What is Data Mining and How Can it Positively Impact the Bottom Line
Modern times require modern solutions to ancient problems. One of those problems is what to do with large volumes of data and what information that may be hidden in that data is of any use to you. In times past data sets like the purchasing habits of customers were sorted through by hand and associations made in that method. However human beings are prone to making mistakes or not noticing things that they aren't expecting. Computers do not have this problem. If you tell a computer system to search for anomalies or associations in existing data sets it will return to you every single instance of that anomaly or association there is, whether or not such an association was expected.
This is where data mining comes in. Data mining is a field of computer science focused on analyzing data and discovering associations and/or variations and anomalies that would be impossible for human beings to recognize, either due to the volume of the data or simply because nobody was looking for it. Data mining implements technologies ranging from artificial intelligence to database management. The main focus is on discovering previously unknown patterns in extant data sources.
More importantly than what is data mining, is the question of how data mining can positively impact your bottom line. Data mining allows for analysis of large data sets for associations and anomalies that may not be readily apparent. Some types of information do not really lend themselves to data analysis as they can be easily inferred, for example someone who purchases large volumes of high fiber foods can be fairly straightforwardly expected to purchase large volumes of toilet paper.
While humorous, these kinds of associations can make a huge difference in sales. While the fiber example above might be obvious there are thousands of examples of data analysis locating associations that nobody would have thought of had a program not alerted it to them. Take for example a common product such as cookies. Most people would assume that a customer purchasing cookies is likely to purchase milk, or perhaps ice cream or in more esoteric applications cake ingredients. What may not be as apparent is that frequently someone who comes to the store to purchase cookies will also be going there to purchase vegetables. While this is an example and not likely applicable in the real world, stranger associations have been made.
What can you do with this information? Well say you have a profit leader in a particular product, and you've found through data analysis that it is frequently purchased alongside something that does not sell particularly well. It would be an easy thing to put the popular product on sale to cause it to move faster, which would also cause the less popular or profitable product (or perhaps a product that is nearly out of date) to move off the shelves. Got a large supply of tortilla chips that are about to go bad? With data mining you could find out that a particular brand of soda is often purchased with it and by putting the soda on sale you'll move the chips. You might even move them faster than simply lowering the price on the chips, which might be selling so low to cost you can't afford to do that anyway.
Another useful function of data mining is locating anomalies in data. Looking at a chart that says each month your business brings in a certain amount of money will never reveal that on the second and fourth Tuesday of each month your southeastern locations do triple their normal volume. With that information at your fingertips you can begin to look into why this happens, and take advantage of it by either holding sales on other days to keep people coming to the store more often, or ending sales on Monday so that your prices are not discounted on the busy days. Further analysis may even reveal how you could increase traffic on days other than those, or at other stores in other locations.
A final, and perhaps one of the most powerful, function of data analysis is as a decision support system. In many situations it is important to make a business decision that could have catastrophic effects down the line. Using data mining and related technologies it is possible to eliminate as many variables as possible from these decisions. In business knowledge can be the one thing that a competitor cannot copy or take from you. Knowledge of how your business functions, knowledge of how your previous decisions have impacted your bottom line, and the knowledge of who works for you. With data mining technologies, you can make sure that you have every scrap of knowledge available to you that you need to make the absolute best decision possible.