Data mining techniques can be applied to many applications, answering various types of businesses questions. The following list illustrates a few typical problems that can be solved using data mining:
Churn analysis: Which customers are most likely to switch to a competitor? The telecom, banking, and insurance industries are facing severe competition these days. On average, each new mobile phone subscriber costs phone companies over 200 dollars in marketing investment. Every business would like to retain as many customers as possible. Churn analysis can help marketing managers understand the reason for customer churn, improve customer relations, and eventually increase customer loyalty.
Cross-selling: What products are customers likely to purchase? Crossselling is an important business challenge for retailers. Many retailers, especially online retailers, use this feature to increase their sales. For example, if you go to online bookstores such as Amazon.com or Barnes andNoble.com to purchase a book, you may notice that the Web site gives you a set of recommendations about related books. These recommendations can be derived from data mining analysis.
Fraud detection: Is this insurance claim fraudulent? Insurance companies process thousands of claims a day. It is impossible for them to investigate each case. Data mining can help to identify those claims that are more likely to be false.
Risk management: Should the loan be approved for this customer? This is the most common question in the banking scenario. Data mining techniques can be used to score the customer’s risk level, helping the manager make an appropriate decision for each application. Customer segmentation: Who are my customers? Customer segmentation helps marketing managers understand the different profiles of customers and take appropriate marketing actions based on the segments.
Targeted ads: What banner ads should be displayed to a specific visitor? Web retailers and portal sites like to personalize their content for their Web customers. Using customers’ navigation or online purchase patterns, these sites can use data mining solutions to display targeted advertisements to their customers’ navigators.
Sales forecast: How many cases of wines will I sell next week in this store? What will the inventory level be in one month? Data mining forecasting techniques can be used to answer these types of time-related questions.
Ref. Data Mining with SQL Server 2005