Data Dilemmas
Business-to-business databases have never been celebrated for their cleanliness. For starters, B-to-B data is complex and fast-changing. Customers conduct transactions through multiple channels, and they often provide conflicting information. People leave jobs. Then there are the faulty legacy systems maintained by many B-to-B marketers.
What can you do?
We asked some experienced database professionals to share their most serious challenges. We then set about finding practical solutions.
DATA PRIORITIES
I don’t know which of my data elements is the most important. Which fields have the most impact on sales results?
The best way to find out is to model your data using multivariate regression. Make sales the dependent variable, and let the other elements sort themselves as the independent variables. You will have your answer fast. But there are two caveats:
-
A model costs real money to build. You’ll need to budget north of $20,000 for this exercise.
-
MISSING DATA
Models have a shelf life. You’ll want to redo the model at least annually.
The plus side? You’ll know which data elements are most important to maintain and keep fresh. And if you find some elements with negative correlations to sales, you may be able to use these fields as a negative predictor
