If you think about it, Google, with its $82 billion market cap, offers just a hint of the potential for online advertising.
The average web surfer spends less than 5 percent of his time using a search engine, according to the Online Publishers Association's Internet Activity Index. That means Google earns almost $3 billion a year from people who devote 95 percent of their time on the internet to doing something else.
Media Hack
Yet Google has been able to achieve dot-com nirvana by perfecting the art of targeted advertising. If I search for, say, "hemp fabrics," I'm greeted by a list of 10 results framed by Sponsored Links at the top and down the side of the page. In essence, I get two types of results: those that are the most "relevant" (read: popular), which aren't for sale, and those that advertisers bid on, which are.
Since a consumer who searches for a product or service is more likely to buy something, this type of advertising is far more powerful than the usual scattershot method of firing ads at us in the off chance one will catch our attention. It's the main reason that paid search makes up more than half of Google's annual revenue, which could hit $6 billion by the end of 2005.
Of course, Google has other ways of making money, namely via AdSense, which generates the other half of its revenues. Google serves up "contextual ads" tailored to whatever a web page is about. But these types of ads are inherently weaker than paid search ads. Just because I check out the weather online doesn't mean I have any interest in ads for Doppler radars.
In fact, of the four distinct activities that people pursue online -- content, communications, commerce and search -- OPA estimates that the average webizen spends about 41 percent of his time checking e-mail and instant messages, 37 percent viewing news and entertainment sites and a little more than 17 percent shopping.
While search and shopping offer the biggest bang for an advertiser's buck, onliners while away many more hours -- 78 percent of their time to be exact -- communicating and reading. Naturally, Google thought of that, too, embedding ads based on content in Gmail, but it's questionable how effective they are. (Google News, still in beta four years after launch, has no ads, largely due to copyright issues.
This is where Roy Shkedi, an engineer and chief executive officer of New York behavioral-marketing firm AlmondNet, comes in. He has patented an idea that he believes could bring tightly targeted advertising to the remaining 80 percent of our time online when we are not searching or shopping.
In the same way that Priceline.com's Jay Walker patented "name your own price" and Overture did paid search (it cost Google $300 million to settle the dispute), Shkedi has claimed as his own the idea of creating an ad network of web publishers that share readers and in the process reap commissions.
Or, as he put it in his patent (entitled "Super Saturation Method for Information-Media," US 6,832,207): "The present invention generally relates to a method for distributing information-media contacts. More specifically (it) relates to expanding the revenue from information capacity of the media."
Here's how Shkedi's concept, which is in beta, would work: Shkedi estimates that 80 percent of the ad space on the web is either unsold or goes for cheap -- something like 15 to 20 cents CPM (cost per thousand impressions, a common metric for advertising revenue). So it's not unusual for a person to visit a site like InfoWorld that charges a premium for advertising -- $102 CPM for a banner ad -- and bounce to a site that gets only 20 cents CPM -- or 1/500th that rate.
From $100 to 20 cents within five clicks. So the person who was worth denting your ad budget for on one site is essentially worthless on another. This, when you think about it, makes little sense. The trick would be to flash relevant ads at people wherever they go on the internet; otherwise, online advertising is like throwing darts in the dark.
"If someone searches for car insurance, he should see car insurance ads when he accesses the weather and sports, or visits InfoWorld or checks web-based e-mail," Shkedi said.
It's a concept that failed in the past. Behavioral marketing had a lot of buzz in the late 1990s, but never lived up to its hype. Publishers like InfoWorld had little incentive to identify readers who left their sites. But if they were compensated, Shkedi believes they would gladly participate because it would open up a completely new revenue stream.
"A company would pay for the right to cookie InfoWorld readers," Shkedi said. "If any of them clicked over, it would pay InfoWorld a commission."
Besides convincing publishers to play along, Shkedi's biggest obstacle might be an indelible aversion to cookies, which many view as a threat to privacy. According to a March 2005 study conducted by JupiterResearch, almost 40 percent of web users delete cookies at least once a month, while 17 percent get rid of them weekly and 10 percent erase them every day.
But Shkedi, who received a prize from Israeli intelligence for a mysterious R&D project he can't talk about, claims he isn't worried.
"Most people don't erase their cookies," he said, "and deleting them doesn't have much of an impact over two weeks. If you haven't searched for something or visited an online publisher in over two weeks, you're not really interested in it."
- - -
Adam L. Penenberg is an assistant professor at New York University and the assistant director of the business and economic reporting program in the school's department of journalism.
