The Right Relationship

What if we got it all right? What if we’ve finally created the tools and procedures necessary to manage and develop a direct marketing business? I think we’re getting closer to this goal now, at least in regard to contact strategy, or if you prefer, customer relationship management.

Here’s what we can and can’t do.

– We can create customer segments using behavioral and demographic data based on samples, and assign people to these segments using some combination of regression, CHAID and discriminate analysis.

– We’ve learned that developing segments based on attitudinal data is a problem when it comes to assignment – it’s difficult and the errors are too large. But we’ve also found that we can research our behavioral/demographic segments for attitudinal differences, and if there are any, the findings can be used to modify our offers and creative presentations.

– We can share and integrate customer data that’s been gathered at various points, so that anyone who deals with customers can have a fairly accurate picture of these individuals’ relationship with our organization.

(The more you’re willing to spend, the more current this information can be. Whether direct marketers require access to real-time data is open to question: If you’re selling airline seats, the answer is obviously yes. But if it’s insurance, customer service may require near-real-time data, while marketing may not.)

– We can respond to incoming calls because we can estimate an individual’s incremental lifetime value, and respond accordingly should he or she choose to contact us for any reason.

– We can use product affinity models to suggest those people who should receive certain offers through direct mail, telemarketing, or when they come back to our Web sites.

(A word of warning here. One of the issues in the Verizon telecom strike this past August involved customer service representatives who were frustrated because they had to make obviously incorrect recommendations based on computer analysis of individual profiles. Marketers need to take a hard look at product affinity models before signing off on their use.)

– We can, in some cases, create customer loyalty and/or reward programs that improve retention and repeat purchasing and more than cover their costs. But we’ve also learned that reward programs can’t make up for a price that’s not competitive, or for unsatisfactory service.

– We can update our customer databases with information from our Web sites, just as we update our marketing databases with information from our physical fulfillment systems. But we really haven’t learned precisely which data is truly valuable and how much of it is simply interesting.

– We can manage our databases and campaign management procedures to handle an unlimited number of business rules regarding contact strategy. For example, don’t call the same person more than once a month on any product; don’t call that person more than twice in three months for the same product; don’t mail the same offer to the same person more than x times in 12 months, and so on.

– We can predict response and lifetime value by using combinations of product, offer and marketing channel and thus can predict, with some degree of accuracy, expected profits per name contacted.

– We can optimize our selection process to maximize response or sales or contribution within some defined promotion period, given a complex set of restraints and conditions. For example, minimum and maximum requirements for any product or product/channel combination; overall min and max constraints for any time period.

– We can automate, for lack of a better word, all our testing strategies, so that once our databases are updated, our marketing plans can be executed in a fraction of the time it took to implement promotions only a few years ago.

– We can trace the effectiveness of our strategies and evaluate those that work and those that don’t. (This is the never-ending task of the direct marketer – to find a better way by analyzing and testing alternatives.)

– We can spend more time on data analysis and strategy development and less time on the data manipulation and routine data processing that often passes for marketing analysis.

Ultimately, all of this should make us much smarter marketers.