BRIAN BAUMGART: We looked at the landscape and said programmatic is seven-plus years-old in display. It’s taken a while to get to the level of traction that they have from both an inventory and a motivation perspective. In mobile, what we see is the opportunity to leverage programmatic is growing much more rapidly, and so it’s process that we’ve seen mature quite quickly over the last 18 months. I think there are a couple reasons why, but before we get into that our vision really is to close the gap on the maturity and the feature set, an provide a more comprehensive ability to execute a buy on a programmatic platform in mobile.
Agencies and advertisers need to trust in the medium. We see the gap really in mobile as the last mile in the overall workflow and chain of vendors and technologies. That means the publisher side has done a really good job of maturing in the last two or three years. The big challenge is with the SDK. Those guys have done a really good job of making 70-% of mobile display programmatically available in the last 18 months.
The last mile really is the buy side and we’ve seen companies go to market with new technology platform that’s not feature complete. We’ve seen people jump directly to new shiny, sexy things like hyperlocal and geo-fencing and trying to bridge a PC display user to a mobile screen. We’ve seen the display guys who have a very mature product offering in the display world extend into mobile with a very, very scaled down feature set.
We say there’s no reason why you can’t have the same feature parity with a more mature DSP and developed channel. That’s been our goal to close that gap and allow advertisers to have a tool that’s on equal build as some of their other channels that they’re buying in.
Do you see yourselves as a mobile demand-side platform (DSP)? Or are you more of a mobile exchange operator essentially catering to both the buy and sell sides?
We’re characterizing ourselves right now as a mobile DSP. Who knows what the future will hold [for us], but we see some great advancements by the exchanges that are currently in-market. They’re starting to go to a more publisher-side model. And that leaves a more efficient playing field for those clearly on either the buy side or the sell side. In display, you’ve got a lot of stuff in the middle and people “crossing lines,” where some companies were positioned as publisher-centric, but now they’re starting to extend tool sets into the agency, trading desk world and that’s causing conflict with their DSP.
In mobile, we’re seeing a cleaner playing field. Exchanges are focused on publisher-oriented tools. That leaves us the opportunity to build [buy side] technologies, with the advantages of having stronger ties solely on the buy side.
One of the challenges for mobile RTB is the lack of a clear tracking tool like the cookie on the PC side. There’s been a lot of talk about the value and the limits of fingerprinting on Apple devices. How does that affect the prospects of mobile programmatic?
You could probably approach the question from a couple different areas. One is looking at the display world and looking at the desktop world. The cookie is something we’ve all relied on and I’ve got a 12-year history in desktop display, as do my partners. There are a lot of challenges with cookies. There are a lot of challenges with cookie refreshing, with cookie deletion. There are a lot of challenges with what’s happening with browsers and forcing people to opt in. We’re not sure what’s going to happen there.
The reality is the cookie refresh rate is very, very high in the desktop world and so it’s not something that’s talked about as often. The cookie challenge is multi-faceted in display. It’s not an exact science by any means and the persistence is not that the level that most agencies and advertisers or just people in the industry would assume that it is. There are a lot of challenges with that and I think that people have done a great job with doing the best that they can to maintain that persistence and be able to have portable profiles that stay with the user, but the reality is it’s pretty difficult.
When you contrast that with mobile and not having a cookie oftentimes, we do have cookies some of the time, you have to come up with alternative solutions. In mobile it’s not an exact science either. I don’t necessarily think that in some respects it’s worse than the challenges of display. It’s just different. Without a cookie, we do get a device ID on a device about 85% on the auction requests that see. That device ID usually is persistent by exchange, meaning they don’t usually change what it is that they’re passing us on that device.
But does the device ID provide enough information for marketing purposes?
What we have to do beyond that point [of getting the device ID] is actually build a profile on the user and we take into account all of the features that are passed to us about that device and about that user so we can start to really understand what it is that we know about this device and about this person. Usually that string of information can be quite long, everything from a unique device ID to a device type, a device family, and OS and OS version. It could have browser information and a variety of other things. We have a lot of data about the user and at the end of the day we deconstruct that, we normalize that, and we roll that up to our own ID set.
We’ve built a data normalization layer that we think gives us a high degree of persistence with our user and being able to track those user and those IDs across the exchange. It’s a different type of science. It’s not an exact science, but I would also remind people that the cookie science is not exact either and the challenge there is persistence. Users really have control over that persistence in their settings. There are a lot of things that can happen that can delete the cookies and there’s cookie expiration and cookie stomping and all sorts of other things. Mobile is showing that there’s more to targeting than the cookie.