Strategy: Maintaining a Relational vs. Flat File Marketing Database
A housefile, or customer list, is a valuable marketing tool if you maintain it properly. It enables more targeted marketing; facilitates various analyses; and generates incremental income from renting and/or exchanging names with other reputable mailers.
When talking about building and maintaining a customer housefile, there are two issues you need to address: 1) collecting data into your order entry system, and 2) extracting relevant subsets of this data to build a marketing database. If you circulate multiple catalog titles or have a large file of records for which you’d like to see transactional history and do complex queries, maintaining a relational database (vs. a flat file) might be the way to go. There are advantages to both.
A relational database enables you to look across all catalog titles to determine who’s purchasing from your other catalog titles or sales channels. Relational databases are different data tables, i.e., multiple titles that relate to each other. A flat file is one-dimensional; a collection of data in columns and rows, often presented as a spreadsheet. If you circulate more than one title and cross mail to your housefile via a relational database, you can determine buying patterns in addition to standard recency, frequency and monetary value (RFM) analysis. It will enable you to mail “smarter” to your housefile. (For instance, you’ll better know how often to mail, which groups to mail and what title to mail.)
Cost is an important consideration with regard to setting up and maintaining a relational database vs. flat file database. One is more expensive than the other. Carefully consider the advantages of a flat file database vs. those of a relational database before you decide which way to go. Here are some points to consider.
Flat File Database
l Initial set-up and development costs are less than a relational file.
l Marketing reports are fairly simple and typically meet most of the marketing needs. If other reports or variations are needed, programming costs could result.
l Often there are no online capabilities.
l RFM information is rolled up with only the first, last and to-date order information visible. No transactional information is available.
l Customers who are multidivisional purchasers can’t have their information viewed across divisions.
l Set-up costs are higher than a flat file.
l Purchase activity is viewable down to the item level.
l Customers can be viewed and selected across multiple titles as well as multiple channels.
l Online access makes it possible to view information and perform selections.
l It allows for full analytical capabilities.
l Promotional activity in e-mail, retail and catalog can be viewed.
l Because of the information it maintains, lifetime value as well as other analysis can be developed.
l All of the history for each order is viewable.
The more catalog titles you circulate, the more a relational database makes sense. If you maintain a flat file, the data relates only to a specific catalog title or brand. Develop a circulation plan for each title independently. If you maintain a relational database, look at customer purchase history and behavior across all titles.
There should be no duplication. This protects the integrity of the data and reporting, and actually can help you reduce the number of catalogs you send to any given customer. It also can help you target which title a customer should receive, etc. You actually can improve the effectiveness of your mailings while reducing your selling expenses.
Some other benefits of a relational database include the following:
Applications for roll-ups and aggregates to various levels using detailed information on the database. With a relational database, catalogers can use the product category level information (even down to SKU), purchase channel and repeat purchase rates for any given period and from any given segment — and back validate RFM selections. With a relational database, you can view full transaction history so you can see, for instance, if a customer has only purchased on the Web or through the catalog.
Application for future housefile promotions. With a relational database, you can segregate your best buyers based on various queries. A cataloger can keep narrowing down its criteria until it has created its best buyer file for the desired quantity.
Application for merge/purge. You can query housefile segmentation to verify service bureau accuracy. You also can keep scores and other demographic selects on the database This, as well as housefile promotion information, can help you create a contact strategy. When multiple titles are processed on a relational database, models can be built for cross-buyer usage.
Housing Database, Updates
Maintaining a database in-house requires a lot of time and resources. If there are any software or hardware updates, have them maintained. Service bureaus can manage the updates, training and changes to the database for a fixed amount every month. It’s difficult, however, to match up an internal database used for data maintenance and storage to an outsourced database. Different factors are used when creating both, as well as different software and logic to update and maintain each one. I recommend that a relational database be updated monthly, or at the very least, quarterly. Keep the data current.
It’s important to decide in advance what information you want to capture, how you will capture it and what you’ll do with it once you have it. Often, I see companies go to great lengths to capture data they’ll never use. Your approach must be convenient and scalable. For multititle catalog companies, maintaining a relational database can improve your selling expense to sales ratio by mailing to the “right” people at the “right” time. «
Stephen R. Lett is president of Lett Direct, a catalog consulting firm specializing in circulation planning, forecasting and analysis. Lett spent the first 25 years of his career with leading catalog companies, both B-to-B and consumer. He is the author of a new book “Strategic Catalog Marketing” (Target Marketing Group Publications, 2006). You can reach him at (302) 537-0375 or by e-mail via his Web site: www.lettdirect.com.