ARUN KUMAR: It’s strengthening what we’ve been doing. The boundaries between what is defined as programmatic and traditional media are blurring. Frankly I think in about a year to two years’ time, the word programmatic will become meaningless. The reason is we’re so in love with the trading mechanisms as opposed to what programmatic has actually set in place.
What do you mean?
For example, I may be looking at targeting millennials who are light-beverage drinkers and I may be targeting them on TV, video and mobile. While I might be using programmatic to reach them on TV and mobile, the difference will be I’m going to be using the same data sets and I will be as granular in targeting TV as I am targeting video and mobile.
So to me, the trading question is far different than the targeting question. And I’m far more interested in data-driven marketing than only data-driven trading, which is the direction Cadreon is headed – we don’t see ourselves as a trading desk and purely as a programmatic unit.
How has it evolved?
Programmatic is becoming less of a media efficiency story and more of a media intelligence story. What do I do with ID-stitching? How do I get smarter on all media vs. how do I extract one more penny out of my media spend? Our goal is to enable our sister units to use that intelligence as well. So we will democratize the data and the tradecraft we have across the rest of the organization. Even if MAGNA or UM is doing a buy, it will be done against a data stack we developed. As far as we’re concerned, the fact that UM buys it doesn’t really matter to us. We’re increasingly seeing boundaries blurring.
Will more agencies follow Publicis’ lead by decentralizing programmatic trading desk operations?
No. You’ve got your tradecraft, knowledge, experience locked up inside what used to be the trading desk. The people who work at the trading desk are very different than those who work at the agencies. I’ve got 70 people sitting in San Francisco who come from a technology, startup, software background. They didn’t join us to become part of a media agency.
At the same time, you have to make sure that knowledge and the changes that are initiated are not isolated to one unit, which is what happened to programmatic in the past. The trading desks held everything. You have to democratize the intelligence across the organization, but you don’t disband the unit.
So your operations won't be decentralized?
I absolutely think it’s the wrong thing to do, pushing it back into the agencies, because that relies on one fundamental assumption, which is all the platforms out there are fairly standardized and that you don’t really need to architect and that you just need to build. The design component is even more important than it used to be.
This whole idea of, “I’ve bought a platform and I’m going to integrate it at the agency and I just do the research and development at the center?” Nope. It doesn’t work. But at the same time, are we a silo that sits separately and says, “Here’s our operating model – take it or leave it?” No. We have evolved and changed.
We’ll make sure the rest of the organization gets access to the intelligence and the capabilities, but there will be a Cadreon. There will always be a Cadreon. Seventy people are just the central technology and engineering group and then there are more than 500 on the campaign management teams across the world. We don’t want to create conflicting groups within the company, which [is mitigated by] creating a collaborative culture internally. You’re not going to find the media agencies or Cadreon or MAGNA on different pages. There will always be disagreements, but that’s a challenge for leadership. Poor leadership is what leads to fragmentation.
Let’s talk about TV. Who handles TV buys, given Cadreon’s development of an advanced TV platform?
You cannot have a siloed unit that is doing advanced TV. A lot of this is change management and how you get different groups of people to work together. Magna has clearly got buying clout. If I stand up tomorrow and say Cadreon needs to do TV buying, it would be stupid. Every time we have a TV conversation, there will be multiple people in the room. The UM and Initiative guys have been doing TV planning and buying for years, so it won’t be one Cadreon guy and one Magna guy sitting in a room.
The reason you can still split these out is because the trading relationships are different. You can use separate agencies to buy digital and TV, but when the two of them merge, particularly when I can use a platform like what we created with TubeMogul to buy TV and video, why would I have separate agencies buying video and advanced TV? It’s going to become cross-screen.
What are your thoughts on the viewability debate?
The root of viewability is people don’t find what you’re offering engaging. It doesn’t matter if your standard says it’s been viewed. In my head, if I haven’t seen it, it wasn’t viewed. As you get closer to making a decision, the number of ads I see multiplies. Accountable marketing has not taken off because there’s still uncertainty and you feel the need to spray everyone with messages.
What's your priority?
Millennials especially don’t want to watch ads. We're examining, "How do we reduce the number of impressions while still hitting the objectives we’ve set for ourselves?" That means fewer people will be irritated by ads we’re showing them. The ability to transfer intelligence cross-platforms is so vital. That’s why we need this data stack. You do a frequency cap in one channel, but if you ignore the others, that person keeps seeing the ad again and again.
The interview has been edited for length and clarity.