Sometimes I think we get so caught up by the hype about how ad tech is, or will, transform advertising that we forget it is fundamentally about tech. I think the reason for this is, many of us in the ad biz don’t really know how to assess the underlying technology powering ad tech platforms, so we focus on what they say and can do, vs. how they actually do it.
Speaking for myself, I know this is true of at least one trade reporter, because I personally do not have the technological chops to assess what many ad-tech firms tell me. And even on the few occasions that they actually open their hood, I have no idea what I’m looking at.
But I’ve always admired the folks at OpenX for the plain-speaking way they describe their technology, how it works, what it is intended to do -- and, perhaps most importantly, why.
So when the company announced a deal late last week to move its underlying data-processing architecture from its own servers to the cloud, I was eager to understand why.
On the surface, it makes the same sense to me that has driven publishers and app developers to cloud-based servers, which is that if it’s not your core business, why are you toiling in server farms? But while many if not most publishers have migrated to Amazon’s AWS, the first interesting part of OpenX’s announcement was that it was going with Google.
The deal, valued at more than $100 million, will shift the entire OpenX exchange to Google’s Cloud Platform over the next several months, with the first benefits expected to come by the end of the second quarter of this year -- most notably a more stable and faster data processing of ad exchanges conducted via OpenX.
The reason they will be faster, OpenX CEO Tim Cadogan explained to me, is location, location, location. By operating on Google’s cloud, OpenX is that much closer to Google’s AdX platform, which is a major partner of OpenX.
People in high-speed trading like Wall Street’s “flashboys” call this “dark fiber,” which means that the closer you get to the trade you’re trying to process, the faster you can see opportunities and transact on the ones you want.
Even if it’s just a hundredth millisecond of speed, in the world of high-speed trading, it can make the difference between your bid and the next bidder's winning the transaction.
The other initial benefit to OpenX is globalization. As it expands its coverage into more distant markets -- Asia Pacific, for example -- OpenX doesn’t need to invest in new servers and can instantly hit the ground running, because Google’s cloud already has dependable, global coverage.
But the most interesting part of the wonky tech deal is how it will enable OpenX to focus on more meaningful areas of development -- what we, in the ad industry, like to call “innovation.”
“It’s the difference between Google buying millions of servers a year vs. us buying thousands,” Cadogan explains, adding that it’s not just the sunk costs of buying, maintaining and replacing hardware, but the opportunity cost of what OpenX could otherwise be engineering.
The first product innovation OpenX will redeploy those resources to is unlocking the power of the “open web.”
By that, Cadogan means enabling advertisers and publishers to create the kinds of audience-based connections on the open web that to date have only been possible via a so-called “walled garden.”
The irony of the fact that Google is one of those walled gardens hasn’t escaped Cadogan, but in the current world of frenemies, Google is a multifaceted partner of OpenX: a buyer, a seller, an enabler, and now part of its underlying infrastructure, too.
“What we’re seeing is that our marketer partners would like to market that way on the open web,” Cadogan says of the explicit audience-targeting capacity of walled gardens like Google, Facebook, Amazon, etc.
“That’s the big thing that we’re going to do, and it’s why we need a lot more computing power to do it,” he explains. The first versions of that opportunity, he says, will come out in the second quarter.