FIGURE 1. CUSTOMER INTELLIGENCE ENVIRONMENT

Customer Data
Starting at the top of Figure 1, Customers and Consumers alike provide information in the form of behaviors and attitudes. Consumer behaviors may be captured internally by sales patterns, channel usage, and campaign responses. Consumer behavior may also be collected externally through syndicated research, behavior assessment such as Nielsen, or attitudinal/lifestyle profiles such as Acxiom or Experian. In addition, consumer attitudes may be captured through either qualitative or quantitative Market Research. Our model for customer intelligence suggests that these sources of information are captured and either loaded into or made accessible by the Analytic Data Repository.

Data Collection
Customer behaviors are directly collected through the major touch-points of the organization. These touch-points include call centers, point-of-sale systems, Web sites and other operational systems managed by the organization. Customer attitudes are being collected through commissioned Market Research studies as well as corporate web surveys, customer panels and emerging technologies for text analysis and customer voice analysis.

Data Storage
Whether from customers, consumers or both, there are a growing number of data sources available that provide organizations with a myriad of behavioral and attitudinal information. In order to derive insights from the data, the data must be combined, managed and centrally accessible.

Monitor
Monitoring is the process of identifying key indicators of business performance at various levels across the organization. These key performance indicators (KPIs) are typically accessed through executive dashboards. Critical KPIs may also be monitored by alerting agents that can send emails or calls when a defined threshold is crossed. Whether by human or machine, KPIs often identify areas of threat or opportunity.

Report
Upon identifying a potential threat or opportunity, enterprise reports are typically available to quickly determine the impact of the trend on business performance. Reports are useful for rapidly accessing business information. However, they are not well suited for exploration due to their static nature.

Explore
Given that the threat or opportunity has been shown to be relevant and substantial, exploration can begin in order to identify possible drivers of the trend. On-line Analytical Processing (OLAP) technology is a valuable tool for examining issues from several dimensions. With OLAP one can narrow the problem or focus the opportunity down to a manageable space. For example, if treadmill sales are on the decline, OLAP can help identify which regions and customer segments are most accountable for the trend. This exploration of the data can be classified as 'data mining' using the broadest definition of the term. However, manually finding important patterns in OLAP ‘universes’ may be like finding a needle in a haystack as the number of business dimensions grows. In such situations, automated Data Mining techniques may be employed to find hidden patterns.

Exploration often leads to the formation of new hypotheses. For example, one might observe that when women buy treadmills they also buy “ab crunchers.” Yet, when men buy treadmills they also buy heart monitors. These observations may lead one to conclude that women buy treadmills to “tone up” while men buy treadmills for better health.

Here is where customer intelligence typically breaks down. The observations in the previous example merely suggest a cause. Many CRM efforts fail because decisions are made based upon one discipline without consideration for the other as described below.

Research
The origin of customer intelligence is Consumer research. Many hypotheses are generated daily within an active customer intelligence environment. These must be properly tested, especially those with strategic implications or costly tactical programs.

Back to our treadmill example, while exploration may have suggested that women were focusing on body image and men were focusing on health, the reverse may actually the case. Research might reveal that women are including this equipment as part of a low-impact program designed to fight osteoporosis and promote healthy aging. While the men intend to use the treadmill to their physical limit to burn off the excess carbohydrates they are consuming as part of their muscle building program.

Consumer research is commonly executed with either of two disciplines - Data Mining or Market Research. Both disciplines provide scientific rigor and allow one to draw conclusions within acceptable bands of confidence.

Deploy
These conclusions are the new findings that expand one’s customer intelligence. They provide the confidence to plan and execute new programs to avoid the threats or capitalize on the opportunities at hand. Done properly, these programs are tested and evaluated prior to being deployed broadly into the operations of the organization.