The phrase business intelligence is pretty old, 1958. In the early 90’s the term came to refer to the “concepts and methods to improve business decision making by using fact based systems.” So business intelligence is the means for taking all of the data in your organization and pull out the relationships. By finding these relationships you can make smarter decisions about how to manage your organization.
Currently there are two goals for modern Business Intelligence
- put as many facts as possible in front of decision makers
- put as many facts in front of everyone
Let’s say you worked in a tele-sales department. You have a fixed number (at least in the short term) of sales makers. You have a fixed number of leads (in the short term). The more efficient you are at converting a call to a sale, the more money you make. What do you do to maximize your profits.
- Identify your strongest and weakest sales people — If you know how effective they are you can either eliminate the weaker candidates, or use the stronger candidates to mentor the weaker ones.
- Maximize the amount of time your sales people are talking to people, and minimize the amount of time they are waiting for a call to be made. — Most of us have caller ID now, and if we don’t recognize the number, we screen it first. Therefore if you call a number and it never picks up, when do you give up on calling that number?
- Identify your easy sales — Are some people more prone to spur of the moment purchases? Of course some are. What if you could find those people who buy time and time and time again?
There are many more relationships you could start looking for, but these I’ve actually done some research on and found a positively correlating dataset. It should show you what types of relationships you’re trying to discover, based on a certain business model.
In order for Business Intelligence (BI) to work you have to have rapid access to a lot of data. Typically your databases are set up to record transactions. Combining data from every transaction can extremely time consuming. BI addresses this issue through data warehousing. A Data Warehouse is a very large database set to optimize data reads.
A data mart is just a mini data warehouse that contains the portion of the whole data base that relates to specific division or group within your company.
The down side to your data warehouse? The amount of time it takes to transform your transactional data into your data warehouse. Also, the amount of data you may need to keep on hand to show a trend could require you to hold on to many terabytes of data… This could become costly.
The good news is hard drive space is always the cheapest component in a server build (per unit). It’s time that you can never seem to buy. Over the coming months, I’ll introduce you to some tools and techniques for using BI in your own company. You’ll see SSAS and SSRS, as well as new and interesting ways of using SSIS to pull over data for analysis.
As always, if you have any questions, please send them in. I’m here to help you learn as much as you can about Microsoft SQL!