Sunday 23 December 2012

Business Intelligence Overview

It is such a long time no posting anything. I feel a bit disappointed of myself when I couldn't spend more time for this blog. As it is a part of my job, I should maintain it more sufficiently thought.

Recently, I joined a long long project which is totally non-related to CRM, it's all about BUSINESS INTELLIGENCE ! Feel being distracted a bit when I kept dragging between CRM and BI...

Anw, I initially was an Information Systems student. My major is working with data, making them from boring numbers becoming interesting things. Today I would like to spend time to provide you guys (who interested in BI) what I have learnt about BI. The agenda for the post is:

1. What is Business Intelligence ?
2. How does BI actually work?

I hope someone out there will interested in this blog.

1. What is Business Intelligence ?

Okay, in general, BI could be defined as the delivery of accurate, useful information to the decision makers. Here I would like to emphasize the "Accurate & useful" information. It means, they are not only just raw numbers, their are "meaningful" and have ability to tell you something.

For example, for sure, from financial perspective, every company has maintained some kind of operation reports which include transaction records, bank records, etc. But finance professionals want to also move beyond standard reporting to analyse details, provide summary to optimise financial performance, such as: Total Sales, Revenue, Profit Margin, etc. To achieve these goals, you must have full insight into the bottom line, cash flow, investments, sales and many other factors. BI is the tool helping you to combine all of those information into one package.

Business Intelligence, to a great extend, is a stack of Data Discovery and Knowledge Management, and is a higher level of Operational Reporting Services. Looking at the diagram below to understand more levels of Data Discovery:
 

The lowest degree of intelligence is Statistics Information. They normally are raw numbers. People normally use standard reports or operational reports to present them. From there you will see what is the fact ? what is actually happening ? Those kind of information is provided daily, weekly, monthly and always be the same. They are good for people who are in operational level. Whereas, while look into the upper level of Intelligence, where BI located. After analyzing raw statistic, BI provide you the insight of information. It not only shows you what is happening anymore, it but also provide you why it is happening and predict the trends what would be happened in the future.

2. How does BI actually work?

I might make you guys feel boring a bit with the all above theory about BI. The next question will be, how does BI actually work ? How could it help us to understand the meaning behind numbers?

Well, pls take a look at the below diagram:

The concept of BI could be understood simply like this: you collect raw data from different sources. For example, in an organisation, you have different departments with different databases, storing different information from: Sales (bills, orders, invoices..), CRM (Accounts, Contacts, Cases..), Supply Chain (products, inventory movement…), etc. Then putting all of raw data into a “Data warehouse”, generating “Cubes” which contains all of those information. And finally, doing analytic queries and  presenting them by using BI Tools such as in graphs or reports…

 When talking about BI, definitely we need to understand the definitions of "Cube", "Dimension" and "Measure".

 
 

In order to provide meaningful information,  BI stores it data in Analysis Database (as shown in the above picture). Different from normal database, organize data in rows and columns, Analysis Database used cubes to present BI data. Cubes used for business intelligence is divided into two units :
Measure: a measure (or fact) is a numeric quantity expressing some aspect of an organisation’s performance.
Dimension: a dimension is a categorisation used to break down an aggregate measure to its constituent parts.
In the picture above, the measure here is Total Sales & the dimentions are Year & Product Type.