This is the last in my three-part series “Lessons from the 1:1 Printing Gurus.” Last week, we talked to Jeff Stewart, chief technology officer of Trekk Cross-Media. This week, we round out the series by talking to Kate Dunn, president and owner of Digital Innovations Group.
HTW: When people think about databases, they often think strategy. Rarely do you hear people talk about the raw data. What advice do you have for them?
Dunn: A lot of people think that once they buy data, they are ready to create a 1:1 campaign. But the data clients provide or the data you buy rarely can be used in its raw state.
For example, you might buy a list that tells you whether someone is married or single. However, you aren’t going to send out a piece that says, “John, you’re single, so you might be interested in this . . . .” Instead, you’ll say, this group is single, so we’ll use this picture and this messaging. This group is married, so we’ll use that picture and that messaging.” Married or single is what the database says. You have to extrapolate how you want the data to fall into your piece.
HTW: You care how the piece is going to read, too. I know you have some thoughts about that.
Dunn: Yes! For example, if you are looking to sell graduate programs to an audience with specific undergraduate degrees, the data might come in abbreviated, such as BA His for Bachelor of the Arts in History. How do you want to portray that in my copy? You certainly don’t want to say, “John, since receiving your BA His. . .” Or if a bank wants to mail to everyone who doesn’t already have a checking account with its institution, it may send you a database in which the people who have checking accounts are flagged. Now you have to turn that around and use that to select out the people you don’t want to contact.
Of course, in some cases, you are just dropping in raw data like an account balance or a product ID. But more often, you need to massage the data so that it reads properly. Say you will be using the name of someone’s alma mater in a fundraiser. You don’t want to say, “As a graduate of Virginia Polytechnic Institute and State University . . . “ That’s how the data will show up, but nobody talks like that. They’d say Virginia Tech.
HTW: Even if you have a lot of data, you don’t necessarily want to let the recipient know how much data you are using. Can you talk a little about that?
Dunn: I agree. If you tip your hand too much, you could end up scaring people. For example, you don’t want to send a piece that says, “Kate, I know you love Stuart Weitzman shoes because you bought 15 pairs last year.” That’s stalkish. You can use that data, but you need to use it in a different way. Instead, when the postcard comes, it should show the stylish pumps you know I like, and the copy should say something like, “Buy one, get one free” or whatever offer you want to make. It accomplishes the same thing without being scary.
That’s what I mean that the data is never ready to flow into the piece. You have to massage the data and apply logic to accomplish the goal you want. In other words, data is data. It’s not necessarily consumable. You may have to make the data consumable.