MELISSA GRADY: We used to buy banner ads around places [like] Food and Wine and [Epicurious]. With this shift to programmatic and data, now we’re buying people. If we don’t understand who those people are, we [need to] get the best probabilistic data and make it deterministic to really understand that consumer. And from an acquisition standpoint when they’re anonymous, how do we give them the experience they want to the point they become customers? It’s all based on who they are and we want to be platform-agnostic, device-agnostic. That’s our biggest challenge.
There must be challenges with attribution.
MG: I was on a panel with Facebook and Tapad recently, and it was like, “Facebook, I love your data” and “Tapad, I love your cross-platform,” but I had attribution already in place from a horizontal perspective. Now they’re giving me some stuff vertically, but you’ve fractured my horizontal attribution in a way. Those platforms and data are great but how do we get those into one single, unified view? And that’s our biggest focus right now.
Publishers like Amazon and Facebook have a big base of logged-in users. Do marketers worry about maintaining a bidirectional flow of data?
JOHN LEE: AOL, Google or Facebook have attribution tools baked into the stack, and you’ll see value for the advertiser. But it goes back to this problem Melissa addressed: I have vertical integration but I lack horizontal integration. So advertisers will need partners that live outside those walled gardens who will stitch together that single view across those walled gardens, because those walled gardens will constitute a lot of the spend in the future. I think it’s becoming very clear.
Partners, as in Merkle?
JL: My prediction is a different kind of agency will emerge from this which understands the media complexities, understands advanced data management and can be tool-agnostic whether Tapad’s building your cross device or you’re using Facebook for Custom Audiences. I know I work for Merkle, so I have a little bit of a bias, but I think we’ll see this developing services market tie together these identities [across walled gardens].
What’s been challenging about unifying the customer database and channel-specific data?
JL: It was less of an any singular technical challenge as it was bringing together the marketing and data skill set and figuring out what do we want to do with that data from an analytic perspective and how to execute on that through media. It sounds relatively straightforward, but most people we’re working with are coming from one of those two worlds. Melissa was one of these rare people who has lived through all phases of this and understands deep analytics, CRM and digital platforms.
MetLife’s done a lot of work with Facebook custom audiences. What’s been useful for you?
MG: Always start with your segment, get the messaging you think will resonate and when you go in, let all your preconceived notions go away and let the technology do its work. We’re very excited about what Facebook can bring and we’re in a beta right now [for Atlas] to see what we can do there. We don’t like black boxes and we don’t just let things optimize. We have teams that go in every week, look at the sales data, go back to our vendors and go over everything. The kinds of variables those models build, this “intelligent machine,” it should all translate into other ads.
JL: There’s this untapped potential of Facebook as a planning tool. You might not make the leap between sports enthusiasts to ESPN on television, but you may make the jump from that audience popping in a Facebook Custom Audience campaign to a guaranteed digital buy. If Facebook gives us more insight into publishers and not just audiences, it could be very cool.