The Goldilocks Rules of Data-Driven Targeting: Getting it Just Right

It has always been my assumption that the story of “Goldilocks and the Three Bears” was written to educate kids on the best practices in data-driven targeting.

You’ll recall that a young girl, Goldilocks, breaks into the domicile of a family of bears. She proceeds to sample their porridge, finding one bowl too hot, one too cold and another just right. She then sits in their respective chairs, finding one too big, one too small and the third just right. Later, she abandons the too-hard and too-soft beds for the one that’s just right.

There’s more to the story, and I think at some point there is even a moral. But none of that matters for our purposes. Goldilocks’ experiences—particularly with too hot/too cold and too big/too small—are meaningful when it comes to using information in marketing. Given the nuances of data-driven targeting, the pressure is on all of us marketers to get it just right.

Too cold or too hot— balancing consumer privacy and message intrusiveness with relevance and usefulness.

There is a wealth of internal and external information available to marketers today. But using it correctly to convey the right amount of personalization isn’t always obvious. Too cold or impersonal, and our intended audiences aren’t inclined to pay attention.

A company with whom I do a lot of business recently sent me a statement of my earnings in their loyalty program, so all of the content was already being customized to me. Well, almost all if it. The salutation greeted me with “Dear Valued Customer.” Way too cold! Other examples include using information that’s inaccurate. I’ve gotten a few emails lately from an organization for senior citizens. Although I hope to be a target for them someday, I’m not one yet.

The opposite approach can be even more dangerous. When we’re too hot we risk offending someone by being a little too close or intrusive. After all, just because data is available doesn’t mean its use is welcomed.

• When we move to a new house, we’re used to seeing “new mover” mailings arrive from nearby home improvement stores and other retailers. Clearly, those retailers know that we’re recent movers and they’ve used this relatively innocuous and widely available information for targeting. That’s great.

• Now try this: you’ve recently shopped for a new watch online. Suddenly, every other online ad displayed for you is about watches, and not just from the retailer you shopped. These ads are probably relevant to you, but their re-appearance makes clear that your preferences are being captured and used by someone with whom you may or may not have a business relationship. Whether this is too hot or just right is a subject of much debate right now.

• And then consider this telemarketing pitch from a company with whom you have no business relationship: “Now that your son has reached driving age, have you compared your current auto insurance premiums with ours?” Lots of companies make use of children’s age ranges to make decisions. But using the information to select an audience is very different from in-your-face messaging that reminds individuals what others know. Rely on expertise (and common sense) to know that some things may be too hot.

Too big or too small – when enough is enough.

In a recent session with a group of marketers, all of whom are heavy data users, one participant remarked, “You know, I’m convinced that I no longer need more data. What I need is greater insight from that data, and that’s what’s becoming a challenge.” Others agreed. Too much data—customer, contextual, credit, demographic, attitudinal, segmentation, lifestyle, transactional, channel behavior—can be overwhelming. As in so many things, 80/20 (and even 90/10) rules may apply. But how do you know which 10-20% of all the available data will yield 80-90% of the predictive insight?

Here’s a simple test I’ve used with clients to assess whether their data is too big or too small for their needs. First, fill in the blank in the sentence below:

• If I only knew _________________ about my customers or prospects, I could improve my marketing efforts.

Second, ask yourself this follow-up question:

• Why don’t I know ______________ about my customers and prospects?

If your answer to the second question points to an information gap, meaning that the reason is that you don’t have the information you need, then your data is “too small.” You need to fill the gap with new or different kinds of data than you’re currently using. If your answer to the second question points to a process or analytic gap, meaning that you don’t get the insights you need but the raw materials are present in your data warehouses, then your data may be “too big.” Investing in analytic and consulting resources to make the most of what you already have makes sense.

In the world of data-driven targeting, like in the world of Goldilocks, getting it just right isn’t easy. We’re confronted with judgment calls with respect to what data to license, what to capture, what to keep, what to use and how to use it. Too hot, too cold, too big, too small—all have costs and risks associated with them. But by getting it just right, we can reap wonderful rewards—and they’re all better than a warm bowl of porridge.

David Danziger is director of data and targeting products at Acxiom Corporation.