In our previous articles on the Acquisition Model (Parts 1, 2), we discussed the importance of a vibrant, targeted email database. This enables you to use your data for improved targeting in your direct messaging campaigns and to better measure the performance of your acquisition sources.

Once you are ready to start acquiring customer and prospect information to enhance contact profiles in your database, you must be properly prepared to store it in the most effective and efficient manner possible. Today’s article will discuss strategies for setting up a robust database, and tips on how to use it to your best advantage.

1. Setup
Ensure that the database is set-up in a flexible way so that it can handle multiple sources. This is important because a record may be identified as originating from multiple channels. For instance, an individual may complete a form online, and also provide their info at a trade show. You may also want to track websites as a source but also have it tracked as a sub-source for how the contact was directed to the website (for example, paid search, display ads…).

A proper database set-up and tracking mechanism is crucial when trying to evaluate the quality of sources. For example, it’s important to be specific when naming your Google Analytics / tracking tags. You want to be able to know whether a click came through from a linked image or body text in your email.

2. Document
Have a data dictionary available for reference for all parties involved in campaign and tracking execution.  Include high level information as well as in-depth technical specifications so that it caters to all individuals. This way only one document needs to be maintained, limiting the chances for “out of sync” information to circulate. For example, a shared word document on your secure local network between analysts, marketing managers & the data team, can keep everyone informed. This also helps ensure that the analytics and marketing teams are on the same page from the very start.

3. Be Aware of Your Sources
Fresh data is better. Sources that have been collected with anything less than explicit permission, and sources that you haven’t used in the past, require a higher level of list hygiene and validation practices. It is quite possible that older email addresses may no longer be valid, and they may even belong to entirely different people!

It’s important to create and implement data into your database in a streamlined, efficient manner so that records do not become outdated. And, with data coming from many different sources it’s inevitable that you will encounter many different data formats and structures. Ensure that the system is flexible and that you have the expertise on hand to handle accurately consolidating varied types of data files.

4. Monitor & Test
Reports should be automatically generated for marketing managers (by email for example).  This will allow them to monitor how many records are coming in from each source on a daily basis. If a source is generating no results for a few days, it can prompt an investigation into potential issues.

It’s important to involve key stakeholders in your testing process. Be sure to share an example final report early in the process, to ensure the correct data is going to be captured. Then, when the new data feeds are set-up, make sure the data team involves those who will be viewing reports on a daily basis in the testing process.

5. Analyze
Analyze relevant metrics and focus on those that make the most sense for your situation. For example, if you currently have a high need for creating awareness, and email open rates are proving to be the most important for creating this, you may want to focus on the source(s) that provide the highest opens. Cross reference your sources with relevant email metrics to help determine their value. Be sure to consider sample size and double check that it is large enough to make confident marketing decisions.

Well-structured databases enable stronger analytic insights which are what drive your digital marketing strategy forward. Think of analyzing metrics not as the last step of the program, but rather a phase in a continuous cycle where insights drive optimizations to the program.