Business Intelligence "Real" World Story #1

Fairly early in my career I worked for a company that sold corporate-logo apparel to customer locations in the United States. Typically these customers were franchisees or retail outlets for large corporations, which were our “clients.” One of my duties, with help from an assistant, was to make sure our sales force got sales reports for the clients and customers in their territories. When I started, it worked like this: each Monday, our corporate head office sent us a sales report for the week just ended. It was sorted by customer location for each client. It was about the size of the USA’s 2010 health care bill, each week. It was printed on wide-ledger computer greenbar paper. Never heard of that? Lucky you. It took the better part of two days to go through this printout and put it in a format that had any meaning for the sales force. That meant that the week was half over (Wednesday) before they saw numbers for the previous week. Apparently this is how it had “always been done.”

I poked around at the process for awhile, and then I found out about a software tool called Monarch. This tool allowed you to take a formatted text or print spool file, and convert the contents into a form that could be easily loaded into a spreadsheet or a database on your PC. I struck a deal with a friend in the corporate IT department to leave the print spool file from the job that printed the sales report on the network for a while after it had finished printing. I would then download it, convert it using Monarch, and load the numbers into a database program called Paradox (all this was before Microsoft Office came to dominate the world). I wrote a report in Paradox that would add up and sort the data by client from highest sales to lowest. It wasn’t totally automated, but we were able to spit out more useful sales figures on the same day as IT put the sales out. And without realizing it, I had begun my BI career several years before it officially started.

Next time: an introduction to the business intelligence life-cycle.

A "General" History (Part Two)

In the first part of this article, we described business intelligence, such as it was, in the days before computers and data processing were part of everyday life.

Time passed. Markets and populations grew. Manual data processing eventually got to be too slow. So automated data processing was introduced to speed up the pace at which business could be done. The first automated solutions were mechanical, such as cash registers and adding machines. Then along came electricity and computers, and the age of electronic data processing (EDP) dawned. Businesses could now process a much larger volume of transactions than before, making it easier to grow very rapidly.

But there was a downside. Once a company reaches a certain size it becomes too big for one person to manage off the top of his or her head as our general store owner did. There are employees to manage. There may be many stores in different parts of town, or in different countries. An item that was in stock just yesterday is suddenly gone from the shelf. With good bookkeeping you can figure out whether you’re making money or not. But you don’t necessarily know why, or even how, you’re making that money. Worst of all, you couldn’t get answers to these questions from the EDP systems. All you found out from them, if you found out anything, was that you did a whole bunch of transactions. There was a desperate need to get data out of these transaction-based systems and use them to make informed business decisions. As a matter of fact, the term “business intelligence” is often used interchangeably with the term “decision support.”

Next time: a story from the "real" world.

A "General" History (Part One)

BI as it exists today grew out of the automated data processing revolution of the late 20th century. “Data” are simply facts about anything that interests us. We “process” the data in different ways and for different reasons, as we’ll discuss later. Before automated tools like computers were common, we were limited in how and how much data we could process.

In the Old West of United States history, the most common type of retail outlet, especially in small towns and outlying areas, was the “general store.” The general store was a one-stop shop for the staple items of the day. The salesperson that worked behind the counter may very well have owned the store. When you bought something, say a pound of coffee, the salesperson/owner would write the details of the business transaction down in a ledger. These might include who you are, what you bought and how much, as well as what you paid or (if you were fortunate enough to have a line of credit) what you owed. It was all done by hand with a pencil, but it was still data processing, and in the slower days of old it was more than sufficient.

Jones, B., one lb. Coffee @ $ .01 – Paid in Full

Our general store owner had all the business intelligence he needed. He likely knew all of his customers by name, except for the occasional drifter with a fancy six-gun rig (but that’s another topic, not to mention another genre). He knew exactly what his financial position was by checking his cash box. If he wasn’t sure about his inventory status he just checked the storeroom out back.

Next time: The modern age of BI.