Sramana Mitra: Can you talk about the workflow of this? Where is the data collection happening? Where is the data integration with the algorithm happening? Is this happening in real-time, batch, or pre-setting of some sort? How is this setup? If I’m a publisher who wants to plug in your system into my site, how does this work?
Kevin O’Malley: Everything is real-time, so is our data management platform, especially the technology that helps segment the data. That is done through real-time transactions. We’ll essentially place our technology on the publisher’s website. From there, we’re collecting, segmenting, and analyzing all that data in real-time. When the publisher has an advertising campaign to run, they’ll select certain segments of users and certain characteristics that they would like to target. We then would run the campaign on our platform.
The algorithm is obviously built into our media buying engine. Essentially, we’re using their data but we’re also using the data that we receive back from all of the impressions that we’re running. As Patrick mentioned earlier, there’s dozens, if not thousands, of different parameters – time of day, day of week, IP address, the types of content they’re engaging with. All of that plus the data that we gather from the publisher’s website is fed into our algorithm to help our machines essentially make smarter targeting decisions. Again, it is all real-time in terms of the data ingestion and all of the campaign-level decisions.
Sramana Mitra: Are you interfacing with the ad network? Who’s serving the ads?
Patrick Shea: We are serving the ad. The ad is serving across ad networks or large supply sources. Ad networks are starting to go by the wayside. Most media is now bought programmatically through large programmatic exchanges such as Google or Facebook. All of these companies have huge exchanges of buyers and sellers essentially buying inventory in real-time.
Sramana Mitra: Your technology is essentially negotiating with the real-time buying engines.
Patrick Shea: Exactly.
Sramana Mitra: What kind of list do you get? In a publisher site, what is the before and after do you see plugging you in?
Kevin O’Malley: It really depends on the use case that we’re talking about specifically. There are two ways of evaluating the value proposition that we’re providing. On the one hand, you can look at these straight performance analysis. There are numerous cases where we’re seeing anywhere from 2x or more that publishers are able to generate on their own site. When you look at sales or more of an e-commerce metric, we see anywhere from 20% to 25%. For some cases, it’s 100%.
For others, it’s dramatically into the multiple hundreds of percent. It really depends on the application. At the same time, one of the great value propositions that we provide is the decreased reliance on human resources on the client side. To try and replicate the different functionality that we provide, if we try to do that back in 2010 or 2011 with a manual team, you would need huge teams of dozens of people to even start approaching the level of sophistication that we can run the campaign with. In some cases, there are some direct comparisons that we can make against what they’re doing currently.
In other cases, there are new opportunities that open up. For example, in the hyper local newspaper, that’s revenue that really couldn’t be monetized. That opportunity really couldn’t be monetized without a solution like ours. In some ways, there is no comparison to what could be done before. There are huge value gains on multiple different sides of the equation, whether it’s improving performance indicators, freeing up resources on the client side so that they can act in a more strategic capacity instead of doing simple lever pulling in a less sophisticated ecosystem. We’re also opening up buying opportunities that didn’t exist before, just because the return on investment against human capital deployed in that fashion wasn’t really there.