Segmentation's Elusive Payoff

By Steve Klinkerman

Experts have labored mightily to define key customer groups, but institutions must work harder still to capitalize on those insights.

People are different. It's the simplest of concepts, but one that large financial services institutions have found exceedingly difficult to capitalize on. Strategists long have been intrigued by the possibilities of using the social sciences and information technology to optimize dealings with key customer groups. But even as frameworks grow in sophistication, it remains to be seen whether customer segmentation will live up to its promise.

While it's tempting to blame flawed models, the problem increasingly seems to be one of implementation. What exactly is an institution supposed to do in response to analyses showing that, say, convenience seekers are far more likely to use online financial services and make Web purchases than complacent traditional consumers commanding nearly four times the resources? Without an action plan, unfortunately, such findings remain sterile.

That turns the pressure back on managers to think clearly about the steps that their institutions must take to capitalize on insights about major customer segments. Following through can be a monumental task, and executives who are either unwilling or unable to do so will never get the full benefit from even the most elegant segmentation framework.

The implementation challenge begins with identifying where each customer fits within the segmentation schematic. How do representatives make that determination quickly and accurately? Real-time database analysis? Live conversations? Web surveys? Paper-based self-diagnostics? The difficulty of placing customers in an actionable segmentation bucket is illustrated by the fact that even in a laboratory setting, Mainspring Inc. found that 39% of a survey group it had created did not cleanly fit into any of the sets identified during a fall 2000 online segmentation project.

The larger question is how to meld segmentation insights into the full spectrum of customer interactions. Segmentation is ultimately intended as a tool for quickly and effectively uncovering customer opportunities. Insights are needed on needs, preferences, perceptions and propensities. What's the channel behavior? What organizational capabilities most profitably align with segment needs? Should initiatives take the form of active outreaches, or be embedded in recurring customer interactions, or perhaps reflected in dynamic pricing?

Then there are the critical feedback and control mechanisms. Segmentation models and accompanying initiatives should be constantly updated and refined based on feedback from customers, representatives and managers. Financial metrics are needed to assess the productivity of various initiatives. Logistical considerations should be identified and monitored.

The fact that numerous institutions remain willing to broach this challenge is testament to the great promise held out by segmentation. Despite some disappointing experiments, it still is seen as a major key in aligning capabilities with the market. The potential payoffs include better-defined and more productive organizations, increased customer satisfaction and a growing share of the customer's business.

The concepts themselves also seem to be improving. In the space of a few years, we've seen the genealogy extend from profitability segmentation to needs-based segmentation; to demographic segmentation; to communities of interest segmentation; to attitudinal segmentation; to behavioral segmentation.

But great care must be exercised in bringing everything into a framework that is both actionable and manageable. In segmentation, conceptualization is not the end of the exercise, but rather the beginning.

Mr. Klinkerman is editor-in-chief of Banking Strategies.

Copyright © 2003 by Banking Strategies, published by BAI.

back to top