I just returned from attending several retail energy and consumer finance conferences where modeling, profiling and segmentation was all the talk of the town. But what makes one data analytics approach different from the other, and what works best for different industries? To help answer this question, I’d like to touch on a previous article from the SourceLink blog that explained these disciplines, and shed some new light on some of the different ways to utilize data analytics.

Have you ever received a piece of mail that was personalized – had your name on it - but just didn't speak to you? Have you ever had an ad pop-up on your mobile device that would have been better suited for your 10-year old daughter? Or, have you ever gone to your favorite website, which you visit daily, and the banner ads speak to a store that isn’t even located in your state? 

These are all examples of how profiling, segmentation and modeling probably haven’t been done to reach you with the most relevant messaging possible. So how are these analytical processes defined and typically used?

1. Profiling
Profiling leverages a wealth of data to determine what elements are statistically different between segments.  In simple terms, it defines what your target audience looks like in terms of demographic, lifestyle, financial attributes and other characteristics.  Profiling is a very good analytical baseline from which to gain customer insights and develop marketing strategies around communications content- right message to the right audience.  In determining your best customer, this methodology builds a picture of your ideal target audience, and can be used to determine pre-selection criteria for other, more complex, types of segmentation. Profiling can provide the building blocks to open the doors to modeling and segmentation strategies that will allow you to harness a multichannel marketing approach with relevant messaging for your acquisition and retention programs.

2. Segmentation
Segmentation is the art of dividing a consumer universe into like, or similar, clusters.  For the purpose of this discussion, segmentation relates to similar clusters of customers.  By tailoring the offering (creative/communication, product, channel, price) to different segments, you are able to better identify and meet the needs of a distinct group of customers. Segmentation built on the basis of the unique characteristics of your customers is extremely powerful for developing highly relevant marketing strategies and campaigns.  Building your unique customer segments is both a data analytics and creative exercise.  The result is an understanding of your customers that is highly aligned with the goals and strategies of your organization, all geared to drive profitability.

3. Modeling
Modeling is essentially predictive mathematics.  The process involves utilizing the name and address of customers and appending hundreds of variables - including demographics, lifestyle characteristics, socioeconomic variables and aggregate credit information. The modeling process allows you to leverage hundreds of variables to determine which variables are most predictive of your desired outcome. Modeling can also help you address key marketing priorities, such as identifying those customers most likely to churn or attrite. Modeling is best used to target predictive behavior.  This methodology is especially important in the industries I work with most – healthcare, retail energy, and financial services. 

All of these analytical processes work well with one another and are even dependent on one another.  Often, you cannot build a segmentation strategy without first developing a profile of your ideal customer. But all three are essential in retention and acquisition strategies, and to better understand the similarities and differences between your customers and prospects. Whether your aim is to use data analytics to find your best customers, or to reduce yearly customer churn, as just two examples, these customer intelligence tools are used to make your company more profitable, and to generate significant return on investment. 

The combination of profiling, segmentation and modeling are critical to best-in-class marketing.  Profiling is excellent for communicating with stakeholders to develop an overall corporate strategy.   Segmentation allows you to talk with customers in a way that is most meaningful to them; what to say; how to say it; what images, colors, and styles to use; and how to motivate them to action.  Modeling is best leveraged to determine who is most likely to take the action you desire – response, renewal, enrollment, etc.  Often, companies make the mistake of choosing one approach over the other when, in fact, all three work synergistically to produce the best marketing campaigns and the best results.