With the hullabaloo surrounding one-to-one marketing, one could easily surmise that every sophisticated database marketer in the country has, by now, developed precisely versioned direct marketing programs aimed at audiences of one.
But practitioners have often encountered difficulties in translating an exciting concept into cost-effective implementation. Why has it been so hard to get from theory to practice?
In a major address to the Direct Marketing Association's Fall Conference in 1993, Lester Wunderman admonished the industry for practicing “mini-mass marketing,” whereby direct marketers used sophisticated tools to identify key consumer segments, but then sent every one of them the same message.
The one-to-one message is clear and powerful; the reason it's not widely practiced today does not stem from disagreement over its merits. The truth is that despite its elegant surface simplicity, one-to-one marketing can be enormously complex to implement. Sending 1 million different messages to 1 million different households is technically possible, but practically unrealistic for most marketers.
The case for market segmentation. How then can database marketers tap into the potential benefits of relevant, customized marketing communications without breaking the bank? One answer is the effective use of precisely targetable market segmentation. Today's marketing database contains an often-bewildering array of data elements. To optimize the direct experience with consumers, marketers often must juggle the transactional data on their own customer files with external database variables encompassing demographic and lifestyle data, reported purchasing behavior and geographic and temporal idiosyncrasies. By searching for homogenous groups of consumers who have similar characteristics and exhibit common behavior, marketers can begin to make practical use of their marketing data. These groups, known as market segments, can become the cornerstones for more effective marketing communications.
Market segmentation has been used in market research for quite some time. However, the broad acceptance of these systems in the direct marketing industry has been hampered by certain inherent limitations of most early systems. The initial VALS system had limited direct marketing success because it was impossible to reliably target individuals comprising the nine psychographic groups. Geodemographic systems, such as PRIZM and MicroVision, have encountered only moderate success in direct marketing applications because of the underlying premise of neighborhood homogeneity, especially in as diverse a society as the United States in 1999.
Custom segmentation systems have been developed for particular brands or products, but often are so brand or product specific that they fail to make the leap from one brand or product area to another. Also, systems based on a group of known product users require a link or bridge to the millions of nonusers in order to produce effective prospect targeting.
As data has become more plentiful and computers faster and more powerful, segmentation systems based wholly on the modeling of response data have emerged. These empirical models, which identify the deciles or duodeciles most likely to respond to an offer, certainly improve response rates and mailing efficiency, but also have limitations, in that they are based on the history of previous offers, provide no understanding of the response process, and thus do not assist in determining how to communicate most effectively with the most important segments.
Household-based market segmentation systems. To enable market segmentation to play a key role in successful database marketing efforts, marketers need a precise yet flexible segmentation tool. The emergence of household-based segmentation systems holds promise for database marketers. These tools enable marketers to speak to precisely targeted, homogenous groups of households. With such tools, the customer database becomes a multidimensional, strategic marketing tool capable of supporting a wide array of one-to-few marketing programs. Specifically, an effective household-based system should do the following:
* Use self-reported, household-level demographic and lifestyle data (in contrast to systems that rely on neighborhood-level, census data);
* Identify groups of consumer households that are homogeneous with regard to demographic and lifestyle data, as well as purchasing and media behavior;
* Transcend a wide range of consumer behavior, from product consumption to sports attendance;
* Humanize a marketing process that was in danger of being defined in terms of impersonal statistics (It was retailing giant Stanley Marcus who stated: “Consumers are statistics; customers are people.”);
* Provide marketers with strategic direction, as well as tactical applications;
* Link transactional, customer data to the outside world of external data;
* Identify a rich pool of known prospects within target groups;
* Allow marketers to understand the relevant behavior of their customers and prospects so as to optimize the communications and offers to target groups.
Customization of marketing communications makes them more relevant to particular consumer groups. If most companies have six to eight distinct groups of consumers that make up the majority of their customers, it makes sense to create six to eight distinct marketing messages. Greater relevance means better response and stronger customer loyalty.
Once a company begins to see its market in terms of its diverse segments, the applications become apparent, particularly if a company appreciates the technological capabilities available today for addressing those different segments. Sending relevant messages to homogenous groups of households represents the key to realizing the future benefits of one-to-one marketing today. Household-based segmentation supplies the critical missing element in effective one-to-few marketing programs.
Scott D. Schroeder is chief operating officer and Jock Bickert is CEO of Looking Glass Inc., Denver, a database marketing applications company.