In hiring digital veteran Mark S. Zagorski (formerly of Modem Media, WorldNow and MediaSpan Group), the Israel-based targeting data exchange eXelate also opens its New York pffice and U.S. operation this month. But how does an exchange bring value to both publishers and networks? As vertical networks proliferate like kudzu, and some publishers worry that further commoditization of inventory undermines their brand, we asked Zagorski, now eXelate's chief revenue officer, to walk us through the model, the proposition, and its place in the controversies.
Behavioral Insider: What is an ad data exchange?
Mark Zagorski: EXelate has a targeting marketplace. Our sellers of data are publishers in specific verticals and the buyers are ad networks that want to target against vertical publisher cookie data. We created the mechanism for connecting ad networks that want to target specific verticals with publishers who want a revenue stream that doesn't involve using any of their ad inventory.
BI: So the publishers are just selling the cookie?
Zagorski: Exactly. It is like a Tacoda or Revenue Science model, except we don't have an ad network. Our customers are ad networks and our affiliates are publishers. On the publisher side, instead of having one company place a cookie on your site and then pay you based on how well their ad sales team may or may not have sold against that cookie, we can have multiple networks sell against that same cookie and pay the publisher for every time that gets targeted against.
BI: Who decides on the segmentation?
Zagorski: Our role is to be middleman. We build verticals based on pre-qualifying those publishers within those verticals, looking at their data, testing data in our lab to see how it works and if it is valid data. We launched four different verticals: small business, travel, Hispanic and a demographic channel, which included age and gender information.
BI: But within a vertical like travel, can you target more precisely or are you selling travel -interested people?
Zagorski: At this point since we are going broader rather than deeper, we are looking at very top level information, so it is travel interests. There is a lot of buzz around data right now. Companies are looking at getting ISP data and capturing every piece of information from a Web browser experience. They know everything about that person. The challenge is that all that data is so deep that they have very little footprint. We basically have said let's look for quick targeting triggers on a very broad audience so that we can look for fresh data on a large number of unique visitors. So we are reaching almost 70 million uniques per month from whom we have accumulated a piece of cookie data. It is more important for us to go broad rather than deep at this point.
BI: But what are ad networks getting from this that they wouldn't be getting from the tons of publishing partners they already have?
Zagorski: A lot of ad networks out there don't have behavioral information at all. They don't have a lot of pixel-targeting data to work with. We provide them with a very easy way to build a BT enterprise. For the larger guys who have their own behavioral information, this is meant to supplement that as well. We haven't found anyone who said they have so much information that they don't need any more right now. There is a sweet spot for us with certain networks sizes that haven't been able to create a publisher relationship with behavioral. The fundamental idea is that if I am selling a run of network ad for fifty cents, now I can sell a targeted travel ad for $2.50.
BI: How many publishers, ad networks are working with you?
Zagorski: We are not transparent on the publisher side, which is important, because publishers can be both buyers and sellers in the marketplace. So you are looking at several dozen sites. On the ad network side we have about a dozen active networks and several that are more active than others.
BI: If it is not transparent on the publisher side, then how does that affect evaluating the quality of the audience on the network side? How can they determine whether the cookie has passed through quality content?
Zagorski: Our job is to make sure that the publishers are qualified. We optimize the data they provide and then we help the ad networks monetize it.
BI: You open an exchange just as a controversy over the value of ad networks and exchanges for branded media is flaring. What is your position in the argument initiated by Wenda Millard of Martha Stewart Living Omnimedia about media fears that the models devalue inventory?
Zagorski: It is a very interesting point, coming from my background where I sold all branded advertising inventory. I had no background in selling DR or BT. It was all brand-based in specific local and national media sites and verticals.
I totally understand where Wenda Millard is coming from in saying we have to avoid this from happening. But the reality is that you can't stop the technology and the way mass volume is going to play a role. You can't go back and say we are going to be like TV now and be transparent and sell the brand idea and not base anything on response. Those days are over.
The days for advertising and big media companies being able to sell that way are gone, because everyone wants to have some type of return metric. And return metrics naturally lend themselves to technology-based systems that can maximize those return metrics by looking at volumes and things like that.
I think that it is definitely a noble way to go, and I think there are certain sites that can do that. But you probably are looking at the top 10% of properties online that will be able to sell only brands and sell on their brand and keep the CPMs high.
The reality is that the majority of the inventory out there still needs help in maximizing the return. And behavioral and other targeting methods that are performed by ad networks help that to happen.
No one sells out. Everyone knows that no one sells out in this business. And that inventory has to move in some way, and the way it gets moved is through ad networks. And the way ad networks maximize their revenues from that is by using targeting tools.
I think there is a role for the branding. There is a role for ad networks. And I do think that programs like ours and programs where publishers can make revenue without selling their inventory are going to help enhance their ability to focus on branding. But it certainly isn't going to supplant their relationship with ad networks at all.