If there are quality problems with your data, there are ways to clean it up — but it’s often more efficient to refactor your processes to prevent “smelly” data.

Sourced from: sloanreview.mit.edu

As someone who often applies Agile programming practices to business and marketing efforts, I really like the approach of using refactoring techniques to clean up data.

Even for our small business and organizational clients there is a lot of data that can quickly start to "smell" with no commitment to standardization and keeping the data clean. Common examples include Customer Relationship Management (CRM) solutions, e-commerce customer lists, email subscriber lists, and accounting data.

Of these, it is often the CRM solutions that create the biggest problem, with out-of-the-box fields that seem self-explanatory but in reality are subject to a wide range of interpretation. To extract the most value out of a CRM solution, the most vital action is not the construction of the system and fields themselves, it is the construction and defining of how the fields and contained data will be used.

I recently had a client with only 5 employees where their e-commerce reporting could not even be done consistently because it turned out that each person had a different interpretation of how to use various intermediate order statuses. On the surface each status seemed "obvious" and therefore had not been discussed. It wasn't until the data was needed for reporting that problems started to arise.

A big part of what we do at Idea Spring is help clients with the pre-planning, and developing appropriate definitions and guides to various data systems. We use an internal process and method that helps us dive into how the data will be used to create actionable intelligence and achieve mission-level objectives.