For more information:
Call: 1-800-447-3269
E-mail: info-ncs@pearson.com
Survey.PearsonAssessments.com
Data Mining
This article focuses on knowing the customer. Subjects addressed are utilizing the customer database, customer retention and cross-selling. Also included are techniques to develop customer profiles that fit in the categories of predictive modeling and descriptive statistics.
One of the most important assets any business has is the database of customer information it has collected. Unfortunately, this is also one of the most under-used assets in most companies.
A well structured and researched database of customer information can be a key strategic tool for identifying new customers, retaining customers, and identifying new opportunities within existing customers.
Prospecting
Every salesperson dreams of finding the perfect customer—that person whose needs are an exact match for the products he or she has to sell. A skilled salesperson will use their knowledge of their customers and the products they have purchased to develop an understanding of what to look for in a prospect. Unfortunately, this can be an erratic process fraught with bias and error, as well as a slow learning process for new salespeople. Fortunately, if you have a customer database you already know who the perfect customer is—even if you're not aware of it.
Through careful analysis you can determine exactly what types of customers have the highest levels of satisfaction, repurchase products with the greatest frequency, utilize the widest variety of your products, are the most likely to respond to a specific type of marketing campaign, and even pay their bills on time. Comparing this information to demographic information can tell your salespeople exactly what type of prospect is worth their greatest effort. This information can also be used to develop highly effective targeted marketing campaigns.
Customer Retention
One of the things we learn time and time again through customer surveys is that many companies don't pay enough of the right kind of attention to customers. Many customers are only contacted through general advertising or when they receive an invoice. A customer database can be used to track customer contact, to send new product and product upgrade information to customers, and to survey customers regarding satisfaction. Often our research will uncover customers who go to the competition because they did not realize their original supplier offered a particular product.
Cross-Selling
Our research consistently finds that a company's existing customer base is the most fertile ground for sales when launching a new product/service. When market research identifies the best prospect type for a product or service the customer database can provide a list of exact matches. These are people who have already demonstrated a willingness to purchase products/services from your business.
Data Mining Tools
There are many different ways to develop customer profiles. Most of these techniques can be divided into two types:
Predictive Modeling
Predictive modeling is used to determine the relationship between data and your desired outcomes. These tools can also be used to determine what types of data are the most meaningful and to determine the importance of each variable. The most common statistical tools used in this type of analysis include: stepwise multiple regression, logistic regression, discriminant analysis, and neural network modeling.
While entire books can and have been written about these tools, they all share the same basic idea. Essentially, there is something you are trying to predict (like response to a direct mail campaign or customer retention) and you have a group a variables that describe variations in your customers. The key to this analysis is to identify which variables best discriminate between customers and then use this information to develop a model that will predict a future behavior.
Descriptive Statistics
Descriptive statistics can be use to describe customers in a database based on the data available. This type of analysis assumes that all data are equally important and meaningful. It also assumes that each data element contributes meaningful information. In other words, it tells you exactly what is in your database and how it can be organized.
Some common types of descriptive tools are frequency distributions, cross tabulations, customer profiles, penetration analysis, factor analysis, cluster analysis and CHAID. While this type of analysis does not attempt to predict a future event like predictive modeling, it does describe past events very accurately. For example, what are the demographic characteristics of your most loyal customers, and how do these demographics compare to the general population.
Both descriptive statistics and predictive modeling are important tools to use when analyzing your data. These tools, when combined with a well constructed marketing plan, can be a powerful asset for any organization.
For more information:
Call: 1-800-447-3269
E-mail: info-ncs@pearson.com
Survey.PearsonAssessments.com
Copyright © 1995-2007, Pearson Education, Inc. or its affiliates. All rights reserved.This document may not be photocopied, reproduced, translated, or converted to any electronic or machine readable form in whole or in part without prior written approval. If portions of this document are quoted in scholarly research, credit must be attributed to Pearson Education, Inc.