Data Modeling & Predictive Analysis

Data Modeling & predictive analysis can significantly improve ROI

At the beginning of each data modeling & analytical project it is important to be in contact with key stakeholders to delineate the overall project orientation, marketing objectives and priorities. This initial step also identifies key internal and external resources, points of contact and project milestones. Each project will start with a review of specific marketing objectives and associated priorities.

The next step will be to assess the data that either is available or could be made available for the project. An effective analytical result usually requires a large and diverse set of information. While it is impossible to 100% accurately model (and therefore predict) human purchase behavior, there are many different characteristics that can help account for differences in consumer behavior. In fact, the best analytical projects are based on B.A.D. (Behavior, Attitudes and Demographics) data. Some of the analytic processes that may apply to your project would include:

• Data Mining – Our take on Data Mining is that the primary goal is to provide actionable business intelligence by presenting information on significant trends, patterns and relationships in company operations as well as the marketplace. To be effective, the Data Mining process must be flexible to meet changing needs, and timely to support ongoing decision making.
• Predictive Statistical Data Modeling – ANCHOR does not use a “canned” approach to developing predictive models. Rather, the specific steps and methodologies employed are chosen according to modeling objectives, nature of the dependent variable(s), type and amount of data available and other client needs.
• Strategic Customer Segmentation – In Strategic Customer Segmentation, the goal is to uniquely identify sub-groups within a customer base that are relevant to the marketing process. These segments (or clusters) ideally contain individuals who are very similar to each other, but collectively very different from other customers.
• Promotion List Segmentation – In Promotion, or Customer List Segmentation, the goal is to identify names in a particular file that have the highest propensity to respond, convert to established customers, generate the highest long term value, etc. List segmentation is more of a tactical tool used to implement specific promotions that were developed around the company’s strategic goals.
• Customer Value – The most commonly used approach is to measure and model the Lifetime Value of each customer. Typically, this involves building three models – one to predict the expected life of each customer, a second to predict the expected revenues over these lifetimes (or at least the next 2/3 years), and a third to assign the costs associated with serving the customer over this period.
• Geospatial Analytics – Many marketing decisions are based on geographical information and when added to other analytics can dramatically enhance the effectiveness of a marketing campaign. There is a measurable correlation between behavioral characteristics and physical location. We can provide a number of geospatial, often called location analytics, solutions including both drive-time and mileage between two locations.