BI

Business Intelligence (BI) is to take advantage of various data accumulated in a company to do a prompt and accurate decision-making for business. Also, BI refers to systems and technologies that support it.

Overview
To understand BI, it is necessary to understand the differences among data, information, and intelligence. Data are " objective facts, such as numerical value etc. extracted from a certain event" and it cannot have any meanings alone. On the other hand, information is "data processed to be useful for a certain purpose." Then, intelligence is, from data, to create, extract, and make use of information useful for decision-making.

Therefore, BI is, in the business, to create and extract accurate, useful information from data related to activities of a company, and to utilize it for decision-making. Decision-making is indispensable in business and understanding the current state accurately and applying it to the decision-making is very important. Today, trends of products and information change so dynamically that decision-making should not rely on just intuition and experiences. Rather, it comes to be desired to understand the current state by collecting and analyzing information (facts), and to do an accurate decision-making. Thus, importance of BI has increased.

Example
To derive a plan to improve business by using BI function, a company firstly confirmed the transition of sales of each sections by looking through the past sales data accumulated in the data base. It was a music business that sales had fallen, and then the company decided to confirm details. The breakdown of sales of the music business showed remarkable decrease of sales of music itself, but it also showed no sharp fluctuations in the amount of the profit. To ascertain the cause, the company confirmed the breakdown of sales of music, and it came to know that sales of CDs notably depressed, while online music distribution greatly increased. Cost reduction of hardware that had been necessary for selling music until then was the reason for the fact that the amount of profit has not fallen in spite of decline of sales.

Confirming the importance of online music distribution, the company additionally analyzed customers, and knew that the main users were women in their 20s and 30s are the main users, while the portion of customers older than 50s was extremely small. Checking the CD sales data in the past, the company realized that, in the music business, customers older than 50s were not few, so the company compared the numbers of customers by generation of CD sales in the present and the past with those of online music distribution. As a result, it became clear that customers older than 50s still bought music in the form of CDs although customers in other generations had greatly shifted to online music distribution.

The result of this analysis tells that the music business is actually an excellent business with high rate of profit although the sales apparently decreases. The result also implies that a further rate of profit can be effectively attempted by taking a measure of making customers older than 50s shift from CDs to online music distribution. In addition, analysis of past CD sales records shows a correlation between the number of sales to women in their 20s and 30s and that to women older than 50s. This result implies that a daughter may recommend that her mother should purchase.

Consequently, the company, first of all, decided to aim at women older than 50s. Furthermore, in order to encourage them to shift to online music distribution business, the company also decided to appeal to women in their 20s and 30s, who are in the age of their daughters, instead of directly targeting them.

Thus, BI supports an accurate and swift decision-making by immediately executing analyses of the data accumulated in the past, such as correlating facts seemingly not related to each other, digging down to a variety of angles, and so forth.

Important Point
Regarding BI, an important thing is "What and how do we use?" Considering the analogy between decision-making and cooking would help you understand the concept. To cook well, we choose good ingredients first, then we cook them well as we like. Similarly, to do a correct decision-making, we choose good data, and find useful information for decision-making. In order to cook well, you should not choose unnecessary ingredients or rotten ones. In the same way, choose accurate and useful data for our decision-making. Moreover, depending on recipes, good material could be totally spoiled, or the result would be different from the expected one. Similarly, depending on the way of analysis, even data useful for decision-making could end up information incorrect or unhelpful for decision-making.

Systems Equipped With BI Function
BI function requires data related to activities of a company, as well as analyses of such data for decision-making. Today, in terms of them, each information system and software are integrated.

Data Warehouse/Data Mart
Data Warehouse (DWH) means a data base, and the data of company activities are organized and accumulated in an appropriate shape for the purpose of decision-making. It collects data concerning company activities from applications that manage the business such as ERP in an integrated manner. If we compare it to cooking, it is something like a market where we can obtain ingredients necessary for the dish. There, different kinds of ingredients are sold at different stores,depending on the kind, at a fruit and vegetable shop, a meat store, or a fish store.

In addition, we sometimes call the data base, which extracts and re-organizes specific data based on finer-grained purposes such as individuals, departments, etc., Data Mart. This is, so to speak, like a market where they sell only Kyoto vegetables for making special dishes in Kyoto or special products of Kyoto.

Data Mining/Online Analysis Processing
Data Mining refers to a technology to analyze data by using statistics etc., as well as to search out relations, patterns, or rules etc. among items hidden in data. Online Analytical Processing (OLAP) is to look into data from various angles, for instance, according to "by region," "by product," "by month," etc. Also, it quickly answers to these demands with analytical processing by digging data down in detail such as from "monthly" to "daily."

These technologies generate information from the data accumulated in DWH, which is helpful for decision-making. If it says with cooking, this is indeed a cooking technique that changes ingredients into delicious dishes.

Use of BI in BPM
BPM is an continuous improvement activity in an business process, and it incessantly executes design, execution, monitoring, and correction of the business process. Applying BI is preferable for decision-making in BPM, while there is also a concept named BAM used for the decision-making by monitoring a business process in real time.

For better understanding, let's confirm these differences in the PDCA Cycle in BPM.



Plan (Design of a business process)
We arrange and decide how to advance the business. Using BI is preferable to set a probably efficient process based on the data accumulated. Also, an index that measures it will be set by finding out elements that become important when executing the process. This index is called KPI, and what collectively displays KPI to the administrator of the process, such as owners and managers, is called dashboard.

Do (Execution of a business process)
We execute the designed process faithfully.

Check (Monitoring of a business process)
By using dashboard, we monitor in real time whether the business process is faithfully executed while executing the business process. This is BAM.

Information obtained by monitoring accumulates in the data base. We analyze data concerning the past execution results accumulated in this data base, and identify problems at the meta-level that requires any restructuring of the process. This is BI.

Act (Correction of a business process)
While executing the business process, if BAM finds any problems to disturb its execution, it responds to them promptly. Concerning problems at the meta-level that requires restructuring of the business process, which are found at the Check stage, we will go back to the Plan stage and will solve the problem by restructuring the business process.

BI takes care of the way to compose the dashboard, while BAM actually uses that dashboard.

Related Articles

 * BAM
 * DWH
 * KPI
 * PDCA Cycle