The relational structure depends on how many data elements need to be maintained. The structure of a typical direct marketing database may include one or more of the following segments of information: Individual Information, Demographics, Purchase History, Promotional History, and Survey Data. These are explained in further detail:

Individual Information

The relational segment for Individual Information contains any information that describes the individual, such as name, company name (if non-residential), address, city, state, ZIP code, phone, and email. In addition, a customer-type field may be established to indicate a piece of information that may be relevant to the target marketing efforts, such as Purchasing Agent versus End User, Customer versus Prospect, Student versus Professor. Many types of code may be established depending upon the number of ways the individual is viewed.

If required, MSC will append and maintain a mail flag to the individual record. The mail flag will contain a flag that indicates if the individual has expressed a preference to not be mailed. During all selections for mailings, the mail flag becomes part of the selection criteria to ensure that only individuals who have not expressed a dislike for direct mail (or telemarketing) are selected.

In addition, on a regular basis, MSC can match the individual record to the Direct Marketing Association ("DMA") Mail Preference Service file that contains the names and addresses of those individuals that have contacted the DMA and requested that their name be removed from all direct mail efforts. Those records that match can be flagged with a do not mail value in the mail flag.

Similarly, the DMA has a Telephone Preference Service that contains the name, address, and telephone number of those individuals that have requested to not be solicited by telephone. MSC will match the individual file to the Telephone Preference Service file and flag the matches as "Do Not Call."

More information can be obtained about the DMA Mail and Telephone Preference Services at the following Website: http://preference.the-dma.org/

Demographics

Other types of information that may reside on the individual file are demographics. Demographics can be used to target prospects or existing customers for various new offers. Demographics can include information specific to the individual such as title, PC software used, dog owner, married or single, or any information that may be pertinent to making target marketing decisions. Alternatively, the demographic information could be specific to the household or business for the individual, such as number of employees, SIC code or own-rent flag, number of children in household and/or number of rooms in house. This information can be obtained by questioning the individual as they place an order, or by appending the data from a compiled database.

Purchase History

Much useful target marketing information is derived from purchase history. The person most prone to buy your product is one who has bought it before. The tried and true rule of thumb for many direct marketers is to select individuals for a mailing based on recency, frequency, and monetary. These statistics are derived directly from the purchase history of your customers.

Recency is defined as the last time your customer purchased from you. Frequency is defined as how many times your customer has purchased from you. Monetary is defined as how much your customer has spent with you. These three statistics can usually predict the most likely candidates on your database for your next direct marketing campaign.

The most obvious source of purchase history information is from the in-house or outsourced order entry system. Inbound and outbound telemarketing orders, inbound direct mail orders, and internet orders are usually placed in some sort of order entry system. Usually these orders are then easily convertible into the marketing database.

Online registrations of product purchases are an excellent source of purchase information. The challenge with online input is to clean the data enough to eliminate erroneous data that can be entered by the customer. MSC can advise you in your online registration forms on how to set up the screens that can minimize the possibility of unusable data from being entered by the users.

The next best source of purchase history information is from the retail chains that sell your product. The challenge is to get that information. Warranty or registration cards placed in the box of your product will prompt some consumers to respond. Incentives to respond may cost extra money, but may be justified by the lift in rate of return of the warranty or registration cards. Rebates are also a way to obtain the name and address of the buyer, but at a cost.

Another source of information is from inquiries. These consumers may not have purchased, but have expressed an interest in your product. Next to actual purchasers, these individuals are the next best target. Inquiry information can be gathered from a variety of sources. Direct information requests are the most obvious and easy to obtain. Bingo or Reader cards in magazines are a good way to learn of interested consumers in your product. In addition, trade shows and seminars are good ways to obtain names and addresses of individuals with interest in your product.

Promotional History

Promotional History is another segment of the relational database that can provide insight for future target marketing efforts. How often do you have to market to an individual before that individual responds? At what point do you stop soliciting? These questions can be answered by compiling and analyzing the promotional history. For instance, you may be able to save on promotional costs if you only mail a postcard announcing your new product upgrade is now available to those individuals who habitually upgrade immediately.

The other advantage of compiling a promotional history is that you now have the ability to calculate the cost to make a sale to a customer, and analyze the lifetime value of a customer based on the original source of the name. For example, if you rent a list for a prospect mailing, you may evaluate the value to your marketing campaign based on the response rate of that list. Alternatively, you could look at the 12 month or lifetime sales of those respondents and find that it could outweigh or negate a simple response rate.

Survey Data

Simple survey data can be housed with the individual information, but complicated surveying will need to be housed relationally. One key complication in tracking multiple surveys is to track the same question and same responses across surveys. If this is done successfully, then a select of all records that responded "50" to a certain question, regardless of the survey becomes possible.