprise business systems. , Conversion and loading are the ETL process of extracting, converting Transformation, and loading Loading, and finally merged into a data warehouse, according to a certain modeling method such as Inmon's 3NF modeling, Kimball's dimensional modeling or both Hybrid architecture model, and finally on this basis, appropriate analysis and presentation tools are used to form various visual analysis reports to provide data decision support for the management decision-making layer of the enterprise.
Business Intelligence BI - Pico Data Business Intelligence BI Visual Analysis Platform
Therefore, it can be seen from here that the position of the data warehouse Data Warehouse is a layer between the visual report and the data source of the underlying business system, and it plays a connecting role in the entire business intelligence BI project solution. If business intelligence BI is compared to a person, the upper body, especially the face, is the appearance, the lower body absorbs the essence of the earth down-to-earth, and the core and core strength of the middle part is the data warehouse.
Then everyone will also ask, isn’t there a lot of business intelligence BI tool products on the market that can be directly linked to data sources and can be dragged and analyzed? Isn’t it also possible to make business intelligence BI analysis reports? This independent, front-end-oriented business intelligence BI analysis tool is more positioned as a department-level and personal-level business intelligence BI analysis tool. For deep-level scenarios that require complex data processing, integration, modeling, etc. is unsolvable. The best way is to build a complete data warehouse at the bottom, standardize many analysis models, and then use these front-end business intelligence BI analysis tools to combine, so as to truly release the front-end business intelligence BI analysis capabilities.
Many companies think that they can solve all the problems of enterprise-level business intelligence BI just by buying a front-end business intelligence BI analysis tool. This view is actually not feasible. There may be no problem with such business intelligence BI analysis tools when the initial analysis scenario is relatively simple and the complexity of docking data is not very high. However, there is a characteristic in the construction of business intelligence BI projects in enterprises, which is a spiral construction process. Because there may be more and more connected business systems, the depth and breadth of analysis will increase, and the complexity of data will become more and more challenging. At this time, there is no good data warehouse architecture support. Front-end BI analysis tools are basically impossible to handle.
Data Warehouse - Pico Data Business Intelligence BI Visual Analysis Platform
Just like going to a traditional Chinese medicine store to grab medicine, the reason why the medicine is picked up quickly is that before the medicine is picked up, others have cleaned up and put away various original Chinese medicinal materials (data from the original data source) in categories, so how to match the medicinal materials (dimension Visualization of indicator combinations) is fast.
There are many such companies in China, and it is also because of insufficient understanding of business intelligence BI that leads to some directional mistakes in the construction of business intelligence BI projects, and finally makes it difficult to continue to promote business intelligence BI projects.
Therefore, in enterprises, we need to clarify whether our business intelligence BI construction is for enterprise-level or individual and departmental analysis. If it is a personal data analyst, it is enough to use this type of front-end business intelligence BI analysis tool. If you need to build an enterprise-level business intelligence BI project, you should not only focus on the level of front-end visual analysis capabilities, but also pay attention to the construction of the underlying data architecture, that is, the level of the data warehouse.
Eleven, data warehouse modeling methodology Kimball vs Inmon and hybrid architecture
Data warehouse modeling is the top priority in the construction of business intelligence BI projects. Inmon's three-paradigm 3NF modeling and Kimball's dimensional modeling are both business intelligence BI data warehouse modeling methodologies. These two business intelligence BI modeling What is the difference and connection in the way.
12. The methodology for the implementation of requirements when actually carrying out a BI project
Business intelligence BI is completely demand-driven. Since it is a demand, interviews and research are required. Before conducting interviews and research on business intelligence BI needs, it is necessary to be familiar with the business characteristics of the industry in advance. Based on the fact that the enterprises themselves need to be familiar with their business processes, as well as the key points that they will probably pay attention to in the interviewed departments, they need to be sorted out in advance. Build up the entire business framework in your mind and practice it repeatedly.
13. What kind of enterprises should use business intelligence BI?
What kind of enterprise is suitable for business intelligence BI? Look at the degree of basic business informatization and the level of detail and granularity of daily business management. The degree of informatization of the business foundation is the basic construction of the enterprise's own IT business system. Without the support of the business system, there is a lack of data foundation for business intelligence BI; the second is the granularity of business management. Is the degree of business management of the enterprise itself more detailed? There is an urgent need to improve the efficiency of business management and decision support through business intelligence BI.
14. How to effectively report and summarize BI data analysis to senior leaders
After completing the business intelligence BI project, you also need to consider how to report to the boss in the end, and master the business intelligence BI data analysis thinking framework and five key points of reporting: user business level and scope, work results, plan execution review, problem feedback, Outlook planning and vision.
Business Intelligence BI - Pico Data Business Intelligence BI Visual Analysis Platform
This is just a simple reporting framework, and there are many points that can be added to it. For example, talk about the combination of industry driving factors and business intelligence BI around the industry; how to decompose and reorganize from the perspective of enterprise management, enterprise vision, CSF, KPI, and performance; such as attribution analysis from a financial perspective; pyramid management The model; the composition principle of the dynamic indicator library, etc. can be selectively integrated and explained.
15. The degree of integration of business intelligence BI and enterprise management
Business intelligence BI analysis is highly integrated with enterprise management analysis. Enterprises with high ROE may have high profits such as Moutai and the jewelry industry, or they may have fast turnover such as the retail industry, or they may have relatively strong financing capabilities and use leverage. ROE attribution analysis depends on industry characteristics.
16. Accumulation of business intelligence BI project industry and business knowledge
To do business intelligence BI, you must also be familiar with industry and business knowledge. Without combining industry business knowledge, it is difficult to implement business intelligence BI projects. The essence of business intelligence BI is actually the implementation of the enterprise's business and management thinking. Why do the senior management of the enterprise and the managers of the business department use the business intelligence BI to read the report, what are they looking at, and what are they focusing on? These contents are the focus of their daily business management in the enterprise.
Data Visualization - Pico Data Business Intelligence BI Visual Analysis Platform
The seemingly scattered reports on the business intelligence BI project actually have a strong logical connection in the eyes of actual users. And the higher the level of management personnel, the more focused the content of the business intelligence BI report is, and what they see is the business result. Personnel in the front-line business department may focus more scatteredly, looking at detailed business process data.
Therefore, for an excellent business intelligence BI developer and development consultant, it is not only necessary to polish at the technical level, but also to accumulate industry knowledge and enterprise business knowledge.
17. Topics about real-time processing of business intelligence BI
Business intelligence BI has a certain lag in data processing, and usually adopts the T+1 mode. The main reason is that the ETL data processing process requires a lot of time consumption, and usually adopts the method of exchanging space for time.
Designing the previously layered ETL scheduling according to the business intelligence BI data warehouse into a scheduling that can automatically find dependencies according to individual indicators will greatly increase the flexibility of individual indicator scheduling and quasi-real-time processing.
The business scenarios for offline data and real-time processing are different, the technical methods behind them are different, and the resource investment is also different. Understanding the positioning differences between them helps to choose the appropriate solution to achieve the enterprise's planned completion of the business intelligence BI project with the minimum resource investment The goal of building.