It’s been over a year since ad tech and video ad distribution company DG acquired semantic data provider Peer39 and folded it into DG’s MediaMind online division. During that time, Peer39-er Alex White has had a front row seat to his company’s integration into DG with his new role as GM of DG’s Data and Trading operations.
White says that although being in a larger company can sometimes make it harder to get things done, the new resources have provided significant benefit to Peer39's business. According to White, DG has quietly built infrastructure support that will allow it “to scale at the speed of RTB,” which might have otherwise become prohibitive for an independent Peer39.
AdExchanger spoke to White last week about Peer39 and the use of his company’s data in digital advertising.
AdExchanger: Looking at the Peer39 products in particular, is it all about integrating your semantic technology into exchanges, or are there other opportunities?
But there are additional product offerings that exist around analytics. We're very similar to how the DMP platforms have evolved over time; eventually, you have to show how the data is responsive.
We're both in-platform and outside of the platforms. What this means is we provide the ability to look at companies' campaign results through the lens of Peer39 data – i.e. discovery.
For example, we have the ability for companies to have an API integration with us, in order to analyze clicks and conversion data across behaviors and overlay it with Peer39’s page-level attribute data so that they can understand exactly what the sweet spots are. It's really about figuring out what the value is. Real-time bidding (RTB) is all about placing specific value on individual impressions based on what an advertiser knows about that impression.
Having Peer39’s data changes that evaluation, even though you're targeting it to the same user, because the user in one environment versus another might have different values.
There have been many trends in data-driven digital advertising in the past few years. It feels like analytics may be the next “wave.” Do you agree?
Analytics could be the next wave. On the other hand, I think analytics has always been there. There have always been ways to utilize data to optimize this or to plan that. But the sheer volume of RTB created an abundance of data – whether good or bad – that then required analytics providers to be able to scale and ingest it. The next question is how do you take all this abundance of data, boil it down to the essence of what's important to a particular advertiser, and make it all actionable.
So is an extension of Peer39 strategy to get more into the analytics world?
MediaMind has an analytics suite based on their third-party ad server. There's definitely the desire for that to move more into RTB, and it already does for any advertiser that's third-party serving with MediaMind. Co-mingling our data with the data that comes out of the ad server is highly beneficial for both analytics as well as verification services. I think that's very important. I view that differently from the large data plays that are just taking huge chunks of data and using independent analysis tools.
But just to be clear, the predominant part of the Peer39 is business providing a contextual data point in a real-time bidded auction, right?
No, I'd say it’s equally important with “quality” and “safety,” which are also large aspects of our business. The ability to proactively enable an advertiser to not show their ad in a real-time environment based on misaligned or negative content, as well as “quality” attributes that go towards page structure – things like viewability, whether or not there are comments on the page, if it's a high density of text versus low, the number of ads on the page...All three of them are equal product lines within our business.
The dominant usage of that data is within the RTB environment and a pre-bid relationship with the technology, such that any advertiser can set up their targeting to target to or away from...any combination of the various attributes that we produce. We do a significant amount of business in custom categorization where we get very granular on behalf of advertisers, especially as it relates to “safety” and keeping somebody off something that they didn't plan for.
Can you quantify how much the RTB space has grown using Peer39 data as a proxy?
Sure. We do have a number of proxies for understanding the scale and volume of RTB. Our AppNexus relationship, for instance, is pretty much everything that they see, we see. I don't think that's necessarily unique; the ability to scale at the rate they're scaling is. We're at a point where we’re processing over 50 billion requests to our platform per day, and by Q3 that's going to expand to 75 billion a day. That's just out of need. Every six months we need to make some major adjustment to the platform in order to enable the scale of it. We know this from our history with this type of data and this type of environment.
RTB really hockey-sticked between Q2 of last year and now. I don't know if it's growing at the same rate. Maybe there will continue to be growth. But we're also becoming more and more efficient at managing that growth and scale with our internal systems...to produce results at high volumes and high accuracy, as well as high coverage.
Any special challenges or opportunities that mobile is creating for Peer39?
Yes, there are. Everyone looks at mobile from two sides. There's either mobile web or apps. Apps don't seem to be going away or slowing down. It seems like that's the dominant channel of mobile as far as ads go. We have not spent a lot of time as of yet focusing on trying to make our product work within an in-app environment, and it currently doesn't. There's no product I know of that does. [With] mobile web, there's no challenge whatsoever. Peer39 has coverage on any mobile web page that exists on the internet. That's just a difference of device as opposed to different content. There are challenges that we have not yet focused on, on the app side of things.
What are you focused on?
Video and international is where we’re focused.
You're talking from a Peer39 perspective. Of course, Media Mind has focused on video, right?
In the vein of all marching to the same beat within DG, video is extremely important. Video is important in the ecosystem as well. I talked about our quality attributes and our ability to extract information as to the type of page or the type of content that's on the page, as opposed to what the content is. We're going to produce a number of video attributes that help advertisers understand or stay away from specific video attributes that they don't want to be near. Things like the size of the player or whether or not it's autoplay.
We do location of the ad unit on the page. We can do that with video as well. As it relates to the DG Media Mind verification suite of products, video verification is something that's near and dear to their hearts, and Peer39 will be the driving technology that delivers the attributes they're able to verify on, similar to how our other verification products work.
Is that timetable this year or next year?
This year. In fact, on the Peer39 side, we'll have our first attributes at the end of this month, which will be size and location of player.
Last question: can you drill down on Peer39’s most popular product use cases?
Sure. The ability to place a different value on different slices of inventory associated with the user is one of the most important aspects of what we can do within the world of RTB as well as Peer39 and what we add to the ecosystem – that’s one. Another significant business proposition for us, as I mentioned before, is custom safety solutions. Most people want to not be associated with alcohol, drugs, firearms or negative news.
As you start talking to more agencies that represent brand advertisers, they're adverse to dipping their toe in the RTB waters until they've got some level of confidence that their advertising is not going to end up on sites that don't align with their brand message. That's a very subjective thing. We have a very subjective approach to walking customers through the development of a custom category and then, because of our unique place in the marketplace, being able to release that to the partner that they've chosen to operate on, whether it be a Turn, AppNexus or whomever. It doesn't necessarily matter, but that's the most successful thing, and that all starts with having relationships with the right people within the industry who recognize that this is something that exists.
Sometimes clients require us to build brand new tools that could take a couple weeks to develop in order to satisfy the particular need of the advertiser – and sometimes we can use some of the existing toolsets that we have. In all cases, it's a very custom solution that gives the advertiser the confidence test within an environment. Once that's created...either they're very quick or it's a longer-term process, depending on how complex the need is. We then release that into the various platforms and within a day or two they're up and running and targeting with it. The taxonomy usually satisfies most people's needs outside of the safety offering.