Remember that you are unique. Just like everybody else.
Database marketing is the art and science of taking individuals, grouping them based upon common characteristics and speaking to them with language that feels personalized. It’s getting the right messages to the right people at the right time. Done properly, it can increase conversion rates, reduce marketing costs and encourage greater engagement. Here are a few basic “spring cleaning” tips to get your database ducks aligned…
E pluribus data.
If your data lives in various systems, start by finding a relational database platform, such as SQL Server, Oracle, IBM DB2 or MySQL. Assuming your record counts don’t run into the high millions, you could even start with a program like Microsoft Access or FileMaker Pro, which both provide a nice entry into the wonderful world of relational databases.
Chances are high that your data isn’t going to be as complete as you’d like. You may have missing data sources, missing fields or values, or “dirty” (unstandardized, full of typos) data. In my opinion, dirty data is the worst and needs to be addressed immediately. This should include standardizing the values in your fields and performing regular NCOA and CASS updates on your address data. For example, if your database has records with “State” values of “CA,” “Cali.” and “California,” your analyses by state will be inaccurate. This is a small example, but it shows how bad data impacts every part of your database marketing efforts.
Remove all variations of the same value.
From the example above, “CA,” “Cali.” and “California” are all variations of the same state. If this state data was part of a mailing address, it would make the most sense to update all three to “CA.” This also extends to fields like “Source of Business.” If you have distinct values for every possible marketing campaign or event, you’ll be hard-pressed to segment them later. It’s usually better to use a handful of generic codes in these fields, with the specific circumstances stored separately as “Activities” or “Interactions.”
Perform an NCOA analysis every 3–6 months.
NCOA is the U.S. Postal Service’s National Change of Address database. This step will help cut down on returned mail and help keep your geographic data updated, which will improve the accuracy of geographic segments and analysis.
Routinely perform a CASS analysis.
CASS is the Coding Accuracy Support System, which the USPS uses to evaluate street address information. It includes a standard address format that reduces the effort and manpower required to deliver the mail. Mailers using it may qualify for a discount. It adds missing data, such as Zip Code, City or State, to ensure a complete address. Like NCOA, CASS-standardized addresses improve the accuracy of geographic segmentation and analysis.
Eliminate duplicate records from your database.
When you start acquiring data, the number of duplicate records you have will probably be fairly small. That said, your industry only has so many companies, and after a while, you’re sure to start seeing some familiar faces turning up in multiple records. Before long, you’ll find out it’s a very small world after all. You can try to eliminate duplicate records at the very beginning of your duck-queuing process, but there’s a good reason to wait. You’ll think you’ve eliminated them all, but they’ll be there, lurking. Soon you’ll find them hiding behind friendly looking abbreviations (“Street” vs. “St” vs. “St.”), in typos or living under an assumed name (“Bob” vs. “Robert”). Knowing this, the surest way to identify and eliminate as many of them as possible is to apply your cleanup process first. Once your data’s in shape, duplicate records will have fewer places to hide, and you can merge or purge those records with greater confidence.
I hope these tips will help you start down the road to a successful database marketing program. From here, you can begin segmenting your audience, personalizing your campaigns and messaging, testing custom content, and watching your conversion rates climb and your marketing costs fall.