Despite the affinity of software publishers with computers yielding sophisticated database capabilities, the events leading to structured data capture, organization, and implementation is, in many instances, ignored.

You first have to decide what type of software will be used to maintain your internal house file. With all of the data entry and database software packages on the market, only careful analysis will determine which one best suits your business needs and hardware platform. There are a few things you should keep in mind during the selection process.

Look for a software package that lets you build customer records to your specifications, allowing the capture of data you need without the constraints of someone else's template.

You should also ensure that the software allows you to import and export files to and from your system to other industry-standard formats, such as DBF, fixed-field, quotes, or comma delimited. At some point, you may want to apply external processes to your data, such as file enhancements or overlays. Such processes require that your data be moved out of its proprietary format to a format common for performing these functions, allowing for the resulting information to be imported back into your internal system. Database products such as Alpha Software's Alpha Five or Microsoft's FoxPro deliver flexible, powerful environments for the warehousing of information.

Once the software is selected, the next critical stage is selecting the information you want to capture. A general rule: the more information you know about your customers, the better your house file is as a strategic marketing tool. So what should you be capturing besides the basic name and address data? There are a few highlights that can yield your organization many long-term benefits in the bigger picture of installed-base marketing.

The date is an important component to the house file. You should retain the date that the customer was originally captured and the last date of purchase, allowing the analysis of 'lag time' between acquisition and actual sales. Products codes and version numbers, stored in a relational format, provide you with the information needed for performing both upgrade and cross-sell efforts.

A source code, indicating where the name came from, is also important for performing back-end analysis on the effectiveness of your advertising and marketing efforts. A promotion code is also useful for tracking the various promotional programs that you have instituted, i.e. - competitive upgrade programs, trade show pricing, various direct mail offers, etc. Email addresses, captured at the individual name level, and web site addresses should also be warehoused. The email addresses can facilitate broadcast communications to your customers via the Internet which can be excellent and inexpensive method of maintaining continuity with your installed-base. And if you are collecting Email addresses, make sure to ask permission for use of the address up-front. It is good practice which will yield addresses useful for your own sales efforts, exchanges, and possibly future rentals.

There is, however, one key method of structuring, analyzing, and segmenting your house file. It is a proven method that has been around for over twenty years that will help you maximize profits when performing installed-base marketing activities. It is called RFM, and it stands for Recency, Frequency, and Monetary value.

By coding your file with RFM, you can rank your customs from 'best' to 'worst', allowing you to target the ones that are most likely to make additional purchases. Let's first take a look at the individual components and how they work together in the segmentation processes.

Recency: It is generally true that a customer who has made a recent purchase is more prone to buy again as opposed to someone who's initial purchase was a long time ago. Recency is a popular selection criteria used by brokers when acquiring outside list media -- applying recency will typically yield a performance boost which could lead to a profitable response rate from the external list segment.

Frequency: The more frequent a customer purchases, the more likely that individual will respond to a new offer. It is a primary measure of customer satisfaction. Frequency can be derived in several ways. When you are doing an initial merge/purge on your raw customer data, duplicate records should be collapsed into one record, maintaining the individual product codes for the items purchases. Making a query into customer purchase history will then reveal an accurate amount of product code entries which reflect the number of overall purchases made by that individual.

Monetary Value: It is the final RFM component, Monetary Value, that truly clarifies both the Recency and Frequency factors. Surly, someone can make numerous and recent purchases, but only spend a very small amount of money. Monetary Value can be equated by cross-referencing your product and promotion codes to determine how much money that individual has spent.

In a typical RFM model, the house file is segmented into quintiles with a simple coding method. Sorting first by date, divide your database into five equal segments. The top 20% (the most recent) are coded with a 5, the subsequent 20% are coded with a 4, etc. proceeding down to 1. The same procedure is repeated for frequency and then monetary value. When you are finished, all of your customers will have a 3 digit code in their record ranging from '555' (best) to '111' (worst). There are 125 different RFM iterations in all. Recoding should be done when your house file is updated with new names -- perhaps monthly or quarterly, depending on your needs and usage.

The RFM coding method allows you to stratify and test different segments of your house file -- specifically, you will know how deep into your own house file to mail before it is no longer profitable, which saves money while improving the overall performance of your installed-base efforts.

Another example is a specialized promotion for reactivating the 'least best' customers. Or, how about a special offer for your 555's -- the individuals who have a tremendous potential on purchasing multiple products and act as a good PR and referral source? How about sending your expensive, corporate magazine to only your very best customers, instead of mailing random segments of your house file? RFM offers marketers many possibilities, and it is very simple to implement.

Take a hard look at the condition of your house file, plan a strategy with your marketing staff, and take control of what is probably the single most important element in the overall profitable operation of your software business.