• United States



by Dan Vesset

Using the data warehouse to drive your CRM effort

Apr 04, 20015 mins
CSO and CISOData and Information Security

Growth in the electronic marketplace and the resultant increase in the availability of and access to customer data have been major drivers of the accelerated pace at which organizations are adopting technology-based solutions for customer relationship management (CRM).

Companies are using technology to enable the alignment of their operations, resources, and strategies to maximize the value customers can derive from their products and services. To date, the CRM applications market has focused on automating customer touchpoint operational processes.

However, customer-centric companies are increasingly embracing a view that goes beyond the functionality of such applications to include a full range of supporting component software. This approach fosters a more sophisticated view of CRM based on the creation of a closed-loop CRM system.

The business process steps encompassed in this view of CRM are as follows:

  1. Customer interaction, interacting with the customer through various customer touchpoints, including online and offline channels
  2. Customer data integration, transforming and integrating the data generated during such interactions into an integrated customer database
  3. Customer data analysis, accessing and analyzing customer data
  4. Customer interaction personalization, establishing business rules based on data analysis and delivering them to touchpoint systems as the basis for providing personalized interactions

The closed-loop CRM system created by traversing these steps is made possible by the underlying CRM-centric data warehouse. Increasingly, companies are realizing the integral part that data warehousing and analysis plays in the process of effective sales, marketing, customer service, and overall business performance management.

At Step 1 of this process, data is tracked at the point of customer interaction. Whether this occurs at point-of-sale terminals in brick-and-mortar companies or on the Web for e-commerce companies, the goal is to track and collect all relevant customer interaction data into a data warehouse or data mart that will subsequently serve as the platform for analyzing this data.

In Step 2, the data from various source transaction systems is integrated in order to create a unified view of the customer. (This step often consumes over 70% of the time and budget of any data warehousing project). The focus on CRM is driving the need to create an enterprise wide, single view of a customer regardless of the customer’s point of interaction with the company. This single view of the customer is becoming a critical success factor, yet today it is more of a future vision than a reality for most companies as they continue working toward laying the necessary foundations of the customer-centric organization. In this step software is used that matches customer records across various internal systems as well as against external third-party demographic data.

Step 3 of the closed-loop process involves the analysis of the data collected in the data warehouse. Here various technologies, including query and reporting, multi-dimensional analysis, and data mining are employed to gain insight into the customer data. The final step in this process involves personalized interaction with the consumer that is based on the rigorous data analysis and empirical evidence rather than decisions that have too often been made on “gut feeling” and resulted in wasteful marketing campaigns and substandard customer service.

Nowhere is this process more evident than in e-commerce, where the customer touchpoint is the Web. Having installed e-commerce software, companies are in a position to harness the vast amount of data that is generated daily through both customer and visitor activity on the Web site. This data, often referred to as click-stream data, is collected from Web server logs and contains a potential wealth of enterprise data that has previously not existed.

Extracting, transforming, and loading this Web data into a data warehouse is no small feat, however it is a precursor for understanding visitor behavior and marketing effectiveness. Where did my Web site visitors come from? which paths did they take on the Web site? which advertisements did they look at? when did they abandon the Web site? which products did they put into their shopping baskets? are some of the many questions that can be answered by Web site analysis applications. In turn, the results of the analysis drive personalized marketing campaigns, which may include personalized emails, dynamically generated Web content, as well as offline marketing communications.

Linking the Web data with other enterprise data creates even more powerful solutions. Imagine a scenario where an unhappy customer calls the company’s customer service center and complains about an experience on the company’s Web site. Having access to history of that customer’s experience on the Web site through click-stream path analysis would empower the service representative to take appropriate action to appease the customer and potentially salvage its relationship with the customer.

Thus, CRM analytics are a group of applications based on an underlying data warehouse with functionality that spans segmentation, cross selling, customer retention and churn, marketing campaign optimization, and, most important, customer-profitability analysis.

The reality today is that the best examples of closed-loop CRM focus on optimizing and personalizing a specific touchpoint. Over time, companies will seek to leverage the analytic infrastructure created (especially the integrated customer database with “a single view of the customer”) to improve other aspects of customer-facing operations. To date, the greatest challenge in implementing the closed-loop CRM system has been the integration of data from a multitude of source systems in to a data warehouse.

Although packaged software tools exist to simplify this process, it often remains a time-consuming effort. On the other hand, the data analysis process can often be addressed through packaged applications and business intelligence tools.