Use Data Warehouses for Better CRM
Many bricks-and-mortar companies have expanded by transforming into clicks-and-mortar companies, providing customers with Internet access to their products and services. This transformation enables customers to interact with the company via the Internet, Web-enabled kiosks, ATMs and other touch points. Reaction to this trend from an operational perspective has been positive because of the reduced costs in customer service resulting from the self-service model. However, the extra channels have made maintaining one-to-one customer relationships difficult.
With the transition to clicks-and-mortar companies, Web sites became the first touch point where implementing CRM proved to be a challenge, often because they evolved rather than being planned. Typically the information technology department developed a company's first Web site. The Web sites were efficient, easy to use and provided the information needed from an IT perspective.
But as the Internet gained importance as a channel, marketers discovered they had little knowledge of the behavior of their Internet consumers. Marketers could see what was bought, but could not see who had searched through the site information or who thought about buying items but abandoned them before purchasing. These details are not trivial: According to a study by BizRate.com, Los Angeles, 75 percent of online shoppers abandon their carts before completing the transaction. If marketers are to improve this state of affairs, they need a way to develop a better understanding of their customers.
The obvious solution would be to plug the existing CRM system into the Web site database. However, the databases commonly used at the back end do not integrate well with CRM systems. These back-end databases are optimized for online transaction processing -- lots of simple operations. They are adequate for answering price and availability queries, but are slow to process the complex ad hoc queries that CRM systems can generate. Speed is essential for CRM systems, because without it businesses miss the opportunity to analyze clickstream -- data representing the navigation path a user followed on a Web site -- and produce a personal, one-to-one relationship.
The solution is to add a central data warehouse where relevant data from the OLTP database are replicated and optimized for CRM queries. Adding a suitable data warehouse is relatively straightforward; while this might initially seem costly, customers report payback times of less than a year.
Using a data warehouse as a central repository allows marketers to analyze consumer behavior through their interaction with the Web site. It collects huge amounts of clickstream data, which helps a company understand how customers access its Web site and what they do when they get there.
Using this data in near real time allows companies to personalize their communications to customers. A clickstream will show marketers what was in a customer's shopping cart before he abandoned it as well as when he abandoned the transaction. This information gives marketers the opportunity to rapidly follow up with the consumer to lure him back to the site to complete the original transaction and possibly evolve into a profitable customer.
For example, a Web travel site might discover from applying CRM analysis to clickstream data that customers who do not buy within two days of surfing the site are unlikely ever to purchase. The company can use its CRM system to watch for these people and take a proactive approach by sending coupons or other incentives directly to them. Bringing customer-centric data to marketers' desks enables marketers to immediately understand customer behavior, draw implications and take proactive measurements to enhance relationships.
Attempting to install a CRM data warehouse often ruffles feathers of those in an IT department, many of whom resent this intrusion into "their" territory. Marketers should anticipate this reaction and be ready to justify their suggestions on both a cost-benefit and technical basis. It is important to get IT department "buy in" for the project to succeed.
As long as they understand that there are two sets of technical requirements and, hence, a need for two databases, it should be possible to reconcile the IT and marketing objectives.
Another obstacle to implementing CRM programs is that companies must consider their customers' privacy. Privacy is the subject of public concern worldwide. It has always been a background issue for database administration, especially in relation to database marketing applications. Recently privacy has sprung to the forefront as a major focus of government, media and consumer attention. With the recent privacy scandals on the Internet, the government has intervened and now appears to be in the process of implementing privacy regulations to ensure consumer protection -- legislation to protect children has gone into law and more of these restrictions will probably come into effect in the future.
Because many regulations are pending, companies should anticipate future upcomings and build a database that can easily be altered to meet further regulations.
In addition, those companies that do not take account of privacy will possibly lose customers to competitors that do practice privacy protection. Contrary to popular belief, CRM and respect for consumers' privacy are not mutually exclusive; it is possible to perform many CRM operations on "anonymized" data where customer information from which all unique identifiers (name, social security number, etc.) have been removed.
In relation to CRM, privacy means individual control and protection over the use of personal information. Consumers should have the opportunity to opt in and opt out of fields to give them control of their personal information. Customers might actually like receiving direct mail, if they get mail that relates to them. The negative perception of direct mail is an obstacle, but only because most companies have abused the practice with spam in the past.
Thus, another reason why one central CRM data warehouse is a compelling solution for integrating CRM across all channels is the simplicity of implementing privacy protection. In the past, most businesses functioned off multiple databases and each database had to have a privacy filter to protect consumers' information. The problem with multiple databases and filters is that every time there was an alteration in the system, or in company policy, it had to be applied to each database. This is a time-consuming process.
By eliminating multiple databases, one nontechnical individual can be given the responsibility of maintaining and updating privacy policies companywide.
It is not enough to simply understand customer behavior; rather, businesses must be able to quickly turn that understanding into actions that will improve business profitability.
CRM solutions help companies solve problems such as customer acquisition, profitability and retention, and help build enduring relationships with customers by understanding and interpreting their behavior across all channels. Adding a data warehouse solves problems by integrating CRM across all touch points, and improves transactions and interaction by providing fast, safe and economical dialogue with consumers.
• Peter Heffring is president of the CRM solutions division of NCR Corp., Dayton, OH, which provides relationship technology solutions to customers worldwide. Reach him at Peter.firstname.lastname@example.org.