How Agencies Are Retooling for the Automated Ad Buy

If you want to know the ins and outs of programmatic, Jay Sears, svp of market development for the Rubicon Project, is probably the guy who should explain it to you. Sears has a long history with ambitious, technically complex ventures, notably the finance world’s vital Edgar Online tool. (Take it from a reporter: Edgar has made getting public financial data worlds easier.) Now, Sears is working to evangelize on behalf of automated buying technology like the platform Rubicon provides. It’s not going away, he says, and regardless of what platform or medium you sell ads for, you’re going to be seeing a lot more about programmatic soon. Adweek: So, I think there’s confusion about what the difference between programmatic and real-time bidding is. Can you help us out? Sears: It used to be that—and still, in many parts of the business, it is—that inventory was sold in gigantic blocks, and when you broke the block down, you’d find that part of it really worked for the situation you were in and part of it didn’t. The magic of programmatic is that it’s about de-averaging and it’s about data. De-averaging means you’re decomposing everything down to an impression level, and you’re using machines to look at it. You’re making decisions in milliseconds—and it’s very easy to make those decisions at the impression level and then reaggregate them so you can continue to buy at scale. I’ve talked to people in a few markets who are totally unimpressed with programmatic buying, notably TV ad sales pros. But your company is called the Rubicon Project. You clearly believe that this is a change there’s no going back from. The automation of the buying and selling of advertising is inevitable. You only need to look at history. Look at the stock market and the evolution of trading—look where it started. It started with over-the-counter stocks, and now every blue-chip name is automated. It started in travel with Sabre Systems, and originally Sabre Systems helped monetize unsold airplane tickets only on American Airlines. Now it’s the booking agent for essentially the entire travel industry. Look at something like eBay where the first items sold on eBay were Pez dispensers. EBay in its initial iteration was essentially a giant yard sale, and now a new car is sold on eBay every single minute. You can draw the same analogy to the buying and selling of advertising. So the TV people are just completely wrong? Automation will envelop anything that can be treated at scale. There’s always going to be a remaining piece that involves deep integration, manual work, custom content. We’re not talking about that. We are talking about everything else. How is that changing agencies? Well, when you look at a holding company, you’re talking about tens upon tens and maybe hundreds of thousands of people around the world, and if you listen to how some of the trading desk people talk, they’re organizing businesses around this march to automation. The Magna Global group is talking about the hub-and-spoke model. The planning function is taking place at their agencies, but they’re centralizing the buying function within Magna. Is Magna’s hub-and-spoke idea the only new model? No, other holding companies are doing it a little differently. Some agencies are doing what they call an “embedded” model. They take their trading specialist from the agency’s trading functions and actually put staff out into the operating agencies. This is about informing the advertising client much earlier about the benefits of automation. It seems like what the models have in common is that they put the automation guy in the thick of the deals rather than scraping around for discount inventory in his spare time. Right. If you and I are running one of these global holding companies and we’ve started to build a practice through our agency trading desk, you’ve got a really nice specialization inside that agency trading desk. For almost every holding company, that’s a very successful and profitable business line. Now the question is, How do you and I take that and make that facility available across our holding company in a way that facilitates and takes advantage of that transformation? Well, the answer is that we make sure the trading facility that’s available in programmatic fits right alongside—and very early in—the planning process. Because if we don’t, we can’t re-engineer things at the end of the process. We want to see the specialists sitting at the same table as the folks designing the strategy and designing the plan. So how fast is this coming? We’re quite certain automation will touch every corner of advertising. If you want to see the way the world is going, you need to be studying the display market, the mobile market and, after those two, the video market. All areas are ripe for automation. If we go back to the Interpublic Group’s Magna Global example, they’ve said publicly that within three years they intend to automate fully half of the media that they buy on behalf of their clients. They’re not talking about automating half of digital display or half of digital—they’re talking about half. Full stop. That’s a pretty big nuance many people don’t understand. It’s coming fast and furious. Are we going to leap into this and say, “Oh, shit. My software has no idea how to make our plan compatible with what’s for sale.” This is another thing I think is so interesting. If you and I were having this conversation two or three years ago, yes, it would be mostly about the technology in the middle. We’d be using horrible acronyms that none of our friends understood and talking about how the pieces of technology in the middle couldn’t talk to each other, and we’d need to have a geekfest to figure it out. Those times have mostly come and gone. It works. [jwplayer sirGMbRS-1kT9Z8rj]

Brand Marketing

How Agencies Are Retooling for the Automated Ad Buy

Q&A with Jay Sears, svp of marketplace development for the Rubicon Project

Sam Thielman

About

Subscriptions

Events

Publications

WORK SMARTER - LEARN, GROW AND BE INSPIRED.

Subscribe today!

Sam Thielman