Database Mining Techniques - Term Paper Example

Database Mining Techniques Success of marketing programs of E-commerce websites relies on the availability data. The number of dimensions of customer data and the number of records play a crucial role in data-mining. Individual customer data can be used to alter the marketing programs accordingly and for a better customer relationship management (Dyche, 2005). The heart of customer relationship management (CRM) and personalized marketing programs is data-mining. Spikes can understand the customer’s behavior and preferences by using CRM technologies such as database mining and personalization tools (Dyche, 2005). Data-mining tools and techniques will also allow Spikes to predict the future behavior of the consumers and to develop advertising programs and promotions accordingly. Customer Profiling Lucinda has been quite keen to develop customer profiles so that they are able to target the future sales campaigns in a better and cost effective way. Customer profiling is the process used by organizations to describe the characteristics of groups of customers by using relevant information from the available databases (Manifold Data Mining Inc., 2009). The drivers for their purchasing decisions and their discriminators from other customers are identified (Manifold Data Mining Inc., 2009) so that they can be used to market new products more effectively using data-mining. These customer profiles can be used to develop group specific marketing and sales plans. Customer profiles will also help Spikes to identify the most valuable customers so that their needs can be differentiated (Manifold Data Mining Inc., 2009) from the other customers. Customer profiling can also help improve one to one relationships with the customers. Using data-mining techniques, the customer data, orders associated with that customer and the data about the shoes associated with that order can be used to develop the customer profiles for Spikes. Therefore the profiles should contain the following data; CUSTOMER_NUMBER, FIRST_NAME, LAST_NAME, CITY, AGE, ORDER_QTY, TOTAL_ORDERS, TOTAL_PAYMENTS, TOTAL_SHOES_QTY. Most of these data fields will be derived from the databases using data-mining techniques and tools. This profiling will help Spikes define a better sales strategy, eliminate products not liked by the customers, introduce new products according to the preferences of the customers and gain higher response rates for promotional campaigns. Data-Mining in Marketing Once the customers of Spikes start using the E-commerce website, individual data of each consumer will start building up in the Spikes database. This will allow the company to analyze the trends in the market and to observe the consumer behavior. Data mining techniques are necessary for Spikes because the marketing to CRM personnel cannot analyze large customer databases manually without some tools or techniques. The marketers have to for the results of data mining in their marketing campaign to get the most out of the campaigns and make them successful. Target Marketing As the Spikes consumers start using the online channel to make their purchases, more and more data will be available about them. This customer data can be used in the practice of target marketing for promoting particular shoes to a certain group of prospective consumers. This can be done by analyzing the data and associating the products purchased by individual customers to their personal attributes such as age and city. Spikes will be able to obtain this data from the customer database. Customer age can be deduced from the BIRTH_DAY field that the customer has to fill at the time of registration. The city in which the consumer is placing the order is also an attribute of the consumer and this can also be obtained from the customer database. Therefore the prospective customers will be divided into categories based on their demographics (Dyche, 2005) age and location. By segmenting the prospective customers, the marketing department of Spikes can initiate a more specialized and personalized sort of marketing campaign for each category. Database mining techniques can also be used by Spikes to determine the most desirable segments for the business. For example, if according to the results of data-mining the customer database, dress shoe named Guana is purchased by majority of the customers of the age group 45 to 60 and who are located in New York. This result from data-mining will be of value to the marketers of Spikes retail shop. This is because the prospective customers in the location of New York and of the age bracket 45 to 60 are most likely to have preference and purchase the dress shoe Guana. Therefore the marketing campaign can be focused on this age bracket in New York to reduce costs and get better results. Customer Retention By segmenting the customers on their line of preferences for the shoes offered by Spikes, the company can reveal many interesting facts about their behaviors and preferences (Dyche, 2005). For example, there might be some customers who respond to promotional deals, others might only visit the website and purchase when they need shoes, others might be loyal customers who increase purchases when there is a sale or the customers who seek new products every time they visit the website. All these customers need to be targeted differently. A different marketing campaign is needed for the customers who are have already purchased from the Spikes store or the website. The data for existing customers is also available and their preferences are also known by the Spikes. Therefore it will be easier for the company to retain the existing customers rather than target prospective customers. Planning new promotional deals for the existing customers is only possible if data-mining is done. According to the shoe category, shoe price and shoe brand, the company can make new promotional and marketing campaigns for the existing customers. Through data-mining of the customer and shoes database, Spikes can gain the knowledge of what the existing customers like and what their preferences are. According to those preferences and past buying behavior, the existing customers can be offered new deals to encourage them to stay (Dyche, 2005). Click Stream Analysis Click stream analysis involves capturing the data that shows the footprints of a website visitor around the website (Performics, 2009). This includes all the products that were viewed by the visitor, all the web pages that were visited and the links that were clicked. It can be said that every move of the customers (Dyche, 2005) is being monitored by the website owner including the period of stay and all the activities performed on the website. All this data can be recorded by Spikes in the form of databases and data warehouses. Then using data mining tools and techniques on this data, the prices of the shoes can be modified, inventories can be re-ordered from the suppliers and lost sales and its reasons can be tracked. Data Mining in CRM Customer relationship management (CRM) is dependent upon data mining activities which are used to analyze data to uncover opportunities and threats for CRM initiatives. These insights that are uncovered using data mining techniques would be very tedious and difficult to uncover accurately if data mining is not used. This is because of the availability of thousands of customer records and inability of human mind to analyze such complex and lengthy data. The results obtained from analyses of data through data mining cannot be used by competitors because they data that Spikes has is unique and different from the competitors. Data mining is used in CRM in order to understand the ways to communicate with the customers in unique ways that the competition cannot copy. If the customer contacts any company representative, the representative will be able to “drill down” (Dyche, 2005) the data using data mining techniques and use it to find the required data. Handling Customer Complaints For a better relationship with customers, their complaints about the past orders or and other queries should be handled with a lot of care and accurate information should be provided to the customer. Therefore data mining can also help Spikes representatives to search for information that is stored deep inside databases or data warehouses. Certain tool for data mining will be required by the representative to search for the data. That tool will help the representative query for the data and generate the information required. An example for drilling down data is that a customer calls one of the Spikes representatives and requests the total billed amount for her last order. She only remembers the date for that order and has lost the order ID. The customer has not yet received the order, therefore, she has requested to track down her order and tell her the exact amount she has paid for the order. In this case, if the representative does not use any data mining techniques, it will be near to impossible to find her order details with only name of the customer and date of the order available. Hence the data mining techniques here will be quite useful here to build and improve relationships with customers. Personalization At the time of interaction with the customer, the ability to customize communications with the customer is called personalization (Dyche, 2005). Based on the knowledge preferences and past behavior of the customer, this message is personalized to fit the customer’s wants. Data-mining techniques are used to extract the key customer data required from the customer data warehouse to tailor the communication. There can be two types of personalization. One would be personalization based on a particular customer and the other one is based on customer segmentation (Dyche, 2005). Personalization based on individual customers would be showing a list of shoes that are best suited for that customer using customer profile and data-mining tools. Each time the customer visits the Spike site, customer’s personal data will be data-mined and will be used to create custom contents or marketing messages. These personalization technologies are only enabled by data-mining and they can help Spikes increase their sales and customer experience. The goal of personalization is to recommend accurately the shoes that the customer would purchase. Personalization, when done in the correct manner, results not only in increase in sales being made to the customers but also customer loyalty being maintained (Dyche, 2005). Personalization can enhance customer relationship in various forms. Customization of the actual web pages and the look and feel of the Spikes website can be achieved according to the individual preferences of the customers. These preferences are deduced by using data mining techniques on the data that has been recorded in the past. Customization of the features of Spikes Website can be done on the basis of geographical location of the customer (Dyche, 2005), the age group he is in or the category of shoes he is interested in. Data mining techniques are used here to extract the customer data in order to customize the look and feel of the site or customization of the contents on the web pages to give the customer a more feeling of importance. This way the customer will feel that the shoe company knows him and they know what he likes. Therefore customer relationship will be enhanced by using data mining techniques. Analytical CRM Capabilities Through using data mining techniques, Spikes can use analytical CRM to analyze the customers’ data and change its strategies accordingly. Once the company has developed a website and has data mining capabilities, it will be able to give rewards to the customers and provide personalized discounts in order to maintain a positive relationship with its best customers (Dyche, 2005). Through using data mining techniques on the data about what the customer has already purchased, Spikes would be able to offer products and services that satisfy the needs of the customer (Dyche, 2005). The purchase rates would be increased dramatically by personalizing the contents of the website according to the visitor’s web profile (Dyche, 2005). Marketing expenditure can be altered per customer (Dyche, 2005) based on the value the customer will generate for Spikes over his/her lifetime. Data mining will also enable Spikes to analyze the data gathered from a variety of touch points to predict the customer’s future purchases. The customer is provided promotions and deals according to this prediction. To improve the website design and understand web use by the visitors, high web traffic can be related to customer segments and individual visitors (Dyche, 2005) with the help of data mining techniques. Bibliography Dyche, J. (2005). The CRM Handbook. Delhi: Pearson Education. Manifold Data Mining Inc. (2009). Customer Profiling. Retrieved December 2, 2009, from http://www.manifolddatamining.com/html/services/service131.htm Performics. (2009). Clickstream Analysis. Retrieved December 2, 2009, from http://www.performics.com/performance-marketing-solutions/conversion-optimization/clickstream-analysis Read More