Resolve the root cause of business intelligence failure

Repost: resolve the root cause of failed business intelligence

Reprinted from: http://www.chinabi.net/CIO/implement/200709/724.html

There is a story that has been repeatedly mentioned. It can demonstrate the application value of business intelligence (BI. For some time, Walmart's store manager in the United States found a phenomenon: Sales of beer and diapers increased year on year every week, but I was confused about the reason.

Later, Walmart used Bi technology to find that almost 25 ~ A 35-year-old male with a baby at home is bought on weekends. Walmart analyzed the relevant data and found that these people are used to watching ball games and drinking beer at night. For children to take care of, they use disposable diapers for convenience purposes. After this conclusion, Walmart put these two products together, and the sales volume has increased significantly.

This example tells us that in today's business environment, the seemingly disorganized and complex information often hides huge commercial values. Whether the analysis can extract valuable information and support managers to make timely and correct decisions is very important for enterprises to cope with complicated competition.

However, compared with Walmart, an enterprise that applies information technology to the extreme, Chinese local enterprises rarely use Bi applications. Although Industrial and Commercial Bank of China, Minsheng Bank, Guangdong Development Bank and other financial enterprises, and Lebai, Shuanghui and other traditional enterprises are working on the establishment of BI systems, it is still impossible to verify the effect of the application.

According to relevant statistics, 60% of enterprises that invest in Bi abroad ~ 70% ended in failure. In China, this data may be higher. There are many reasons, such as lack of support for historical data and misunderstanding. Liu Xiaobing, CIO of Shuanghui group, believes: "Although Bi technology is not complex, there must be at least three ~ 5 years of historical data accumulation to assist research, otherwise investment will be futile ."

　　Three major obstacles to Bi applications

Enterprises that can successfully use their information resources must be aware of the commercial benefits that information can bring-not only to improve hardware efficiency, in addition, it has benefited a lot from soft areas such as improving sales performance, improving work efficiency, and enhancing customer satisfaction.

However, for most enterprises, information is still an asset that has not been fully mined. That is to say, most enterprises have not yet benefited from the information systems built with huge investment. What are the obstacles that enterprises often encounter when deploying bi?

First, it cannot provide reliable information.

According to a study last year in business weekly, 43% of business people are not sure whether their internal information is accurate, 77% of respondents said they had made bad decisions due to lack of trusted information. In this case, there is a big problem with data quality.

For example, local banks have tens of millions of records and hundreds of millions of records in their databases. The quantity is huge enough, but what is the quality? An account with the same name or an abandoned account has multiple accounts. Professional tools are required to clean and improve the information and prevent low-quality data input systems. In the business process, foreign banks place special emphasis on the accuracy of customer information collection, which gives them an advantage in grasping customer needs and product innovation.

The second is whether a simple and visual user interface can be designed for non-technical personnel.

For non-technical personnel, how simple the system operation is likely to become a key to popularizing bi. That is to say, end users want to obtain information easily and automatically, rather than being forced to learn and recognize a new environment, or rely on IT department help. In particular, the preparation of business decisions has never been completed by a certain individual in a vacuum environment-this requires the user to coordinate and discuss information data before making decisions, this requires an easy-to-operate environment.

Deployment and Management of Bi applications should also be simple. More and more enterprise users want to find packaged, easy-to-implement mid-market solutions or software service projects, in this way, they can directly access Bi applications online without installing any other software programs.

The third is the misunderstanding of greed and perfection.

No matter which kind of management software you use, you will almost hear the same voice: do not be greedy for perfection, start with the most urgent business, Bi is no exception, it can be made into an independent huge system, all business data of an enterprise is stored in a data warehouse for multi-dimensional analysis. It can also be embedded into individual business data for separate business analysis.

　　Four suggestions for successful implementation

Although the application value of Bi has been recognized by many enterprises, some enterprises that want to build BI systems are still confused. So where should they start? What preparations should we make before going to the system?

To make bi bigger and smaller, start with the most urgent business. First, manage the most critical business to quickly respond to market demands and make the best decisions. After accumulating a certain amount of experience, the Bi system gradually increases the number of other businesses to continue decision-making and analysis, which can avoid risks to a certain extent, because Bi also needs to restructure the process, the rectification of a department has a much less impact on the company than the rectification of the entire company, just like a minor operation is always less pain than a major operation.

Drive by business instead of it. Many enterprises understand the principle that the requirements of business departments should lead the design and development of software. However, the level of business personnel is insufficient to reach the goal of proposing a future management model architecture. For this reason, enterprises should give the IT department absolute authority, it has the right to require business departments to cooperate with IT departments to negotiate and propose demand models, so as to grasp the business development direction of enterprises as far as possible.

A dedicated Data Analysis Department should be set up. In China, there are not many data analysts mentioned in this position. Some experts suggest that if you are preparing for the Bi system, you must implement the corresponding data analysis department. The purpose of enterprises to optimize data is to find the most valuable data from the data. If there is no corresponding person to track and process the valuable data after it is mined, its value is only at the rapid report level.

Bi is also a top-notch project. From the perspective of top leaders, the goal is to increase the value-added business of enterprises. Without strong support from top leaders, the adjustment of business processes of enterprises will face great resistance or even difficulty in implementation.

Therefore, before deciding on Bi, enterprises should carefully consider which businesses need decision-making support first, whether the internal process of the business is clear, and how to restructure the internal process. Relevant departments must be provided to track and optimize data analysis, so that enterprise decisions become more rational and objective.

Link to this article: Repost: resolve the root cause of failed business intelligence
Reprinted please indicate the source: Business Intelligence blog-dinosboy