Answer to Question #225932 in Marketing for Sid

Question

Suppose you are Marketing Strategy Manager at XYZ Supermarket. Your Supermarket uses Loyalty cards to maintain their customer’s data like Purchase History, Delivery Information, Contact and Personal Information etc. The Supermarket uses Loyalty card data to send regular promotional and discount offers to relevant customers. The Supermarket planned to launch their own Ready-to-Cook food items. For this purpose, they asked you to prepare predictive analysis of the products to check the sale capability and target audience for these products.

So as a Marketing Strategy Manager, which of the following Data Mining technique will you use for the above mentioned task and why? (Briefly explain in only 2-3 lines)

• Classification
• Association

Answer 1

Classification as a Data Mining Techniques

I will use classification as a data mining technique for predicting future sales.

Classification allows us to forecast future data patterns and trends with models that detail vital

classes of data. Intelligent business decisions depend on the functions of models that make

recurring predictions. Furthermore, the classification uses algorithms that foretell clear-cut class

characterization. It involves connecting two variables: the required variable and the one of

interest. Since we want to predict our customer's likelihood of purchasing food, this becomes a

qualitative variable.

For purposes of prediction, algorithms develop a link between the two variables.

Classification then organizes data into different sets for easier prediction. It allows us to

categorize data sets both simple and small; large and complex data set. As a manager, I can

easily connect customer purchases as a variable of interest and the future likelihood of

purchasing other foods. For example, the likelihood of a customer purchasing fries after buying

soft drinks is easier from analyzing the data variables. It's also easy to predict the purchasing

patterns of the millennials and the likelihood of purchases fast foods from the given sets of data

variables.