How a Data Driven Marketing Database is Different than an Operational Database - dummies

How a Data Driven Marketing Database is Different than an Operational Database

By David Semmelroth

As a database driven marketer, you should know that your operational systems are not designed to answer the kind of questions that you are asking. Your accounting department, for example, could quickly tell you how many widgets your company sold between 3 and 5 o’clock on Monday. What they probably can’t do is tell you much about who bought them. That’s your job, in a nutshell.

The questions that naturally occur to you are not top of mind to the folks on the operational front lines. But even if they were, the data may simply not be available to them. Operational systems tend to operate in data silos. This means that they only store the data that is directly relevant to their specific function.

A movie ticket kiosk doesn’t need to know how far a moviegoer lives from the theater. But if you were trying to decide how widely to mail a movie ticket offer, you would certainly care. You might also care how old they are and how many kids they have. These things could affect how you might communicate with them.

What you need is a database that’s organized around the customer. And that means pulling together data from all (or maybe just some) of these data silos and organizing it in a way that serves your marketing needs. This is the basic notion behind a marketing database.

In your discussions with your IT partners, you will almost certainly hear someone refer to the ETL process. ETL stands for extract, transform, load and is in reality three distinct sets of processes:

  • Extract: Pulling data out of the operational data silos. The most important thing IT will want to know from you is how often extraction should be done. For example, do you need the data every night? Or is weekly enough?

  • Transform: Anything that is done to clean up, standardize, consolidate, or otherwise make the data ready for prime time. Whenever a change is made to the database, you should be asked to approve what is done here.

  • Load: The actual loading of the database.