Commentary

Come Undone

As audiences move further away from content, what roles do ad networks need to play to stay in the game?

Ad Networks: Come Undone

You may not be your khakis, but you are your data.

Media buyers always thought they were buying audience, but, in fact, they were buying eyeballs. Now, media executives agree that they really are starting to buy audience in a meaningful way. And part of the evolution of the audience becoming uncoupled from content is the inevitable paring down of ad networks.

The power is shifting from ad networks to the large ad agencies and the brands themselves. It's a moving away from a digital media landscape dominated by companies who can deliver billions of impressions to companies who can deliver precise audiences based on specific data, according to Darren Herman, president and founder of Varick Media Management.

"In my opinion, the networks that will survive in the medium long-term will be those who flip their entire model of selling impressions to selling data," Herman wrote in a white paper regarding the disintermediation of online display advertising. "The data that networks have collected over the years is extremely valuable." For example, Specific Media has dominated the automotive segment because of its wealth of data.

"Because of the rising data industry, we're seeing upstarts like BlueKai, Lookery, Datran, TargusInfo, Bizo and Exelate emerge as essentially data exchange players," Herman wrote. "We're also seeing some of the current ad networks starting to create dashboards that give media and brand folks insight into specific audiences that are visiting a particular site. Many of the major ad holding companies are moving into this space via units Havas Digital, VivaKi, B3, Ecosystem and others."

Curt Hecht, president of the VivaKi Nerve Center, the hub of Publicis Groupe's digital media services, says pressure is being put on the media companies in many ways. "The way that most advertisers have placed their dollars, and the way most agencies have placed them on their behalf, has been based on context: 'I want to be in this show, or near this program, because it helps support my brand and I like the association between the two.'?" Hecht tells OMMA. "And the other component, audience, was more generically defined by the age-gender demographic."

Then, display arrived on the scene. The ad networks and companies were creating their own techniques and approaches using technology and data to go back to the marketers and say "Now we can do this in the display marketplace," Hecht says. Dollars really started moving into the ad network marketplace from 2006 to the present. "The next iteration became the DSPs, the Demand Side Platforms," Hecht says. "What we saw in that was our ability to use scalable exchange platforms and be able to leverage at the time - because BlueKai and companies like that weren't in existence - client data coming directly from their Web sites for basic retargeting, which cut out the middleman. It cut out the ad networks, based on us really being able to take client data and put it back into the market on behalf of that individual marketer. And there still isn't any pulling of data and all the other things that become the next round of questions."

Ad Networks: Come Undone

The Light of Day
The days of ad networks multiplying like rabbits are over. "They're going to get thinned out," Hecht says. "If you break down the model, you can see the problems. Black-boxed, arbitraged inventory sold at a premium - too many of them, not enough transparency. Buying on an exchange, it's an open marketplace and anybody can do that."

Two years ago, VivaKi took a long look at that model and asked: "What is in this that we can't do?" Hecht says it looked a lot like the first iteration of search and the dependence on SEMs. "Pure-play SEMs are a thing of the past," he says. "You just don't see that anymore because it's one-dimensional and the SEMs are moving into display and social, and they're going to take on more than just doing search.

"The corollary is that the exact same thing is happening to the ad networks - the exact same thing," Hecht says. "Maybe there's one or two or three that we continue to work with because they have a model that we appreciate or they've got some technique or approach that we can't replicate. Like they've really built a competitive advantage. But if they haven't, they're probably going to go by the wayside. They absolutely will."

Hecht estimates addressable display will move into the 20- to 30-percent range of the overall display spend. "Over the next year to two you'll see more of that flow directly through the exchanges, and the middleman will be pulled out of that," he says. "The absolute number is going to grow in terms of dollars spent, as well as the percentage. For us, for both agency and publishers, it's really a matter of controlling your own destiny and you'll be fine. If you ask around with the media companies, some are totally in control of this and understand it. I think CBS is a good example of being in control. And then there's other publishers who are smaller and are not in control. When we meet with them and say 'Would you like to transact in a private exchange' the response is typically 'No, we don't want to do that.' Basically, they're afraid."

Joe Apprendi, CEO of Collective, sees both winners and losers as site-buying shifts to audience-buying. Ad exchanges won't have the biggest margins, but they will have scale. Data providers are irreplaceable in the supply chain. The field is wide open for administrators, the companies that provide ways to visualize and manage all the buys.

"It used to be the publisher controlled how to reach a consumer," Apprendi says. "There's a huge shift in terms of how media planning is bought and sold. The shift is from buying sites as a proxy to reach an audience, versus just buying the audience. If you can just buy the audience to reach a consumer, versus having to buy a site as a gatekeeper to reach a consumer, that's the catalyst for disintermediation today."

It's not just the inventory you use to target the audience, it's the data you use to target the audience, Apprendi says. "So now you've got the data provider disintermediating as well," he adds. "You need demographic data, psychographic data, purchase-intent data, whatever type of data is meaningful to that advertiser. Then you've got an ad exchange between the ad network and the publisher - Admeld, Rubicon, Google Ad Exchange, etc. - and then you've got the inventory provider at the end, the last point where the ad impression is actually delivered."

Ad Networks: Come Undone

This shift is creating havoc, Apprendi says. "The traditional cost of proprietary content development and the demand that they're getting from advertisers directly for buying site-specifically, is difficult to justify. It's harder for them to monetize proprietary journalistic homegrown content. Better models there are lower-cost content production, so [content that is] blog-driven, user-generated driven, where the pricing model makes a little more sense, you can still make a profit."

Who is going to be successful in the aggregation of data? "There will be a lot of companies there; it's not just going to be Google and Yahoo," Apprendi says. "Will there be as many as there are today? No. You can't justify an ad network aggregating access to long-tail sites - probably not that exciting. Ad exchanges are probably going to replace you. And if you're an ad network just monetizing ad exchanges and don't have any proprietary, direct-to-publisher relationships - probably not going to add a lot of value there. So you're definitely going to see a lot of compression on the ad-network side of the businesses."

Data providers are likely to come out on top. "Look to see how data providers start maybe tapping into the money being lost to inventory to better monetize audiences through whatever technology platform you choose," he says. Apprendi agrees with Hecht that DSPs are all the rage. "There's a lot of companies that could be categorized as a DSP," Apprendi says. "Some say they are; some say they're an ad network; some say they are an agency. This is what's going to be interesting: how companies that help advertisers through technology, data, targeting, analytics, plug into inventory to make the best use of it. Who are advertisers going to turn to to be more successful at audience buying versus the tried-and-true site-specific media planning and buying? There you'll definitely see a number of players being successful: the holding companies, whether they build their own proprietary technology or use third-party technology to help them do that; ad networks morphing to display demand-side platforms like Collective; or even ad exchanges offering some more demand-side platform solutions, or the ad servers themselves looking more like demand-side platforms."

The ad exchanges are best at aggregating inventory, but data targeting analytics - the services you need that manage and monetize that inventory - is a different skill set and a lot of what the advertiser wants to control, especially the data used for targeting analytics purposes. "So how that demand-side platform business plays out is kind of the middleware to plugging into inventory, whether through ad exchanges or directly to publishers," Apprendi says. "I think that's going to be the most interesting part of disintermediation in the future."

Eric Porres, chief marketing officer of social media integrator Lotame, says the "days are numbered" for networks that act as aggregators of vertical sites and continue to rely on indices of site visitation as a proxy for real people. They are numbered for those who did not invest in enhanced audience targeting capabilities (via direct integration with third-party data providers), optimization capabilities (algorithmic or otherwise), data capture and warehousing of audience behaviors beyond page consumption, and integration with media exchanges for audience extension when necessary. 

A common thread you hear is discussion of holding companies bypassing networks to go directly to media and create their own islands of inventory. Acquiring audience once, then remarketing as necessary across the exchanges through light, medium or heavy integrations with platforms (e.g., Invite Media, DataXu) is in vogue, Porres says. "And yet, completely open access to a publisher's inventory - as some advocate - undermines the ability for that publisher to discretely value its media and data assets," he says. "'Place a pixel and pray' can't be a publisher's sole recourse, nor can a publisher go with a head-in-the-sand strategy to avoid making choices around the valuation of data and media and the rise of audience marketplace demands in the past two years."

Ad networks with an established track record of understanding, interpreting and activating data and media assets will provide a superior digital-supply-chain solution. Porres says the ad network of the future will have full access to multiple third-party data sources and have those sources available in an interface that makes audience building as easy as search, and don't go "behind a curtain" to provide a solution. They will need to report on all digital marketing metrics that matter, not just clicks and conversions. They also have to provide measurement of time spent with advertising (real-time with ads in view, and not just loaded on a page) and the creation of an Attention GRP to capture value that online brings to audience experience. The evolved networks will be able to optimize campaigns in real-time - using age, gender, interests, behaviors, actions and third-party data in a marketer-centric and easily explainable way that doesn't require pulling out another black box, Porres says.

Publishers will require the networks to provide transparent views into how they use publisher data, and payments based on quality and quantity of media and data discretely, Porres adds. Advertisers and publishers will ask for integration with current ad exchanges for audience extension programs as necessary, and for audience data solutions that leverage the best of a network's existing understanding of online behavior, whether that behavior be a critical mass of known moviegoers, sports fans, auto enthusiasts, political junkies, etc. "Wave One of audience solutions lived discretely with ad networks," (e.g.,Tacoda, Audience Science) Porres says. "While there are distinct and superior optimization and reporting advantages that the network would have by running that audience within its publisher footprint, Wave Two unlocks data value and makes it available in an audience marketplace to append to any media source." 

"Adaptability, adaptability, adaptability" is the bottom line, Porres says. Ad data exchange providers "provide a clearing house for data with two competing models for payment. With that said, just about every major data provider offers the same or similar apis to access data independently and directly. To the extent that BlueKai and Exelate today, and other data exchanges of tomorrow, provide access to unique data that cannot be purchased directly and build insight solutions that extend beyond their data assets, they'll thrive."

A Modest Proposal
In just the past two years, we've seen an extraordinary shift in the opportunity to plan media based on audience concentration, which is a sea change from the audience composition standard created long ago (and still ever present to this day in traditional media and the many buy-side digital planning groups) that uses indices as proxies for whom marketers really want to reach. We're in the first mile of a long solutions foot race.

BlueKai, which has been working with ad networks from the beginning, started seeing a significant demand and activity from dsps and agency trading desks last year, says BlueKai CEO Omar Tawakol. "In order to sell very targeted audiences, the trading desks need granularity, scale and performance," Tawakol says. "Our aim was to offer the granularity of search in a banner-video world by disaggregating the data from the media. Our goal was to provide scale and coverage at the most granular aspects of targeting commercial searchers."

The existence of inventory exchanges made it very easy for networks to do their job, Tawakol says. This led to more ad networks entering the market. The creation of the data exchange made it easy for these ad networks to add behavioral and audience targeting, which is increasingly important. "The only problem is that it also made it easy for the end customer - the agency or marketer - to roll out their own capabilities," Tawakol adds. "This is a recipe for disintermediation. We already see an increase in demand directly from agencies who are rolling out their own internal ad networks. Having said that, smart networks provide a lot of value that will continue to keep them growing."

Not all networks need to go away, Tawakol notes. "There will be consolidation for sure, but there is an also an opportunity for evolution," he says. Ad networks serve two main purposes: arbitrage and the aggregation of audiences across multiple sources of inventory and data. "In addition, many networks differentiate themselves through high-touch service, low cost, and specialization within a vertical," Tawakol says. "As agencies try to roll out their own internal networks, they will find that doing all of the above is very hard. Demand-side platforms have emerged to help them do this."

A successful ad network can become a successful demand-side platform if it unbundles its offerings, Tawakol says. Some networks, for example, have excellent operational infrastructure an agency could benefit from. Other networks are very good at arbitrage and that is something many agencies won't want to do because it might make them appear misaligned with their marketer. "Demand-side platforms represent the next iteration of what a network could be," Tawakol says. "Many networks are figuring this out, but their current revenue streams are trapping them from adapting. Unfortunately, in a situation like this, if you aren't willing to cannibalize your own revenue, someone else will."

Context and context-specific buys - such as fantasy-league sponsorships or custom programs - aren't going away. "Our clients like them, they feel like they differentiate the brand in an impactful way and if you need to launch a campaign, launch a product, launch whatever, they're really meaningful," VivaKi's Hecht says.

Apprendi agrees. "Advertisers that want to buy sites like ESPN and sponsor the hockey section - that's where publishers are in a great position to not be disintermediated," he says. "The advertiser wants that brand. They want to do a fully integrated type of sponsorship or advertising package with a lot of bells and whistles. That's a distinct, differentiated and very specific initiative that a brand advertiser might want."

But if the brand advertiser really just wants a reliable way to reach a target audience independent of a full-on integrated buy with a publisher, that's where publishers have the greatest risk of being disintermediated, he says.

"Media in the middle," which is the age-gender-demo buy, is really under pressure, Hecht says. "So if that's the way you are still selling media and it's not really connected with the content - because when they say age-gender and demo I'm not really talking about any content, I'm just talking buying mass eyeballs - it has got to be more specific and more addressable. It's really incumbent upon the publishers to get to that next level of being specific on who exactly it is they are selling. And they can do that - they just need to take control of their destiny. If they take control of their destiny, they're fine." 

Next story loading loading..