How to Successfully Manage One-to-One Campaign Databases

The key to success when deploying any one-to-one communications campaign is proper database management. It may seem like a foreign subject to many print providers new to variable data printing (VDP) or cross media, but it can make or break any campaign.

We have all, at one time or another, received personalized direct mail or e-mail messages with a name misspelled or an irrelevant offer. Not only are these communications ineffective, but they can damage the reputation of the organization sending the communication. In this article, I will highlight the ways you can prevent data mishaps and ensure your client’s database is campaign ready.

A Clean Database Is a Happy Database

A clean database ensures the right message with the highest relevance reaches recipients. It also means fewer errors and returns on print and e-mail messages, reducing wasted resources. Though it is usually the client’s responsibility to supply good, clean campaign data, providing a data cleansing service may be a great opportunity for your operation to expand into a new service-based offering and generate additional revenue.

When cleaning a database, it is good to avoid duplication of effort and do as much as you can at once. If you are in the database cleaning up addresses and phone numbers, make sure other data meets standards as well. For example, are names all upper case when you prefer mixed case? Or are there extra spaces in text fields or unwanted characters, such as percent or pound signs floating around? Handling any potential problems at one time will take slightly longer, but the effort will pay off greatly in the long run.

It is also helpful to have another colleague who is familiar with the client, project, and information to double check the data. When reviewing similar information repeatedly, you can often overlook glaring mistakes.

Lastly, once the database is clean, it is imperative to develop a plan to ensure the cleansed database does not lapse back into disrepair as existing data is altered or new data is added. It will make the process before deploying a campaign much faster and easier if you know the data has been regularly maintained.

An important place to look when planning for ongoing data cleanliness is in applications that write back to the database. Some cross media solutions, for example, have a built-in rules engine that not only reads databases and applies business rules, but also has the ability to apply business rules or logic to data being written back into a database. This capability provides a way for a set of rules to be created that automatically ensures new data meets existing rules.

Use Applications Designed for Data

Many organizations store data in applications that are not designed to be databases. For example, Excel is often used and referred to as a database, but in actuality it is an application designed to perform calculations. That is why several issues come up when running a VDP or cross media campaign from an Excel spreadsheet. Who among us hasn’t had a zero drop off the front of a zip code when working with mailing addresses in Excel?

To avoid potential data pitfalls, use Access, SQL, Oracle, or another application designed for data. If a client provides you with an Excel spreadsheet, you can simply convert the file to CSV format and deal with any potential issues, such as missing zeros in zip codes, during the conversion process.

Another option is to use a VDP or cross media application that allows you to write business rules addressing issues, such as the common zip code problem, ensuring your data appears perfectly every time. Many companies use both techniques in conjunction. That way, if something slips past the conversion process it can be caught by the business rules.

Limit Freeform Text Fields

There are many instances when it seems necessary to include freeform text fields in a database, such as when entering the name of a product or service level. However, this manually entered information can create more harm than good, allowing errors or typos during data input, increasing the likelihood of incorrect information in the final product.