Direct marketers depend on marketing databases to steer and evaluate their marketing programs and to better understand their customers. However, few marketers know enough about how their databases work or the relationship between their marketing efforts and the database. This knowledge gap exposes them to multiple error and financial risks, even when seemingly knowledgeable vendors are involved. The following red flags represent the most common causes of database marketing errors and unnecessary spending:
Databases must be optimized for specific functions. A properly built marketing database is not one database, it is three — one each for execution; reporting and campaign management; and business intelligence and analytics.
When databases rely on a single structure to meet all needs, they are forced to be optimized for execution, complicating reporting and analytic tasks and impeding data access for marketers.
Marketers and their agencies can cause problems in the database as well. Improperly designed surveys, web forms and business reply cards can introduce inconsistencies in the database that impair campaign management, execution, reporting and analysis.
For example, if you ask the same question three different ways with three different answer sets across three different surveys, the data that flows into the database from these surveys is likely to cause errors in the campaign, reporting and analysis tasks that rely on this data. To truly seize the opportunity presented by self-reported data, data collection must be reviewed and coordinated to ensure that the data collected are consistent, usable and error-free.
Data interfaces must accommodate a constant stream of changes to survey questions on an ongoing basis. Inflexible interfaces require moderate to significant programming work each time a survey change is made, which is costly and error-prone.
Marketing databases should only use data-driven interface schemas that do not require programming when marketing needs change.
The value of a current or potential customer is usually far too great to let missing or incorrect data prevent processing. That is why most database vendors use some kind of error-correction process. But when that process is manual, it can be costly, involving availability of resources for phone calls and research.
Automating the process is more efficient, but can be costly upfront. The best approach is to take proactive steps to avoid issues before they get to the database. For example, a Web form that asks for “channel preference” should display an error when a customer selects “e-mail” but leaves “e-mail address” blank.
David Bernard is a managing director of DB Marketing Technologies. He can be reached by email at [email protected].
