Driving Direct Marketing's Next Generation

Direct marketers today can be forgiven for feeling harried now that they have to scramble to prop up sagging response rates in traditional vehicles (which recently prompted a client to remark, "The harder I bail, the faster I sink!"). To keep up direct marketers must deploy ever more sophisticated techniques while managing a rapidly proliferating array of channels, products, and offers. The task of shifting from traditional arrangements of single product, single offer, and a few channels to multiproduct, multioffers, and multichannel is daunting. The good news? The opportunity for strong returns continues to grow. The bad news? The opportunity to destroy value by competing against oneself is growing much more quickly, and in some cases exceeds management capacity.

Take as an example a clothing retailer that recently introduced an off-label line. Initial tests might reveal a segment of young, well-heeled customers who are extremely responsive to email offers for the new line, but the retailer may not realize that these same customers previously dominated the retail sales of the marquee line. Without careful management, the "successful" introduction of the new line into the new channel could cost the retailer a small fortune by eroding its higher-margin flagship business. Because these two lines of business actually conflict with one another, the company's strategy must transcend its business units to maximize overall enterprise profitability.

Direct marketers must make fundamental changes to navigate this new multiproduct, multisegment, multichannel world. They must transition from finding the right target and treatment for each product and channel--a task dominated by champion/challenger thinking and traditional response modeling--to solving the larger problem of right product, offer, and channel for each person, at each customer contact.

The framework for solving this problem has existed since 1947, when the field of linear optimization was established. It allows companies to mathematically determine the correct sequence of marketing contacts for each individual. Even better, enterprises can solve the problem while accounting for constraints like budgets, sales targets, and margin requirements. Until recently, however, practical barriers--computing power, modeling requirements, and, most important, technical know-how--have forced marketers to rely on rules of thumb and manual processes.

This is beginning to change, as computing power and software design catch up with linear optimization, and marketing practices are now poised to catch up to marketing theory. Specialist software providers can put the power of optimization into the hands of marketing laymen. But the software tool kit (homegrown or vendor provided) is only one part of the answer.

The real transformation for companies lies in an optimization approach that informs overall strategy, aligns marketers' incentives, and provides guidance on trade-offs between competing product lines. In one recent example a multiproduct marketer increased profitability in its call centers by 15 percent when it optimized the sequence of offers between three competing product lines with different margins. Sales agents applied it in the call centers, and this sequence minimized cannibalization of the most profitable product, while allowing the less profitable products to grow market share.

When combined with models of campaign profitability and responsiveness, this approach enables savvy marketers to design marketing strategies that cross product lines, segments, and channels to maximize profits.

Michael Lamb is an engagement manager and Humam Sakhnini is an associate principal at McKinsey & Co.