Data Mining Can Help Retailers Realize the Promise of Enterprise CRM Systems

Organizations have spent vast amounts of money in pursuit of the perfect CRM system, yet the full promise of CRM is rarely realized. The original premise of CRM was threefold: To understand customer behavior to improve relationships; to share this information throughout the enterprise to allow customer-facing systems and staff to be more effective; and to reduce costs. Unfortunately, most organizations still cannot present a consistent and coordinated customer experience. Instead communications often seem to conflict across channels or at best appear to compete for customer attention, resulting in a perceived dysfunction that leaves customers perplexed.

Companies that want to get it right need to take a long, hard look at what it takes to establish a CRM system that fulfills its promise. One obvious notion is to turn the old segmentation approach on its ear and, instead, rely on individual user behavior to establish better targeted campaigns that mean something to customers. This can be done only by consolidating processes and technologies into a single, common system that can automatically segment a customer, then dynamically generate targeted content.

For example, Sephora, a leading worldwide retail beauty chain, achieved productivity gains of 70 percent for the creation and execution of campaigns by implementing database-driven marketing and automating the campaign management process.

Sephora was experimenting with a customer loyalty card program utilizing an outsourced service provider. Though the loyalty card program was valuable in helping to identify a select group of repeat in-store customers and their preferences, Sephora was rapidly coming up against the limits of using a marketing database managed by a third party. Seeking to gain more timely access to the specific buying habits and behavior of its customer base, and the ability to effectively identify links between online and in-store shoppers, Sephora took its database in-house and developed its own customer-centric marketing program to assist in driving sales across channels and building better, more personalized customer relationships.

Workable channel
With the company's new enterprise marketing system firmly in place, Sephora marketers were curious to find out if customers using loyalty cards were using both online and in-store locations to purchase products. More specifically, was the loyalty card really working? Was online really a viable sales channel? How valuable were individual customers and how much were they spending in store and online?

Using enterprise marketing software and data mining technology, Sephora began to drill down into a centralized marketing database and segment its customers via RFM: recency, frequency and monetary segmentation. With the ability to accurately calculate each customer's unique transactional history, Sephora could assign each customer a value—and target campaigns accordingly. The results were immediate. One example: Sephora initiated a direct-mail campaign offering private sales to its best customers twice each year, a tactic that reinforced customer loyalty and generated greater revenue for the company.

After defining RFM segments, Sephora developed a key performance indicator (KPI) for each segment, based upon previous purchase behavior, to adapt and personalize communication in the hopes of increasing campaign ROI. Analysis and campaign management functionality gave Sephora the ability to calculate buying-behavior indicators by segment according to the number of products bought during last 12 months, the number of different brands and the number of different stores where purchases were made. The ability to define customer loyalty by brand, market and channel preference made personalized marketing a real possibility—and gave Sephora the opportunity to cross-sell and up-sell products accordingly.

Sephora was also able to overcome a disconnect in the loyalty card program. Despite the program's initial success, over time, loyalty card owners appeared apathetic to the coupon program that offered a 10-percent rebate. A deep market analysis using enterprise marketing management technology showed that the problem was not the market saturation of 10 percent coupons as initially thought but, instead, the ability and interest to use them by the expiration date. Sephora subsequently tweaked the campaign according to four different categories of user behavior and came up with two new campaigns. In the first program, the 10 percent discount was eliminated in favor of a gift giveaway, which resulted in increased customer satisfaction, measured by customer loyalty. In the second program, shoppers had to collect points to achieve the 10 percent rebate, which boosted average spend.

Additionally, Sephora was able to determine the value of online store activity by comparing profiles of online and in-store customers and prove that, for a loyal store customer, the online purchase meant additional revenue and frequently resulted in increased in-store spending. With this knowledge, Sephora developed specific campaigns that catered to the online consumer.

Since deploying enterprise marketing management technology, Sephora has recognized estimated productivity gains of 70 percent for the creation and execution of campaigns by implementing database-driven marketing and automating the campaign management process. In fact, even though the direct marketing budget has remained the same, response rates have doubled.

Sephora is also now able to deliver campaign activity reports in as little as one day, down from five days. And customer response rates consolidated from all the touch-points (point of service, web and call center) can now be assessed in real time; the previous method took 10 days of work each month.

In this heated battle for the consumer dollar, retail marketers are looking to establish a personal and consistent relationship with their customers. Though it's often overwhelming to figure out where to start, you truly begin an effective marketing program with the creation of a centralized customer database. And an enterprise-wide marketing automation platform that pushes and pulls individual customer data from that central database can do wonders for establishing profitable, long-term customer relationships.

Patrick McHugh is the executive vice president of North American Operations for Neolane, Inc. He has been in sales, marketing and business development for technology solution companies since 1987 and is a proven thought leader in the software solutions market. He can be reached at patrick.mchugh@neolane.com.