YouTube's Buried Treasure

Making the most of its riches will take a lot longer than Google expected

In mid-April I spent some time at Google and YouTube. Even this far into Google's remarkable run, there's nothing quite like a sunny spring day at the Googleplex to make a print media guy's shoulders slump; to make that guy feel like a third-term Democratic senator running for reelection in the fall of 1980, just as the implications of the phrase "The Reagan Revolution" were becoming brutally clear. That's not to say everything's golden at Google, which last year took in $16.6 billion in revenue. Its great ad successes--AdSense, which places text ads across several hundred thousand partner sites, and AdWords, which displays text ads beside search results--are both simple, easily automated, and can scale to just about infinity. (Practically anyone can afford such ads and create them with a few snippets of text; you can sell truckloads without a sales force.)

Video ads on YouTube, though, do not yet lend themselves to easy automation. They're also more expensive and still primarily the province of big-name advertisers. Selling to them nearly always requires the participation of an ad agency and calls for more labor-intensive sales methods. There are high-level concerns inside Google that the excitement around YouTube--which continues to increase its share of the Web video universe--isn't readily translating into sales and ad dollars.

"It takes longer to bring in a YouTube dollar than it does to bring in a search dollar," concedes Tim Armstrong, Google's top U.S. ad-sales executive. "Can you make [that process] more efficient? We think "yes.'" He adds: "If you talked to me" about this in early '08, "I'd have been more anxious. But we're making nice progress."

Others wonder. The most successful ad format for Web video sites has been display ads that run near video clips, says Dave Morgan, founder of ad network Tacoda, which placed ads on YouTube before the Google deal. "No one is able to sell [these] display ads well in an automated way," he says. "It requires a human sales force."

Of course, Google and YouTube have a massive human sales force. And, like all big online players, Google has invested serious time and dollars (the $3.1 billion DoubleClick deal, for instance) in beefing up its display-ad capabilities.

YouTubers point out that most of their key ad formats--including the "overlay" format, which places an ad over a small portion of a video without interrupting it--have been available only since August. And, in a corollary to the saying at Google that warns "don't bet against the Internet," I wouldn't bet against Google. But it will take time, and perhaps more than the company first reckoned. (A JPMorgan Chase analyst estimated recently that YouTube will gross around $100 million this year, with the vast majority coming from display ads.)

There is a delicate dance between users' expectations and advertisers' desires for a site as naturally woolly as YouTube. Its tens of millions of users expect a degree of graininess, a lo-fi feel to what they see there. (Interestingly, internal research suggests that ultra-glossy ads on its front page tend to get tuned out.) And the site's free-form format, wherein user-generated videos ranging from the innocuous to the stomach-turning still mingle with copyrighted content, makes many uneasy. "Clients are concerned about adjacency to inappropriate content," says Mohan Renganathan, a vice-president at media buyer MediaVest. YouTube "is a haven for things that are not appropriate" for top brands.

There are also factors within the greater media ecosystem at play. At ad agencies, Web video ads can be subject to tug-of-wars and turf battles between TV buyers and digital buyers. An additional perversity of putting video ads on the Web: It can require agencies to take their most prized assets--the cinematic 30-second TV spot--and recut them for shorter attention spans, if not outright start over from scratch.

There are ways around these issues. YouTube is holding contests in which users upload videos to a sponsor's page and thus, ideally, generate heat and page views for ad messages. (Among others, Toyota has sponsored multiple comedy-related YouTube contests.)

There's nothing wrong with this idea, and others for "safe" advertiser sites within YouTube. But they're not a new notion that will set the world ablaze, and some ad execs sound lukewarm. "On balance, I haven't been impressed" with YouTube, shrugs one digital media buyer. "Because what they're offering is basically another venue for a micro-site"--Web pages within a larger site--"on steroids." (Such micro-sites can be had with a minimum YouTube ad buy of around $200,000.) From what I saw at YouTube, the most interesting new ad format uses an algorithm to select videos from known YouTube contributors--be they cable networks or homegrown stars--to surround an ad on a dedicated advertiser page. This capitalizes on both YouTube's vast programming archive and Google's technical underpinnings.

I'm still not convinced that this can scale to Google-esque proportions, since these still can't be automated as simply and elegantly as AdWords and AdSense. All the video, all the users, and all the data YouTube can claim as the Web's biggest video depository means there's an awful lot of treasure buried there. I'm just not sure how fast Google--or anyone, for that matter--can surface it.