Plans & Pricing
Does My Company Subscribe?
Glenn FishbackHead of Global DisplayeBay Enterprise
Glenn Fishback, head of global display for eBay Enterprise, eBay’s advertising division, works with Internet 500 retailers to drive sales and demand. Increasingly, those efforts mandate a cross-device approach. Fishback spoke with eMarketer’s Lauren Fisher about how cross-device targeting tends to be a more effective tactic for customer retention vs. prospecting.
eMarketer: Where is the industry in terms of its cross-device targeting sophistication?
Glenn Fishback: The space is evolving, but we’re still in diapers. With cross-device ad targeting, you’ve got your two main approaches: deterministic and probabilistic.
With regard to the probabilistic space, that’s where most of the players are, but there’s still very much a black box about how these companies are building their panels and algorithms. So it can be difficult for advertisers to understand which cross-device tracking technology is optimal.
eMarketer: On one level, you have the deterministic approach, which relies on identifiable user information, such as a login or email address. And on the other hand, you have the probabilistic method, which uses publicly available ad-serving data to tie an individual’s identity to his or her device using statistical analysis. Is eBay Enterprise using both approaches?
Fishback: Yes, we’re using both. We’re uniquely positioned because of our access to eBay’s proprietary commerce data. Last year, we generated over $200 billion in commerce across the web, with $20 billion on mobile devices. That gives us access to a significant amount of deterministic data because we know when people sign in on their websites or mobile devices.
One of the most important things we’re doing with cross-device targeting is we’re trying to be mindful of putting the consumer at the center. Privacy is very important to us, and we believe that whatever solution we create has to be really privacy friendly because of the amount of data we have access to.
eMarketer: What are some of the ways your clients are using cross-device targeting?
Fishback: Clients are discovering it’s extremely effective for customer attention vs. demand generation. It’s much better to talk to existing customers on multiple devices vs. new ones, because targeting new individuals in this way really narrows our view.
Even if you have CRM [customer relationship management] data from a retailer or a target list, in terms of reach, by the time you refine that list by every other component, you end up with quite a small sample that you’re hoping to reach and that you can identify with certainty.
It’s very effective for upselling or retaining existing customers when you can factor in things like login and transactional data over a period of time to talk to them at different moments and personalize that on different devices. That’s much more effective than targeting someone who read three articles about Porsches on their mobile devices and desktop and then assume they might be in-market for a car.
I compare this to traditional ad retargeting in that you need to consider how you can be personalized yet polite and understand a consumer’s expression of purchase intent. That’s harder to do with new customers than it is with existing ones where you know what they have been responsive to in the past and can hit them across devices at appropriate times.
eMarketer: Going back to the two main approaches, deterministic and probabilistic, are there any players in particular you see emerging as leaders in the space?
Fishback: There are probably going to be a few winners in the deterministic space. Facebook and Google are obvious ones. Amazon could be another one, and so could eBay—really, any of the big dogs and major portals where a company can connect an individual’s identity all the way across devices.
On the probabilistic front, you see all these startups that are algorithmic-based. But those who are taking an algorithmic-based approach need to be transparent when it comes to what their methods comprise, because it’s very hard for advertisers to know what they are looking at, such as patterns in operating systems, carriers, locations and demographic targeting, to name a few, and it’s really hard to understand how they’re doing it.
There’s another piece to this, though, and that’s the fact that some of these probabilistic vendors might be using your data. In some instances, you see advertisers becoming part of this data co-op. So the marketers’ data is being used to build out [a platform’s] network, and it becomes part of their data, and the company isn’t being transparent about that. That worries me from a privacy perspective.
Learn more about eMarketer data and insights »
You've never experienced research like this.
Nearly all Fortune 500 companies rely on us.
Inquire about corporate subscriptions today.