Use Your Data to Describe and Predict

Jan 16, 2012

“Describe and predict” starts with understanding what your customers look like. Once you understand this, you can compare their profiles against well-known demographic and psychographic patterns to anticipate how they might behave. This becomes the basis of a more effective one to one marketing campaign.


The first step is to understand what your customers look like. What are their ages, incomes, genders and races? Where do they live? Where do they prefer to shop with you (brick-and-mortar store or online)? What marketing channels do they respond to (direct mail, e-mail, SMS text messaging, or social media)?


Next, filter this customer information through general demographic and psychographic patterns to predict their behavior. This includes what products or services they are most likely to purchase and how soon they are likely to do so.

Let’s look at a simplified example.

Three Separate Models

Let’s say you are an auto dealership and discover that your lease customers fall into three basic categories: young singles, families and retirees.

Because these are all current customers, you can use descriptive modeling to understand what each of the categories looks like. You know their ages, incomes and ages of their children (if any) at the time of initial lease. You know their current vehicles and the options they selected. Based on past interactions, you know the marketing channels they prefer. Next, you can use predictive modeling to match appropriate upsells and cross-sells based on the behaviors of each group.

+ In the young singles category, for example, it would be reasonable to assume that, after three years, they might be more established in their careers and have higher earning power. Therefore, you might be able to trade them up to the next class of vehicle with more options.

+ In the families with young children category, you might assume that, after three years, they might have had more children. If they currently lease a sedan, they might need to move into a minivan or crossover vehicle. Families with older children might need to move into a vehicle with greater towing and storage capacity.

+ In the retiree category, predictive modeling suggests that these customers might be looking to downsize. Those with higher levels of disposable income might be looking for sportier cars or luxury vehicles.

In all cases, you know when the customer will act—at the end of the lease period. If you are selling rather than leasing, however, you might look to industry studies to help you anticipate when each customer category is most likely to be in the market for a new vehicle.

This information in hand, you can now craft marketing campaigns with appropriate messages, offers and incentives based on what you know about each customer. Because you also know which marketing channels each segment prefers, you can “personalize” your contacts based on these preferences, sending personalized SMS text messages to young singles, for example, or a high quality, variable printed brochures to retirees.

Your customer base might look different than the one described here, of course, but you can use this process against your own customer profiles. To boost the effectiveness of your campaigns, just remember the letters “d” and “p”: describe, then predict.

More Insights To Enjoy:

+ Psychographics Can Help Your Marketing Communications and Improve Your Direct Mail

+ Finding the Right Data is Easier Than You Think!

© Action Graphics, 2011. Article taken from Action Insights, Vol. 2, Issue 7. Sign up today to receive future issues of our newsletters.


Like what you see? Stay in touch!