Why Silicon Valley Can't Sell

Or, why luring big brands to platforms is its next big problem

Big brands continue to spend relatively insignificant parts of their ad budgets on digital: In 2009, Procter & Gamble, for example, devoted only 7 percent of its ad budget there; at Unilever it was 3 percent, according to IAB figures. In P&G’s case, that percentage actually declined in 2010, by Adweek estimates, to 5 percent, even as the total dollars tripled to $169 million. Meanwhile, Unilever CMO Keith Weed claimed at Cannes last year that the company would merely double its digital budget in 2010.

One big reason for the relatively small digital ad spend is that brand managers don’t want their ads caught anywhere near much of the existing online inventory, derived as it is from the direct marketing world of junk mail (the ubiquitous banners, for example). “We’ve got decades and decades of experience that says building my brand drives sales—and all that direct response and discounting does is drive my brand into the toilet,” says one former brand manager.

And yet it’s their exploitation of all that direct response inventory that has Google, Facebook, Yahoo, and Microsoft (which is a Silicon Valley company in temperament if not in physical locale), with their vast scale, pulling away from the rest of the competition online—together they’re expected to grab some 65 percent of the digital ad market in 2011, according to eMarketer, and as much as 70 percent in 2012. It’s the ad sales equivalent of rent-seeking, on inventory that is often the digital equivalent of a tenement slum (and, as the CPMs demonstrate, priced accordingly).

All that excess inventory keeps prices down at the expense of just about everybody producing content—but ultimately, at the tech companies’ own expense, as well. It’s what’s fueled the growth of content farm companies like and Demand Media, for example, whose response to low CPMs has been to produce more low-quality, low-cost pages—pages that actually reduced the value of Google’s own search results.

A growing proportion of these tenement transactions are taking place via the dizzying array of automated services—exchanges, demand-side platforms, supply-side platforms, agency trading desks, real-time bidding—that have turned much of the “tonnage” referred to by Rothenberg into a low-margin commodity business. Digital advertising people these days often sound more like Wall Street quants than Mamet pitchmen.

The ad agency holding companies, used to getting their piece of media buying through the traditional firms that they long ago split off from their creative agencies, have invented their own quant systems to trade with the supply-side platforms that Google and the others have devised (and, perhaps, to hedge their bets about their own inability to deliver for their brand clients in digital). Sorrell’s WPP, for example, added the latest iteration to this in late June when it unveiled Xaxis, a subsidiary intended to buy digital audiences in bulk.

David Kenny, the since-departed founder of VivaKi, Publicis’ digital arm, which launched the first agency trading operation, acknowledges the need to create better context for brand advertisers, but is unapologetic about the commoditizing effects of the trading platforms decried by Rothenberg. “Content commoditized awhile ago,” he says. “I don’t think the [agency] platforms created a Demand Media. And if the media has become commoditized, all a trading platform does is expose it much more quickly.”

Arora, who manages revenue operations at Google—which took in more than half of the available online ad money last year through its mastery of these automated systems—turns almost gleeful when discussing the opportunities in disaggregating audiences from content and indeed context. “We’re selling network,” Arora says of the 2 million Web publishers Google reps via AdSense. “To me, that’s a longer sustainable business, because I can build a network based on how many publishers I can acquire [as Google clients].” If Google doesn’t have enough men 18 to 24, for example, it just expands the pool of sites that draw them.

Only a fool, Arora seems to suggest, would want to be caught in the business of trying to attract an audience, the way his competitor Yahoo does just down the road in Sunnyvale, Calif. Compared with Google’s network scheme, in which it simply adds website publishing partners whenever it needs more audience, “Yahoo can’t grow,” Arora explains, leaning forward on a couch in his Mountain View, Calif., office and bouncing a small toy ball. “Their media business is going to be dependent on the number of viewers of their content properties. You can tell me how they’re going to fare,” he says, dismissing the possibility of a positive answer with a laugh.

Of course, both Yahoo and Google have hedged their bets here: Yahoo is now trying to develop its own network of websites through an affiliate sales organization; it’s also gotten into the content-farm game with the purchase of Associated Content. And Google itself appears to recognize the need to create owned-and-operated premium spaces for advertisers; witness its recent additions of Hollywood content to YouTube and reported pursuit of Hulu.

Beyond Yahoo, others representing digital companies who accept the media label (particularly those grown out of traditional media companies) dismiss an overreliance on automated trading schemes to sell long-tail inventory as quickly as Arora dismisses Yahoo’s reliance on its own content. The skills required to make a market for stocks or bonds or pork bellies, they suggest, aren’t the same as those required to build brands. “It’s either about trading commodities, which means the people most adept at trading those commodities will win, or it’s about strategies and client relationships,” is how one digital media executive puts it.

Curt Hecht, Kenny’s former colleague who oversees the automated ad trading services of Publicis’ digital properties through its VivaKi Nerve Center, agrees that ad platforms won’t ever be enough for the ad world’s most important clients: “I think on the branded side of it, when it comes to building ideas and experiences, I don’t think that can ever be standardized. I think we do want a level of customization for it.”

Rothenberg is even more pointed. “If the economy were based solely on efficiency, we’d have to bleed all desire out of it,” he says. Responding to desire, Rothenberg suggests—and more importantly, prompting it—is what brand advertising has always been about, and the media business has always been about creating a place for that kind of advertising to work most effectively. It’s the difference between a low-margin business and a high-margin one.

To different degrees, Silicon Valley players—even those protesting otherwise—understand the need to create spaces apart from those managed by automated selling tools, so they can give brand advertisers a chance to stimulate desire. What remains in question is their ability to deliver on it.

Desire demands storytelling, which is why brand advertising has always been intimately involved in the creation of mood, emotion, character, look, and feel. The people who run digital companies for the most part don’t have those talents (except to the extent look and feel is required in a software interface), and aren’t interested in them—and only recently have they started to recognize the need to add employees who do.

“It’s really about you and I remembering that awesome Nike ad from three years ago,” says Neal Mohan, who oversees Google’s display business, about the emotional impact a brand can have. But Mohan also notes that creating those sorts of experiences is still far too expensive in digital, with as much as 28 cents of every dollar spent online lost to the costs of actually producing ads, versus 2 cents in television, leaving much less of the relative budget for buying eyeballs.

Continue to next page →