AdExchanger: In this report, are you focusing on the real-time bidding (RTB) element of programmatic?
KEN SENA: My thinking of programmatic is an openness of supply. You have these pools of inventory connected to one another, which enables these interfaces that look across the inventory. It’s optimizing across a bigger pool of supply. The notion of RTB is important and it’s driven out incremental yield, but to me the bigger event in the industry was when all the ad networks got together through exchanges. You’ve seen all the inventory become interconnected.
The takeaway seems to be that tech providers are well-positioned but the companies that provide services around programmatic media buying are not.
I’d caveat it by saying the tech providers are subject to the same competitive issues and opportunities as tech providers before them, where they’re competing against one another and there’s a certain amount of leap-frogging. What makes it more challenging for them is that a lot of the technology tools they provide are moving toward open-source as well. They need to stay ahead of that as well as compete with each other [through innovation].
You indicate programmatic media companies – analogous to old ad networks – are being squeezed from both the buyer and seller side, as well as from tool and data providers. What’s going on?
You get this period of time, especially in digital and maybe in linear advertising too, where technology [hits] an inflection and buyers and sellers aren’t quick to adapt the improvements around that technology. This creates opportunities for ad networks to come in and help both sides leverage additional deficiencies. But as those buyers and sellers start to adapt those tools in-house and get up to speed a little bit, they wonder why not do it themselves? So [the ad networks] start to get marginalized.
You position some of these programmatic media companies, like Rocket Fuel and Criteo, as the next evolution of ad networks. In the meantime, traditional ad networks like Specific Media and Conversant are rebranding as tech companies. Will they be successful?
They have to do a certain amount of repositioning because they’ll face increasing pressure from buyers, sellers, agencies and even data providers, especially as they get smarter about tools around programmatic.
Why is the lack of transparency a weakness of the ad network model?
For us, programmatic, [is] all of this inventory coming together and becoming transparent to optimize for your own goals, your own yield.
But if you’re not presenting something differentiated in a transparent environment, where the marketplace is disaggregated, you have no competitive barrier. If you are a premium publisher and you have content you know folks want, the last thing you should allow is to let your content be sold in a disaggregated marketplace.
That’s why television programmers aren’t running to embrace programmatic. And you’re seeing guys like AOL give premium content through programmatic methods, but you have to buy it through them. That’s because they don’t want to allow full transparency. They want some control over it to make sure they have the pricing power.
Do you feel similarly about Condé Nast and Hearst? They have premium inventory but they’ve embraced programmatic in-house.
That’s kind of like the idea of the private exchange, which isn’t very programmatic at all. You aren’t allowing folks to buy across all inventory. You have to hook into them to get that inventory. In some cases, it’s closed off and that becomes the notion of the old ad network, where you compete on the basis of the scale of your network.
In contrast, The New York Times has been criticized for its reluctance to embrace programmatic. Based on what you’re saying, their motivation is understandable.
Other examples would be, besides television, LinkedIn. LinkedIn isn’t embracing programmatic players. You see Facebook and Twitter do it because they have an abundance of inventory. LinkedIn really doesn’t have it yet. But they can call whatever price they want because they have something really unique in terms of their B2B angle. They’re going to keep their inventory outside of what is essentially a programmatic marketplace, where everybody gets to see what the impression is worth and let the market dictate what they think the value should be.
Programmatic media companies that aren’t transparent argue that opacity is necessary to drive results.
The old ad networks competed on the scale of their network. Now their scale has gone out the window since all the inventory is somewhat connected. Now you compete on the basis of your algorithms. It seems fine to protect your algorithms and your competitive advantage, but that’s not consistent with what marketers want when they buy impressions. Yes, they want return, but they want some sense of where that return is generated.
It’s a long-term risk for the [black-box] guys. Even if an agency can’t get the same return as Rocket Fuel, they can tell [clients] where impressions are placed, that works [in favor of the agency’s] selling proposition.
So far, companies like Rocket Fuel, which has a black-box model, are doing very well.
A lot of it is sales. Ad networks are bundling together new technology features, new data sets, new inventory sets. If you have a winning selling proposition, you can get into the graces of the major accounts and you’ll show strong growth. (Ed: Rocket Fuel has stated in its financial calls that much of its growth comes from existing customers increasing investment.)
But what’s changed from a buyer and seller proposition? When we look at total display dollars, once you strip out video and Facebook, we haven’t seen a lot of growth. We also haven’t seen huge reductions in cost per acquisition, and aggregate pricing across display hasn’t changed much either.
Our argument is the buyer will win if they produce some of these tools. But sellers are equipped to realize what buyers find valuable, and adjust pricing. You win on both sides and [the value proposition] becomes neutral. The way [buyers and sellers] extract value following this is to go after any middleman.
But there’s always a time delay while buyers and sellers get smarter.
So this balancing act causes display ad spend and average cost per acquisition to plateau.
Exactly. Ad networks filled that void in the beginning, helping clients buy more efficiently. But as buyers and sellers get smarter, the ad networks’ value erodes. They need to change and find new ways to add value to existing or new customer sets.
That’s what’s happening to some extent right now. Folks that fit more squarely in the ad network area will have to move quickest and most dramatically, because they face pressure from the tool providers, the buyers, sellers, the full-service digital agencies and the data-management platforms who bring in unique data sets.
Will programmatic media companies become tech shops?
Tech shops? It’s hard to say exactly how they pivot because you can see different ways where they might [bring] on new inventory or data sets, and repackage it into newer tools to sell into agencies. That’s one direction.
And that puts them back to the ad network model. That gives them time, but doesn’t provide them with something sustainable.
You’re still bullish, though, about programmatic.
You should still be bullish about programmatic because you can still drop tremendous insights from the data you're seeing, and integrate different data sources, to find where your money should be spent efficiently. But in doing so you’ll find areas where you should be spending less.