|
Database Mining What is data mining? 
Data mining involves using the today's increased computing power, coupled with advanced statistical techniques, to discover
useful patterns and relationships in large databases to drive business decisions and increase company
sales. Data mining thus involves the creation and/or subsequent enhancement of a database to create a
profile of "good" prospects. The database can be created from internal and/or external sources. MRA can
model your data and create a set of measures or criteria for you to use in your marketing efforts.
There are a host of different methods used for modeling the data. Selecting the "correct" method is not easy, because
a "one size fits all" mentality is clearly inappropriate in creating a data mining
solution. MRA's expertise in creating sophisticated data mining models, which have been applied to the
pharmaceutical, industrial manufacturing, credit, banking, and other industries will help you create a
highly effective marketing program from your in-house data warehouses.
Why use data mining? Data mining allows you to focus your resources on marketing efforts with the highest potential. This saves you money, time and effort. For example, using data mining to help guide a direct mail campaign can save money by eliminating those persons who are least likely to respond. Often, the simple savings in postage - not to mention print material production costs - are sufficient to pay for the data mining effort. Alternatively, you can target specific efforts at selected respondents. If we identify a substantial cross-selling relationship among persons in selected industry groups, then targeting non-purchasers with a structured offer that highlights the benefits of using both products simultaneously is an effective marketing tool. Data Sources There are two primary sources that are used for data mining. Internal data is collected as part of a company's normal and usual business activities. Examples of this would be product sales tables, accounting records, customer relationship tables, and the like. This information is stored in your internal corporate databases. Typically such data is easily accessible by a company, and does not cost an exorbitant amount of money to obtain. An example of this sort of data and its application to data mining is our financial services product, CVI Profit Manager. Alternatively, external data may be acquired and used for data mining activities. A host of different data elements may be appended to customer lists, and the mining can focus on that information which is available from these public sources. These outside sources can give you a wealth of information about your existing clients. With data mining, we can then profile your typical customers and use that information to acquire list of potential new clients. Knowledge Driven vs. Data Driven The analytical techniques used in data mining can, broadly speaking, be broken down into two main groups. When used effectively, most data mining projects use both of these approaches. Knowledge-driven data mining techniques use pre-existing information to guide the analysis. As an example, the hypothesis that customer loyalty, as measured by a survey, is related to increased sales to existing customers (corporate database) can be tested in a data mining project. Alternatively, data driven models search the data for underlying patterns which are then isolated. Most efforts at developing market segments using cluster analysis will fall into this data-driven category. Both approaches have value and should be used to better understand your customer base. A vendor who tries to tell you that one approach is always right for you is probably not taking the time to truly understand your needs, or maybe they don't know how to use multiple approaches. MRA's statistical and database expertise will help to ensure that your investment in data warehousing takes the next, crucial step toward increasing sales and profits. Contact us for further information on how our data mining solutions can help you achieve your business goals. |