Modeling, Segmentation, and Database Profiling: Who are your best prospects and what will they do?
Modeling allows you to predict a prospect’s defined action, such as responding to a lead generation package, purchasing for the first time, responding to a cross-sell offer, or defecting to a competitor.
Models are created using multiple variables, and they are often performed in conjunction with segmentation and profiling. We test many different modeling methods (linear regression, logistic regression, CHAID, CART, neural networks) and select the best-performing model for testing. The data comes from your organization, and may be modeled with or without appended data (age, gender, household income, etc.)
Once the model is developed, each record (individual or household) is scored and assigned a segment value.
Every model includes a profile of each segment in the database as well as strategic and tactical marketing recommendations.
Mapping: Where are your customers and prospects?
Mapping used in combination with modeling and database profiling can provide a powerful way to visualize the data relating to patrons, prospects, campaign results, etc.
Mapping is performed using your data, either with or without appended demographic information. We map from the census block to the national level as appropriate.
Appending Data: Do you have enough information in your files?
If required, we can append additional consumer or business data to your files. Consumer data includes direct response purchasing; demographics such as age, gender, and household income; lifestyle, and other variables appended at the household level. Business data includes revenue, number of employees, SIC, and other information.