CRM - February 2008 - (Page 40)

CUSTOMER SEGMENTATION As the volume of customer data increases, intuition isn’t going to cut it, and neither is mere manpower. requires two simple, straightforward questions: • How long will it take to have this tool in production for all users? • How long will a necessary change take? “If the answer to [either of] those questions is measured in months or years, that’s not acceptable,” Deighton says. He believes customers should settle for nothing longer than days, minutes—or even seconds. But start-up speed isn’t everything: Any segmentation tool also needs to be as flexible as a marketing department’s campaigns. The biggest mistake is thinking that one system can be applied to another (e.g., expecting a segment tailored for an email blast to be useful for a magazine-advertising campaign).“You’ve got to think ahead of time,” Hren says. “Make sure you have all the hooks necessary so that you can actually make those bridges [later on].” So how do you decide what you need before you need it? Map out your business processes to the best of your ability, think about how you’re going to use your segments, and then take it slowly. “It’s going to break [companies] if they go in one step,” Rheault says,“So we say, ‘Why don’t we design the first step? Let’s categorize your customers into just two groups.’” Set it up so that you leave room to build additional levels of granularity: Don’t create a closed system that has no flexibility, Rheault warns. Companies eventually realize they can’t stay in this phase any more than they can continue to mass market. Hren’s dose of reality is bracing: “Sometimes ‘simple’ works out well because the cost is low to maintain, but sometimes it’s a really competitive marketplace—and ‘simple’ loses.” THE INTERMEDIATE: CUT CAREFULLY You’ve identified the needs of your business, established the basic levels of segmentation, battled the software-selection process—now what? “Sales and marketing has changed,” Pombriant declares: Further segmentation is no longer a luxury, it’s an imperative. The key to being more sophisticated is determining whether or not new segments can justify the effort required to identify them. There is such a thing as too much. The goal is to have the greatest amount of difference between groups and high similarities within them. To avoid unnecessary complexity, frequent test-and-learn implementations help to reveal whether your efforts are being compromised by the law of diminishing returns: Does creating this other segment have, as Hren puts it, an “incremental impact”? Here are some questions to help gauge your segmenting situation: • Are response rates going down? • Are some segments doing better than others? • Is the competition segmenting? If so, how? • What can I learn from those efforts? • Am I doing better, and how can I stay ahead? • Is it more efficient to keep these segments apart or, even though they look different, will they both benefit from the same treatment? Another crucial component to effective testing? A good baseline comparison. It’s the difference between decision by intuition and decision by data. Consider the anecdote shared by author Tom Davenport at an SPSS conference last October: Gary Loveman, chief executive officer of casino operator Harrah’s, reportedly fires his employees for any of the following violations: 1) stealing; 2) harassing women; or 3) failing to use a control group. Katrina Lane, vice president of channel marketing at Harrah’s, likely passes the test. Every marketing program, she says, is measured by its impact on both profit and the customer. Similarly, Thomas & King’s Clark says a new value menu recently launched in several test restaurants will sink or swim based on analytics done by QlikTech’s business intelligence product, gauging success in each of several customer segments. By constantly evaluating the return from each segment, you can systematically and confidently execute your marketing strategy. In addition to regular testing, there are more overt corporate changes that would signal the immediate need for segment evaluation or modification: new products or price points; a new acquisition or new competitor; or an expanded distribution network. In the end, though, how many segments can you realistically support? Although segmenting enables you to personalize a message to a greater number of people, keep in mind that abandoning some might be just as important. “Each customer has a cost associated with it,” Pombriant says. “If you can identify someone who is not a buyer and select them out, you’re going to be 40 CUSTOMER RELATIONSHIP MANAGEMENT | FEBRUARY 2008 www.destinationCRM.com http://www.destinationCRM.com

Table of Contents Feed for the Digital Edition of CRM - February 2008

CRM - February 2008 Contents Front Office Reality Check Customer Centricity The Tipping Point The Loyalty Riddle CRM Drives Down-Market Out of the Gate: Marketers Rate ’08 Traits The Pulse Consultants Adapt to CRM’s Changing Landscape Required Reading Cover Story: CRM Gets Serious Contact Center Solutions Always On Rumble in the Office The Smallest Slice Tying Up Cable’s Loose Ends Burning Up the Paper Trail Sunny Skies for Knology No More Bumps for BlueRoads Secret of My Success Re:Tooling Scouting Report Pint of View

CRM - February 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.