How to Create a Customer Database

[Note : Because of dramatic advances in technology since the publication of this article, certain technical points are no longer applicable. However, the majority of the article focuses on the business issues that are involved in building and leveraging a marketing database. These remain very relevant.]

So, you have just been assigned marketing responsibility for your company's new customer database. Congratulations. If you are like many direct marketing managers, you may be wondering what your first steps should be. You know what is wrong with your current customer file, but you are a bit intimidated by the constantly changing technical environment and the difficult-to-understand "buzz words."

Marketing managers must understand the basic principles of database construction in order to help insure that their objectives will be met. You need to rely on technical experts, but you can be helped by a modicum of database education. The purpose of this article is to help non-technical managers understand and become articulate about the key elements involved in building a database that will best meet their needs.

Building a customer database requires many skills. However, from a management or marketing perspective, there are four important non-technical skills that you may not hear much about in these days of "database" hype:

  • Communication "All users and other participants must be honest with themselves and each other as to the needs, uses and limitations of such a system. This includes formalized documentation of perceived database needs, as well as informal day-to-day brainstorming. Put aside the old rules and look at your needs with fresh eyes. Operational people can be particularly helpful in the design stage, because they have to deal with the results of poor data design and retention.
  • Common sense "Know your business. Each issue must be thought through, identifying the options that make the most sense for your business. If you do not know the answer to a particular issue, find someone who can help. Database design is rapidly evolving as new solutions are developed.

Sometimes technical experts, both internal and external, can become get bogged down in the search for the "perfect" solution, ending up with one that is not entirely practical. Retain your ability to step back from the "experts." After all, you know the most about your business. The final decision rests with you after you have thoroughly evaluated the options.

  • Creativity "If some need appears to be impossible to meet in a cost effective manner, seek other options. If building an on-site database, be creative in thinking through your specific needs. If, like most people, you find it more cost efficient to use an outside service bureau, work with the technical experts there to find creative applications of their existing systems for your particular needs.
  • Attention to Detail "What appear to be small decisions can lead to inefficiency and high cost if not thought out carefully. No matter how advanced the technical aspects of your database, they will be worthless if they result in a clever solution to the wrong problem. Remember, you are the one who knows what you need.

"Customer Database" Defined

"Database" is a word that is often used and misused by today's direct marketers. There can be many nuances to its interpretation. For the sake of this article, it will be defined in the following way:

A compilation of data about each consumer who has interacted with the products and/or services offered by a company, division, or catalog. A database maintains individual data elements that can be brought together in an efficient manner to create useful information. They can be manipulated in order to meet many needs of an organization, including product fulfillment, customer service, the generation of management reports, quantitative research, and the selection of customers for subsequent contact.

Here, "customers" are assumed to be individuals or companies that have responded to you via mail or telephone. They are qualified buyers because they have spent money on your product. This makes them most likely to purchase again, and should allow you to contact them profitably in the future. They form the basis of a successful direct marketing operation.

Inquirers are different from customers in that they have identified themselves as having some interest in your product, but they have not committed themselves. Inquirers might include those who respond "no" or "maybe" to sweepstakes offers, spend a nominal sum for a special introductory offer, or request a free catalog.

The uninitiated sometimes believe that acquiring a customer database is as simple as "buying" a mailing list and treating the names as though they were customers. In reality, a customer database is built slowly, as you test, expand, and re-test name acquisition methods and media. More importantly, a database is much more complex, flexible and efficient than a traditional customer file.

You are likely to have heard the term "relational" database. Without getting into technical details, "relational" simply means that data is kept in separate yet linked logical groups. The structure is such that tasks can be accomplished without having to pass every individual record during processing. Thus, processing options are very flexible.

The alternative to relational access is what is known as "sequential access." As its name implies, sequential access databases must pass every individual record to execute a task. This results in processing that is somewhat less flexible than for a relational database.

Relational databases are often held up as the ideal by many direct marketing consultants. The visionaries talk about ultimately flexible relational systems, in which every data element that a direct marketer might ever conceivably need is kept on-line for instant access and manipulation.

This view is somewhat simplistic. The unfortunate fact is that, with today's technology, such systems are impractical for many businesses. In the real world, relational databases are superior only for some applications, such as individual record lookup. For many other applications, such as large-scale transaction processing, sequential access databases are superior, because transactions can be applied much faster, and therefore more cost efficiently.

It may be helpful to keep the ideal of a relational database in mind, but it is important to be practical about the cost-effectiveness of such a system for your particular business.

The key thing for the non-technical direct marketer to remember is that the structure of every database must be carefully tailored to the quantity of names and the information needs of the business. For example, these two businesses could not build databases from the same blueprint:

  • An operation selling to retail franchises "The number of retail outlets and quantity of stock keeping units are relatively modest. This database can approach the ideal, with every aspect of sales and service to each outlet captured in a relational format. Separate yet interconnected files can be maintained on-line to allow instantaneous updating and reporting.
  • A mailer with 6 million active customers and a large number of corresponding orders and items that must be applied on an ongoing basis "Current technology does not allow efficient processing of such a large number of transactions in a truly relational format. Therefore, a significant portion of the system must be maintained on tape, and accessed sequentially. Hard decisions must be made as to what data is kept and how summary fields can be judiciously used. Separate files can be maintained, however, for such needs as customer service.

Data Uses

It is important to understand the criteria by which you should select data worth keeping. Simply, the more useful the data will be, the more worthwhile it is to keep.

Start by thoroughly examining your needs. Put them in writing, forcing yourself to be very specific, thereby generating thorough internal discussion. That will give you important clues as to what data should be maintained.

  • Needs should be translated into dollar benefits wherever possible. This will help you more clearly prioritize your options and may give you important insights into your data needs.
  • Separate "nice-to-know" from "have-to-know" in order to establish clear priorities.

There are many reasons to create a database, but it is important for everyone involved to understand the specific rationale behind such an undertaking. Many companies create a customer database in order to help them meet one or more of the following corporate objectives:

  • Acquire new customers at a specified percentage of their expected lifetime value.
  • Generate increased list rental income.
  • Fulfill a strengthened commitment to customer service.

In terms of day-to-day usefulness, a database can be designed to accomplish a number of important functions, such as:

  • Select names for customer mailing (based on segmentation).
  • Generate management reports.
  • Tie sales into inventory management.
  • Maintain names in consistent format for suppression on outside lists.
  • Enhance customer research to better target prospecting efforts.
  • Increase list rental revenue by augmenting the selects available.
  • Improve the ability to base financial forecasts on actual data.
  • Generate more detailed input for the evaluation of test results.
  • Improve order turnaround time.
  • Speed up and clarify customer service efforts.

Data to Be Kept

As mentioned earlier, it is generally impractical to keep all possible data relating to the business. The decision to retain particular data elements should be based on the following criteria:

  • Appropriateness to the business and the ability to help achieve objectives.
  • The offsetting cost of storage, maintenance and access time.

A customer database is constructed from a variety of sources, depending on your particular business. The more traditional sources used to generate customers include outside lists, compiled lists, package inserts, space ads, friend-get-a-friend offers and retail customers.

There are certain primary types of data that should be included for each customer and are, therefore, the foundation of any new system. You must take these building blocks and add to them as appropriate for you. Although not comprehensive, the following basic types of data elements should generate additional ideas specific to your business:

  • Recency (e.g., date of last purchase).
  • Frequency (e.g., total number of purchases).
  • Monetary value (e.g., net sales to date).
  • Product category (e.g., footwear, electronics).
  • Catalog (for multiple catalogs under one corporate database).
  • Relationship with company:
  • First purchase.
  • Customer service contact codes.
  • Purchases through other distribution channels.
  • Purchased gifts for others (and tie to the giftee records).
  • Credit card usage & card name.
  • Credit information.
  • Miscellaneous types of data:
  • Size of apparel.
  • Presence of children by age.
  • Presence of personal computer at home, by brand.
  • Demographic overlay data.

Basic Database Organization and Planning

No matter what physical form your database may take, optimal organization of the data into logical sequence and/or group is critical to its effectiveness. If you are a catalog business, you will likely need the following logical groups of data. If you are able to build a relational database, each group would be separate, but tied together by unique numbers or other identifiers:

  • Basic information "Used as base that ties all customer information together; basis for name selection if segmentation techniques have been applied. Examples: Name, address, customer number, point score.
  • Purchase history "Used to generate sales reports; develop scores. Examples: Original source, additional sources, items, returns, dollars spent, dates of purchase.
  • Promotion history "Used to generate response reports by list and medium; acts as a suppression file for specific categories of promotions. Examples: Catalogs mailed, key codes mailed.
  • Service history "Used to generate customer service reports; manage bad debt; help evaluate quality and value of each customer. Examples: Codes for complaints, excessive returns.
  • Miscellaneous data "Used for list rental; customer research; scoring. Examples: Change of address (including last address), name change (including previous name), gift purchase ("ship to" different name at different address; including a code to identify the related gift recipients), overlay data.

One of the most significant decisions concerning a database is the choice between internal or external database resources. Factors that should help you make an informed decision include:

  • Cost alternatives "It can be cheaper to use an outside service bureau than to build and maintain your own system. In any event, some aspects of any system will have to be internal, to meet timely requirements for customer service or management reporting.
  • Expertise "You need access to a knowledgeable systems expert.
  • Service "Your needs must be a priority for the staff assigned to fill them.
  • Technology "Applications software at the leading service bureaus is likely to be more sophisticated than any developed in-house.
  • Quantity of raw data "This will directly affect the number of options available.
  • Control "Your database is a significant asset. You will want to be sure that it is in trusted hands that understand your business needs and objectives.
  • Be sure that all levels of personnel are included across organizational functions.

Everyone will have to live with the new system for a long time, so they need to participate in it, which will make them more supportive and understanding.

The form in which the data is kept can be as important as the data itself. Common oversights related to the form include:

  • Data fields that are too small, particularly for the name and address. Do not short change your ability to identify your customers on other lists or hurt deliverability further by having to truncate these crucial data.
  • Data that take up more space on the record than necessary. Use codes where appropriate in order to save space (e.g., title codes, gender codes).

Companies often have elaborate procedures and controls within certain departments, such as finance or even marketing, but the database is often under the jurisdiction of many. It is imperative that controls are put in place so that a specific aspect will not be changed solely for one person's or group's expediency.

  • Techniques for changes and updates should be approved by someone who understands both the needs of the marketers and the technical people.
  • Changes must be communicated to all key areas of the organization. Many marketers have had rude surprises when changes have been made "either because of very independent systems personnel or unconcerned or ignorant marketing personnel.
  • Once again, teamwork and communication are critical at all levels and across all departments.


There is no one blueprint for building a database. Of course, there are basic design rules, but the needs of your business must be met in a way that is most cost effective for you. In the long run, one of the most important things that you can do is to be actively involved, lending and encouraging support throughout the company for this powerful business tool.