Home Depot is a major retail chain specializing in the sale of construction, home repair, and maintenance products. The company has more than 2240 retail stores in North America and employs over 340 000 people worldwide.

Suppose you are a buyer for the clothes washer and dryer product line at Home Depot. You work with seven different brands and numerous models within each brand. One of your goals is to turn your inventory as many times a year as you can, and to do so you want to identify poorly selling models (and even brands) as quickly as you can. This identification is not as easy as you might think because competition is intense among washer and dryer manufacturers, and a new model can quickly capture a substantial portion of another models market share. Thus, a big seller this year can be a dog (a poor seller) next year.

Another problem is that while some sales trends are national, others pertain to specific regions. A strong seller on the East Coast may not sell as well on the West Coast. In other words, a brand can be a big seller in one region and a dog in another.

When answering the following questions, assume that you have total sales data for each brand and model, for each store, and for each month. Also assume that you know the stores city and province.

Questions

1. Explain how reporting systems could be helpful to you.
2. Show the structure of one or two reports you could use to identify poorly selling models. How would you structure the reports to identify different sales trends in different regions?
3. For one of these reports, write a description of your requirements that would be suitable for communication to an IT professional. Be as complete and thorough as you can in describing your needs.
4. Explain how data mining systems could be helpful to you.
5. How could cluster analysis help you identify poorly selling brands? How could it help you determine differences in sales for different geographical regions? Is the unsupervised nature of cluster analysis an advantage or a disadvantage for you?
6. Do you believe there is an application for a KM system for identifying poorly selling brands? Why, or why not?
7. Do you believe there is an application for an expert system for identifying poorly selling brands? Why, or why not?