‘A fragmented, partial view’: MullenLowe Mediahub president Sean Corcoran on attribution woes

You’re seeing more but it’s fragmented?
You have more partial pieces of a pie. You just don’t have a great view. [For example,] now you have location data, but is location data accurate? Is the partner you’re using really willing to give you the true data that they’re using? Is it a panel? Not a panel? Is it “census data?” Is it statistically significant? The more we introduce, the more little pieces we get.

Also, the stuff we always had a nice clean view like the desktop cookie world is being reduced down, whether through ad blocking or regulation or whatever have you. It creates a fragmented partial view.

Holding companies have made major acquisitions of data marketing firms at the same time that attribution has become more difficult. IPG acquired Acxiom. Publicis Groupe bought Epsilon. Will these big acquisitions help?
Potentially. They use a people-based model. Instead of tracking devices and software, which is essentially what we’ve been doing in the digital space. They’re people based in the sense that we’ll be able to take first-party data, go find more of them and actually find humans across their devices. It feels like more of a targeting capability.

So agencies can use these databases to target people with more information?
It’s like a database of human beings with technology around it that allows us to go and find those human beings. It’s kind of like direct mail, basically. We’re all going to direct mail. I can find “Katie” but her name is scrubbed out. Her personally identifiable information (PII) is scrubbed out. But I know that she has certain traits, where she lives or demographics, some of the things about her. She comes with devices that I can find and I can target her on that laptop, on that phone, on our TV at home.

How does that help to look at attribution?
The idea behind that is like a closed loop system, quote unquote. I think everyone’s trying to move in that direction. A lot of these systems are moving towards a people-based model. And the idea behind it is to get to a place where yes, we can convert, and understand who converted and then look to see and provide weighting and attribution towards which platforms delivered on that. So, the answer is yes because you can get to like a one-to-one match of people.

But I still don’t think it solves for the other things. If you can’t track them through the things you track before, like cookies and other things, it’s still a partial view, right? It’s a partial view into what those people did. You’re still going to have a challenge.