• by Tom Hespos , April 23, 2002

Jim Meskauskas made some great points in his last article about how marketers use data, and he reminded us that sometimes we lose track of what really matters about advertising when we bury our noses in data. However, the final paragraphs of his article plunge a meat skewer into the idea of one-to-one marketing, and I don't think it was justified.

One-to-one marketing doesn't entail custom tailoring of messages to "100 million different people in 100 million different ways." Rather, it has more to do with defining a set of variables – things that matter to your target audience – and tailoring messaging accordingly. This doesn't necessarily mean that 100 million people each get a personalized message from the advertiser. A far more realistic scenario would involve those 100 million people being separated into 25 or so DR buckets such that the marketing message is increasingly relevant to people in those groups. Okay, so this doesn't technically equate to "one-to-one," but it makes an effort to address specific interests, which is how folks expect to be treated by interactive media. And it's effective, which I know from experience.

Maybe one-to-one marketing isn't 100% feasible at this point in time. That's okay. What's important is that we use data to move closer to one-to-one marketing – that we take what our customers tell us and use it to give them more of what they want.

Moving closer to one-to-one makes sense only when people are interested in hearing one of your messages in the first place. A more cost-effective application of segmentation technology would be to apply it once prospects have indicated a minimum level of interest. In other words, one-to-one may not make sense for an introductory advertising message, but it becomes a lot more effective once a prospect tells an advertiser "I'm interested in your offering."

A campaign I worked on last year for a liquor advertiser used only four pieces of creative to acquire email addresses for a content offering, around which we wrapped the advertiser's brand. Only after prospects opted into the program did we apply techniques that could be considered part of a move toward one-to-one marketing. What we found was that the more we asked program participants about their interests and lifestyles, the better we were at giving them what they wanted, and thus the better the loyalty to the brand and the program.

Based on answers to surveys on preferred brand and consumption, we were able to separate our program participants into buckets that mirrored what you might see in a more traditional direct response program. We had our "competitive" bucket, "switchers" bucket, "loyalists" for various brands and many others. Looking at these buckets in a traditional DR way allowed us to tweak the content offering such that we preserved the loyalty of folks who were already loyal to our brand, while we made maximum impact with the prospects who were likely to switch from a competitive brand to ours.

During the course of my career, I have worked on quite a few campaigns that work toward a one-to-one model. While it's true that none of these ever generated different messages for everyone touched by our campaign, that really didn't matter. Just as in life, the true value is in the journey, not the destination. As we move toward a one-to-one model, we learn more about our audience and we learn how to make the next communication more valuable. And how do we learn? By gathering both observed and declared information from our prospects, analyzing it, and formulating assumptions based on it.

Sometimes data do mean a damn thing, Jim. And I don't think our industry needs to give up on one-to-one marketing. Maybe we'll never get there, but we'll learn plenty from the journey.