Data Mining is the process of analysing and extracting meaningful patterns and previously unknown and actionable information from large databases. The results allow the marketer the ability to make sound marketing decisions in such a way as to improve the profitability of the business.

Data Mining helps the marketer to:
Realize higher Direct Mail returns from smaller segments of a database.
Remove randomness from direct mail campaigns.
Identify meaningful customer clusters.
Understand the profiles of established customers.
Plan all media more effectively.
Identify opportunity areas for additional sales expansion.
Develop meaningful, targeted communications.
Identify customers with highest profit potential per campaign.
Maximize up-sell or cross-sell opportunities.
Identify customers most likely to churn.
Maximize Return-On-Investment for all media activity.

There are two basic styles of Data Mining: Hypothesis Testing and Knowledge Discovery. Hypothesis Testing attempts to substantiate or disprove preconceived ideas, while Knowledge Discovery starts with the data and tries to get it to tell us something we didn't already know. These styles can be broken into four general categories:

Simple Queries
Stand Query Reporting (SQL)
On-Line Analytical Processing (OLAP)

Complex Analysis
Statistical Analysis
Knowledge Discovery

Stand Query Reporting is a simple system of asking questions from the database, like "How many people did...?"

On-Line Analytical Processing (OLAP) pre-groups the data so the relational data can be reported easily in one query. "How may of my customers shopped 1 time or 2 to 4 times, over the last year?"

Statistical Analysis uses historical data patterns to predict which customers are most likely to be your best customers over the next year.

Knowledge Discovery systems utilize artificial intelligence tools like neural networks and machine learning to create a model to answer the same questions as Statistical Analysis. The difference being that the Knowledge Discovery tools can find complex, non-linear relationships in the data used, whereas Statistical Analysis methods are typically limited to finding linear relationships.


a collection of Data Mining & Predictive Modeling solutions for Canadian marketers. The PinPoint Series provides a complete solution for all database mining, profiling & modeling problems based on four Data Mining modules:

Response Modeling
Customer Valuation
Customer Segmentation I Profiling
Cross-Selling

The following articles are designed to give the marketer an overview of the power and capability of Data Mining and Predictive Modeling to produce valuable information for the purpose of creating better and more profitable Direct Marketing Campaigns.