Reaching the right user at the right time with the right message is the goal of digital advertising. Critical to that goal is the ability for ad tech companies to understand identity through the use of cookies. Unfortunately, it's not easy for these companies to work together without a universal ID for all to use. MediaVillage spoke with Garrett McGrath (pictured below), Vice President of Product Management,Data Services at Rubicon Project about why they are involved with the Advertising ID Consortium and the issues the consortium hopes to address.
Rob Beeler: How are cookies used by digital marketers?
Garrett McGrath: Cookies can be used as a representation or record of a person visiting a website, at a particular time, using a particular browser, etc. Based on the site that a person visits, cookies can loosely become a proxy for intent, desire or interest.
Beeler: Where do cookies fall short?
McGrath: Cookies are by definition domain based. A cookie from one domain means nothing to another cookie set in a different domain. So in order to transfer the concept of identity from one web-based entity to another, we engage in what is called a cookie sync. In order for a supply-side platform (SSP) and the various connected demand-side platforms (DSPs) to have a common understanding of a cookie set on a particular publisher's domain, they have to synchronize or share the representation of this "identity" between the various platforms. They have this sort of handshake, where they build a match table between the two sets of cookies so that in future ad requests the matched user is sent to the DSP. This enables them to refer to their own internal record of that cookie, how they value that user, what a bid for this impression opportunity should be and so on.
Beeler: What is the Advertising ID Consortium?
McGrath: The goal of the consortium is to get a meaningful number of arbiters of programmatic advertising -- the SSPs, DSPs and exchanges -- to agree to use one or a few namespaces, or cookie pools. In some ways that would act as a universal ID. The hope is this would make a significant amount of cookie synchronization unnecessary because we'd all be using cookies set in a common domain. This is really an attempt to make web-based, cookie-dependent requests act more like in-app/device ID-based requests, where the ID is the ID and no synchronization is required.
In the future, as we get more and more off the page, and more and more device-based and/or server-to-server based, the concept of identity will very likely evolve. This will start to involve the modeling of data with machine learning, something we're already thinking about. We're already thinking about a world where the models that we're using today, even based on device IDs, pivots to something different. We take privacy and transparency very seriously, so those concepts will be north stars as we consider our future product plans.
Beeler: What's the benefit to marketers and to consumers?
McGrath: Cookie syncing is a wasteful and inefficient practice for all involved. For marketers, it would result in increased fidelity of identity resulting in a better understanding of consumers. For consumers, they won't have to wait for pages to load because synchronization is happening. The volume of cookie syncing impacts user experience.
Beeler: How does header bidding make cookie syncing easier or more difficult?
McGrath: Before header bidding, publishers would place SSP tags on a page and they would have access to the domain and set a cookie. It was pretty straightforward. With header bidding, and especially server-side transactions, those tags are not on the page until the end of the transaction (if you get there at all). The impact of this is the frequency with which an ad tech vendor is on the page, and therefore able to set or synchronize a cookie, is greatly diminished.
Beeler: Is the idea to have only one universal ID?
McGrath: No one is under the illusion that we can have just one common ID across the web for advertising. Currently, there are hundreds or maybe thousands of systems and therefore millions of IDs that have to be synced up. While we're still in this web-based world, if we could simply sync less -- perhaps three or four IDs -- it would be a massive improvement.
Beeler: Is the universal ID concept a response to platforms such as Google and Facebook?
McGrath: Google, Facebook and others have an inherent advantage because they have first-party data. If parties outside these walled gardens have a more common, constant and transferrable sense of identity, then that's better for publishers and marketers. The commonality of your identity across different websites goes up. The fidelity goes up. The success of advertising and the potential for the industry to deliver consumers advertising that they actually care about improves as well.
The Rubicon Project attitude toward this whole issue is that we'll support any universal ID system that we think has merit and is worth pursuing. Our mission is to be the No. 1 independent ad exchange, which means we talk to as many different sources of supply and demand as we can. The whole world is clearly moving toward a header bidding (model) and also toward server-side interactions. That makes cookie syncing even more difficult or impossible in certain circumstances. So, at least for the web let's stop the madness and have an ID system, or a couple systems, that we all agree on.
Click the social buttons above or below to share this story with your friends and colleagues.
The opinions and points of view expressed in this content are exclusively the views of the author and/or subject(s) and do not necessarily represent the views of MediaVillage.com/MyersBizNet, Inc. management or associated writers.