How do you know when you have had too much data lately? By and large, we have come to believe that more is better—look at the huge portions served in restaurants. But when we have had too much of a good thing, we can become bloated, slow, and uncomfortable. That doesn’t mean we can’t have the treats we love, but we must begin to practice portion control. The same applies to direct marketers’ view of data.
With New Year’s right around the corner, it is time for overweight data hogs to consider a resolution to get a handle on their “weight problem.” Some disciplined eating habits applied to our data appetites will help us become more-discerning carnivores.
Let’s begin by addressing the first step in making any change: admitting that we have a problem. We have become addicted to more and more data. We are always looking for more when it comes to understanding customers. We want more information about their needs. We want more information about their spending habits. We want more information about their preferred marketing devices. We want more information from more of our transaction systems. If that’s not an addiction, I am not sure what is.
Sometimes, however, more is just more. In many ways, it is not surprising that the most overweight nation in the world also has the world’s most voracious, and often indiscriminate, data appetite. The weight-loss tactics below should also help us streamline our overly curvaceous data figures:
Remember to work out. In the same way that an effective diet is about more than the type and quantity of food that is consumed, so too is changing traditional views about data consumption. One key to helping make the transition is to view direct marketing as more than just a direct response process. Look at ways to empower the customer touch point instead of focusing simply on driving traffic to the customer touch point. This allows marketing to view its data needs and obligations in a streamlined fashion. When exercised, this new perspective helps marketing eliminate information that wouldn’t be useful or meaningful during a real interaction.
Cook without the excess fat. Be judicious about what you add to your key data repositories and what data you add to your purchased data requests from third-party vendors. Getting all 500 variables for all your customers and prospects is not always a great deal when compared with purchasing and retaining the most predictive 20-30 variables. Additionally, challenge yourself about what the useful lifetime of historical information truly is. Don’t keep it around just because you have it. Adding history and unused variables tends to complicate your data structures and makes them more difficult to use over time.
Try smaller portions. While data storage costs have decreased dramatically over the years, the total cost of data acquisition can still be relatively large. This is often driven by the costs to extract, transform, and load information from a data source into an accessible data repository. Before incurring such an expense, consider obtaining a one-time extract from that data source and performing your data-mining tasks using that extract. If any of the new data contained in that one-time extract prove valuable, then proceed with the transformation and load processes to obtain a recurring extract that is populated into an accessible data repository.
Focus on complex carbohydrates. Not all data are created equal, just as not all sources of energy for the human body are equal. The complex carbohydrate of data is the behavioral piece of information. It is this information that is often essential to generating the most useful customer insights. Transaction systems and customer interaction points are the most common sources of this information. Most often, only transaction information is available in source systems. Key data about the interaction are often lost due to a dependence on company resources to input into arcane systems nontransactional facets of the customer interaction. Gathering this information on a sample basis (see “Try smaller portions” above) may provide some customer insight and justification for deploying customer touch point systems that improve the quality of these data moving forward.
Plan meals. How many times have you stood in front of that refrigerator or pantry and concluded that there was nothing to eat? When this occurs, often you try a multitude of solutions hoping that one will satisfy your desire. This is true of organizations that do not plan their customer contact strategies. Organizations should challenge themselves to leverage data mining to develop customer segmentation and to understand segmentation migration patterns. Armed with this understanding, the organization can focus its contacts on items it has a data-driven basis to believe will be effective in accomplishing the overall goal. This, in turn, will serve to focus the data needs used to derive the contact strategy.
The customer data diet is not a fad. It is a long-term approach to a healthier marketing organization. By actively applying some of these suggestions, you and your IT department will manage your ever-expanding data appetite more effectively.
Steve Schultz is executive vice president of client services for Quaero Corp. (www.quaero.com), a marketing and technology services firm based in Charlotte, NC.