Jason Ramsey is Director of Program Management at Microsoft and helps guide the ongoing development of Microsoft's advertiser-side ad serving technology known as Atlas. Ramsey helped drive development at Atlas prior to Microsoft's acquisition of aQuantive in May 2007.
As part of AdExchanger's "State of" series, we spoke to Ramsey last Monday about the state of Microsoft Atlas ad serving and industry trends.
Click below or scroll for more:
- The Major Trends Today
- Engagement Mapping & Atlas
- Differentiation In Ad Serving
- Going Cross-Channel
- Brand Dollars and RTB
JR: The major trends that I've seen in the last few years – since the end of the period when Atlas was part of aQuantive - are programmatic media and the increasing amount of inventory that's available via that mechanism both in display and, increasingly, in mobile and other channels. The complexity that creates for Atlas customers is something we're focused on through partnerships in our Atlas Technology Partnership Alliance (ATPA). We have deals with several DSPs. We're working to ensure that the integrations and partnerships are successful for our customers.
The other big trend is audience [buying] which was initially done through context. In other words, you would look at publisher‑side context and then get an audience composition indexed against who you thought was there. Now you are able to go after audiences directly.
In Atlas, we've addressed this by building out audience management tools. We released a first new version of that back in November, and we'll continue to iterate on that.
Then, putting these two trends together - audiences and programmatic – our customers' need to understand and measure across all of those things is paramount. That's been the goal of ad servers from the beginning - help advertiser marketer customers understand exactly what their media dollars are doing, how their spend is being executed and translate it into performance.
Atlas was at the forefront of attribution management five or six years ago with "engagement mapping." What can you say about how that technology and methodology has progressed? Do you call it engagement mapping now?
We do. It's one of our priorities. Our first, full release of engagement mapping was four years ago. What we've done recently is increase the frequency and powering up what's going on with the reporting system behind the scenes. We're running our reporting four times a day now versus daily, and integrating additional analytics reports on top of all of that.
There's a sense that Atlas has fallen behind, certainly from a market share point of view. What can you say about the resources behind the development of Atlas today? Is there a commitment there to continue to grow and integrate Atlas or is it more about the status quo?
We're definitely committed to the success of Atlas. A lot of that perception has come out of the fact that in the past couple of years we did a lot of work around building APIs and transforming Atlas into a platform. For a period of time, we had releases that didn't have the flashy customer impact, but in the last six months, we released something customer‑facing and impactful every month. We're also now seeing the dividends of moving to a platform with the number of partnerships that we can sign, and the integrations we can pull off.
For Atlas customers, we see all of their impressions across all of their buys. Our customers will use Atlas in addition to DSPs or other toolsets that plug into exchange inventory. But that's just one channel for Atlas customers. To provide the kind of insight and management and performance attribution that they require, they have to use a base platform like Atlas that goes across everything including RTB, search, guaranteed display, social, and all the other channels. We continue to integrate with DSPs as that's a critical piece. So I don't see a conflict there, necessarily.
How do you think about your offering in terms of Google's DoubleClick offering?
We have a decidedly different strategy than Google. We're very much about being an open platform and allowing our customers choice and allowing them to pick the different solutions out there in the ecosystem that work for them in different areas.
Another part of it is in regards to the pace of innovation ‑ the one thing that surprised me the most in the last five or six years is how the pace of innovation and the complexity of digital advertising has continued to accelerate. In that world, taking a very vertical approach and locking everyone into one set of tools is very limiting to customers.
Companies out there are very open to partnering with us because we have this [open] approach. Our customers continue to push us to help them integrate and make sense of all of this. Google has their perspective and we differ in this regard.
Can you talk a little bit about where the sweet spot is for Atlas in terms of the target market?
I would say our target market continues to be the same, which is top agencies and their top clients. We've long had a mix of brand and direct clients. I don't see that changing. The goal for us is to really help them navigate through some of the challenges I've mentioned earlier, particularly around understanding how to take their data and audiences and load it into Atlas, where they can track, measure and manage across all the media channels that they have access to.
A publisher ad server and an advertiser ad server would appear to share many capabilities. I wonder if you could talk about why it's important that you don't offer a publisher‑side ad server?
You're right, in that to some degree, with the ad servers and its underlying technology - particularly once you get to the point where Atlas is at high‑scale with a broad set of capabilities and a good set of APIs on top - you can take that platform and use it for multiple businesses. In fact, we've done things like that where we've built the technology that powered DRIVEpm and Microsoft Media Network off of that, but for Atlas specifically, we continue to be focused on our buy‑side customers of advertisers and agencies and marketers. That's our focus going forward.
In that area, we've struck a number of partnerships to help us go across those channels. We're in process right now with setting up some mobile partners. That's in process so I won't talk about that. In video, we have partnerships with Extreme Reach, Flashtalking and YuMe, and search partnerships with Kenshoo and Marin. As new channels continue to evolve, we'll work with other companies who have specific expertise in how those differ.
In the general sense of tracking digital across multiple channels and devices, from an Atlas standpoint, we would view that as a generic set of inventory. We would try to normalize what that looks like for our customers inside of Atlas and give them a single experience and a single measurement platform across all of those.
When do you think this more efficient cross‑channel world will really come together and what might be some of those triggers?
We talk to partners all the time who are trying to solve this problem. The cross‑device, cross‑platform ID problem is obviously one that is still emerging and looking for solutions. In Atlas today, we can track that and report on that up to a point, but there continues to be a ‘hold' for everybody really as you go across devices. Via our partner program, we'll continue to look for ways to bridge that because you're right - getting a single view for a single user as they migrate across different ways that they interact with a brand or an advertiser is key to what Atlas is about.
From a timing standpoint, I don't know. I know that there are a lot of people working on it really hard and we talk to those guys all the time, but it would be speculation for me to guess on a date.
The key there is to understand - for brand dollars - what happens at the top of the funnel. For those exposures and customer interactions where people are looking at and interacting with your brand, yet don't necessarily click or travel to the site and deliver some kind of action, it's telling a better story.
Engagement mapping is certainly a solution that is driven to answer that question. Through our partner program, we have worked with rich media and DCO vendors to instrument a better way by which we can track interactions inside the creative units themselves.
This is an extension of what we did with Atlas Rich Media, extending it to partners and getting that full view. When you start to do that, you can start to rationalize the spend between on‑ and offline and get it closer to the time spent in these different environments.
What's the reality of real-time audience buying today - is it still early days? Where do you see it down the road?
Real-time bidding (RTB) is one part of the campaign today and it's a very efficient way for customers to go execute on an audience that they've constructed, for example. But, direct buys and other buys will continue to be an important part of the plan.
From our perspective, it's about tying those two together and bridging them, and telling that as one complete story. What you can do in RTB versus non‑RTB is slightly different. But from an advertiser perspective of saying, "Hey, I'm trying to talk about my brand with these customers," they want to try to deliver the same message across both of those.
Our goal is to figure out how to creative a seamless, yet powerful experience for our customers.
By John Ebbert