Data Driven: Data Dealings
In any company, strategy development for future planning hinges on one critical assumption: the data used to drive decision making is sound and the data's integrity is uncompromised, robust and reliable. It starts with making sure the data is correct on the input side of any database. Here are two issues common to all data input practices that marketers must deal with:
No. 1: Quality
To further marketing strategy development down the right path, it's crucial to invest in quality checking (QC) your data program. Always take the time to do a reality check on your data and ask questions. Let's look at a couple of case studies:
• Case Study No. 1: A business implemented new segmentation to expand its media channels that would divide the customer file into groups based on Average Order Value (AOV), among other factors. When the initial data was delivered, the analysts looked to see that all the proper fields were accounted for and held values. They gave the OK and the planning and decision-making process began.
However, there was a problem with the data, and if they had done a "reality check"—asked "Do these numbers make sense?"—they would have seen it. The problem: All the merchandise had not been included in the "Gross Sales" definition. What the analysts should have seen is that discounts made up over 50 percent of gross sales. Unfortunately, the data looked acceptable when given only a cursory glance and was passed along. As a result, the merchandise total that was calculated from these two fields was wrong, which threw off the AOV segmentation and tainted all decisions made downstream.
• Case Study No. 2: An email campaign was designed, coded and sent to IT for deployment. The creative department had already tested all links to verify they led to the appropriate landing pages. But when the file arrived at the IT department for launch, IT made a change to the opt-out link and re-saved the file. When ready to deploy, the only link tested was the one that had changed. Time had run out—the file needed to be sent immediately and it was assumed the other links should not have changed. Unfortunately, something did change. The result: 1 million people who received the email could only click through to an error message instead of the main offer presented.