I will be the first to admit that our industry, predictive analytics to support retail marketing strategies, is dense with technical words, phrases and methodologies all wrapped up into various umbrella groups or plans such as “Market Segmentation Analysis”, “Customer Analysis”, “Predictive Marketing”, “Retail Pricing Strategy” etc. Although each sounds a bit different all of these approaches strive to reach the same outcome:

• Use available data to describe customers and understand their needs, wants and actions

and each takes a different approach but all the approaches are based on well established principles of computer science and statistics.

As much as I would like it to be, the truth is the algorithms and data analysis procedures that are used in market segmentation have been around for decades. In fact Bayesian approaches (A common algorithm used to analyze data) have been used since the 1950s. The military used the same statistical principles to create the first radar systems, code-breaking methodologies in WWI, and plan various logistic strategies in WWII. There is no magic and no basement-scientist voodoo in our Market Segmentation Research or our buying predictions; just mature, proven calculations sprinkled with statistical know-how. So if Market Segmentation is more historical non-fiction then what makes it exciting? Innovation.

Although the fundamentals have been around for decades, modern day computers and storage have enabled us and others to apply those proven ideas to large sets of data. Today’s computers allow for innovation to take place within the domain of Market Segmentation Research in the form of applicability. In fact the Association of Computing Machinery has published several recent papers documenting how traditional segmentation analysis is applicable to today’s retailer (see references below). At Istobe our contribution to innovation is demonstrated by our ability to quickly convert data to actionable predictions to help define:

• retail marketing strategies,
• retail pricing strategies

and optimize and create:

• direct email marketing campaigns
• targeted email campaigns
• email promotion for internet marketing

So Market Segmentation is an old academic subject but the science fiction is in how we apply it.

ACM Market Segmentation Research Examples:

Research on customer segmentation model by clustering

In the paper, we use credit card consumption data as our model-building samples and present a modeling framework for building segment-level predictive models that utilize pattern-based clustering approach and signature discovery techniques. We devise monetary matrix and fluctuate-rate matrix to study various modes. Through clustering on both matrixes, we uncover different customer characteristics. Utilizing these characteristics, we can build two-dimension Consumption-Based customer segmentation model.

Intelligent value-based customer segmentation method for campaign management: A case study of automobile retailer

Most marketers have difficulty in identifying the right customers to engage in successful campaigns. So far, customer segmentation is a popular method that is used for selecting appropriate customers for a launch campaign. Unfortunately, the link between customer segmentation and marketing campaign is missing. Another problem is that database marketers generally use different models to conduct customer segmentation and customer targeting. This study presents a novel approach that combines customer targeting and customer segmentation for campaign strategies. This investigation identifies customer behavior using a recency, frequency and monetary (RFM) model and then uses a customer life time value (LTV) model to evaluate proposed segmented customers. Additionally, this work proposes using generic algorithm (GA) to select more appropriate customers for each campaign strategy. To demonstrate the efficiency of the proposed method, this work performs an empirical study of a Nissan automobile retailer to segment over 4000 customers. The experimental results demonstrate that the proposed method can more effectively target valuable customers than random selection.