Statistical Analyses


The role of database is to help select names for modeling, implement the results of the modeling process by scoring names and assigning them to the appropriate decile, and selecting names by decile and other criteria for marketing programs. Most companies use statistical analysis for two principal reasons: a) segmentation, and b) predictive modeling.

Segmentation techniques are used to identify and profile groups of customers whose characteristics are similar. If the objective is to segment customers based on their performance, then the procedure is to group people according to their performance characteristics and then develop profiles of each performance group. Typical segmentation variables are performance measures such as recency, frequency, and monetary value of purchases; types of products purchased; or types of promotions responded to.

By linking this data with customer performance data, marketers can analyze who buys what and use the profiles of customers in each segment as a means of finding other customers like them.

Once the segments have been created, individual customers will be assigned to segments and these assignments will be recorded in the database. This makes subsequent selection of individuals for promotion based on the segmentation criteria relatively simple.

Predictive Modeling, based on previous purchase history, based on recency, frequency, and monetary value, models can be developed to predict who is most likely and least likely to purchase at the next opportunity. This scoring model would be used to determine who should be promoted and what they should be promoted with.

Once scoring models have been executed and customers assigned to deciles, this information is recorded in the database so that subsequent selection of customers who have the highest probability of responding to a promotion is easily accomplished.

End users would use a selection menu in which they would indicate which scoring model they wish to use and either a specific cutoff score or a desired number of names to select. The database would then perform the selection and produce an output file to the specific medium. This would either be a file, a magnetic tape, or mailing labels. A file could either be used for further analysis, or in many cases, the file could be combined with a patterned letter file to produce personalized mailings.

My Consultancy–Asif J. Mir – Management Consultant–transforms organizations where people have the freedom to be creative, a place that brings out the best in everybody–an open, fair place where people have a sense that what they do matters. For details please visit www.asifjmir.com, and my Lectures.

Customer Value Checklist


  1. What are your current and targetted CR rates?
  2. Given your current defection rate, how often must you replenish your customer pool?
  3. Has your CR rate increased during the past 3 years?
  4. What is the lifetime value (LTV) of a customer?
  5. What is the cost of a lost customer?
  6. What percentage of your marketing budget is spent on customer-retention activities?
  7. On average, how much do you spend on current customers annually?
  8. What criteria does your company use for developing targetted retention programs by market segment?
  9. Do you invest more on high-value (A) customers?
  10. How does your firm use recency, frequency, and monetary value (RFM) analysis?

My Consultancy–Asif J. Mir – Management Consultant–transforms organizations where people have the freedom to be creative, a place that brings out the best in everybody–an open, fair place where people have a sense that what they do matters. For details please contact www.asifjmir.com, Line of Sight

Usage Analysis and Customer Retention


Segmenting markets by consumption patterns can be quite insightful for understanding your customer mix. Differentiated marketing strategies are needed for the various user groups—first-time users, repeat customers, heavy users, and former users. By classifying customer accounts based on usage frequency and variety, companies can develop effective strategies to retain and upgrade customers. There are many highly informative, low-cost applications of usage analysis that should be considered by management.

By classifying customers into usage categories, management can design appropriate strategies for each market segment. The objective is to move customers up the ladder, where possible. The implication of usage analysis is that all customers are not equal; some (the heavy users) are clearly more important than other categories.

The Pareto principle, or 80/20 rule, is insightful in the context. In a typical business, approximately 80% of sales comes from about 20% of customers (also, note that generally about 80% of your sales comes from 20% of your goods or services). It is important to defend this core business, as heavy users are primary attraction targets to key competitors. These highly valued customers require frequent advertising, promotions, and sales calls and ongoing communication efforts.

By knowing who better customers are—through geographic, demographic, psychographic, and benefit research—we have a solid profile of “typical users.” This information is very helpful in playing subsequent customer attraction/conquest marketing efforts. Realize that the marketing information system, the database, plays a key role in customer analysis and decision making.

For unprofitable customers, the company often needs to find new ways to serve them more effectively. Technology such as ATM machines, ICT, can be used in this regard. Quarterly contact through a newsletter and direct mail or access options such as toll-free telephone numbers and websites maintain adequate communication with low-volume users. In some cases, it may even be desirable to sever the relationship with certain unprofitable customers.

A good understanding of our customers’ purchasing patterns helps us keep our customers and gain a larger share of their business. Share of customer (customer retention focus) has supplanted market share (customer attraction focus) as a relevant business performance dimension in many markets. Share of customer is adapted by industry and goes by such names as share of care (health care), share of stomach (fast food), and share of wallet (financial services). If a company can increase a customer’s share of business from 20 to 30 percent, this will have a dramatic impact on market share and profitability.

Recency, frequency, and monetary value (RFM) analysis is a helpful tool in evaluation customer usage and loyalty patterns. Recency refers to the last service encounter/transaction, frequency assesses how often these customer-company experiences occur, and monetary value probes the amount that is spent, invested, or committed by customers for the firm’s products and services.

A more effective strategy is to classify customers via usage analysis and design differentiated marketing approaches for each target market. In sum, usage analysis can greatly assist us in our customer retention activities. Think about how to “hold” heavy users and key accounts, upgrade light and medium users, build customer loyalty, understand buying motives to meet or exceed expectations, use appropriate selling strategies for each targeted usage group, win back “lost” customers, and learn why nonusers are not responding to your value proposition.

My Consultancy–Asif J. Mir – Management Consultant–transforms organizations where people have the freedom to be creative, a place that brings out the best in everybody–an open, fair place where people have a sense that what they do matters. For details please contact www.asifjmir.com