6 fatal errors in implementing business intelligence

How do I get data to drive a business? This is a key issue that companies must think about in the big data age. Today, enterprises need to deal with too many business unit needs, and the complexity of data determines that enterprises must use business intelligence to cope with the change of business needs and uncertainty, information processing capability determines the success or failure of the enterprise key. Many companies spend a lot of money, human and material resources to build online transaction processing (OLTP) and Enterprise Resource Planning (ERP), accumulated a large number of data, but the traditional analysis tools are difficult to timely and accurate business analysis of these data, the emergence of business intelligence technology for these problems to provide a solution.

However, the business intelligence platform based on data analysis is facing more and more opportunities and challenges in the era of large data, many enterprises have many misunderstandings in the understanding of business intelligence, and the following six kinds of common problems are listed below.

Let the IT department manage too much

Getting it to do business intelligence platform purchases is often less than desirable, because the IT department focuses on factors such as stability, scalability, security, and vendor reputation in order to minimize the risk of purchasing.

The analyst at Boris Evelson--forrester, who warns us that it is best not to do so, gives too much say to end users (IT staff) in purchasing decisions is a costly mistake.

Boris Evelson also said: "Desktop systems or cloud systems can meet the needs of business users, and this does not rely on it reporting developers, but also may lead to the solution is insecure or unreliable, so it is best to compromise." ”

Ignoring the business intelligence needs of users

When an enterprise invests in business intelligence, one of the most likely mistakes is not to combine system performance with the actual needs of the user. This may not seem important to many people, but many companies fail to implement the business intelligence system because of this basic error.

Rita Sallam, a Gartner analyst, told us about business intelligence: "These systems cost companies millions of of dollars, and sometimes they offer little or no real value, and the key reason is the mismatch between purchasing and user requirements." ”

To avoid such problems, it is essential to establish a specific user requirements reporting function to ensure that users participate in the procurement and implementation of business intelligence systems. Sallam also told us: "It may appear that the user needs interactive reports, and the system can only generate static reports such cases, but also may be too simple or too complex system functions." ”

Underestimate the cost of user training and user support

"Many companies are making business intelligence budgets only considering the cost of software purchases, and of course may consider short-term (for example, two weeks) user training costs. Today, the complexity of business intelligence systems cannot be underestimated, and the need for a longer period of user training to gain real value from the system. ”

Ignoring future business intelligence needs

In a recent survey, about 30% of companies have planned to use cloud-based business intelligence platforms. The figure is now over 45%, according to a Gartner survey. "This means that even if your chosen business intelligence system Provider does not currently have a cloud-based product, it should at least have such a plan to meet your future business intelligence needs," says Sallam. ”

Of course, for future needs, cloud is not the only factor to consider, but also to consider how to make complex analysis easy to understand users, consider how to turn the interactive discovery into automatic discovery, the vendor's product roadmap has at least relevant plans.

Many enterprises choose business Intelligence system suppliers, lack of some long-term considerations, such as whether the supplier can meet the future needs of enterprises, lack of long-term consideration will lead to business intelligence system can only play a short-term role.

Lack of overall consideration

Business intelligence is essentially used to analyze data, and if you plan to access data in JD Edwards, PeopleSoft, SAP, or other large ERP systems, you can't underestimate the role of business intelligence. Evelson reminds us: "Access to data is not easy, access to data is not to say simply access to the database, but also need to understand the metadata and how the data is laid out." ”

Use of unprofessional business intelligence tools for cost savings

Evelson tells us that about 80% of all Business Analytics are done with simple tools, including Microsoft Excel and Access. The use of unprofessional business intelligence tools can also be beneficial, such as cheap, easy to use, and efficient (for simple business analysis).

But non-professional business intelligence tools apply only to small businesses: they can analyze terabytes of data, but it's hard to handle larger amounts of information; they produce an "isolated spreadsheet library (ToolStrip Silos)"; For the same problem, for different parts of the enterprise, may give different answers because they have no uniform description of the same event.

Worse, non-professional smart tools bring security and business risks, Evelso warns: "For ' Who can access data, who can process data ', you can hardly limit it, and once the data or a formula is wrong, the information based on these processing results can cause a lot of problems." ”

In addition, different enterprises in the stage of business intelligence is different, the problem is not the same, some enterprise business intelligence platform has been developed to the data mining stage, some are in the data analysis stage, and even many enterprises are still in the reporting stage. In the report stage, the business intelligence of enterprise often faces the problem of large amount of data and too little value information, and the data processing is difficult. A customized report lacks flexibility because the business often has to analyze the problem from multiple perspectives, so users need interactive reports to understand the combination of different data and generate new information to solve new problems.

Written in the last

Correctly understanding these problems is the key to the development of "enterprise" business intelligence, where the enterprise actually contains all walks of life organizations. For example, government departments, educational institutions, medical institutions and public utilities, business intelligence has a wide range of applications. Business intelligence problem is also a kind of data management problem, including data storage, extraction, cleaning, conversion, loading, integration ... A series of data processing in order to improve the quality and security. Enterprises should give full play to the advantages of business intelligence, we must rely on more powerful tools, which depends on artificial intelligence, machine learning, data warehousing technology, expert intelligence system and other scientific and technological progress and development. The establishment of business intelligence system is a long-term, arduous task, enterprises need strong leadership and execution ability to ensure that business intelligence play a real advantage.