To Be or Not to Be an Online Adby
The online advertising industry—with all its data on who people are and what they like to buy—is a sophisticated animal … sometimes. At other times, it looks an awful lot like a gangly adolescent trying to figure out its place in the world.
This week, for example, I was told that some folks in the industry had reached an unofficial consensus around what it takes for an online ad to be “viewed.” Put another way, they’ve figured out what an ad has to accomplish before someone leaves a Web page in order for it actually to be considered an ad with a chance of influencing a person. The answer: At least 50 percent of the ad must be visible on a person’s screen for at least 1 second. (This does not strike me as a particularly high bar. A true, standardized metric is expected to arrive from the advertising industry next year, which may be more demanding.)
It turns out—and this may surprise you—that advertisers do care if their ads are actually appearing in front of paying customers. Earlier this year, ComScore released the results of a study of 12 big-name advertisers in which it found that about 30 percent of the online ads purchased by these companies never reached an eyeball. “Just because an ad is served doesn’t mean that it shows up on the screen,” says Anne Hunter, ComScore’s senior vice president for ad effectiveness. “It may end up too far down the page for someone to see, or someone might go to read a new Web page before the ad actually loads.”
ComScore loves to measure things and has spent a good chunk of this year hawking its “validated Campaign Essentials” tool, which includes technology that can determine if an ad—as defined above—did make its way onto a page and appear in front of a human. “There is a movement in the advertising industry to provide a lot more transparency about what is actually happening,” Hunter says.
While some ad brokers may be displeased to see the introduction of nosy technology that wants to count whether or not an ad is delivered, others have welcomed the technology. This week, VivaKi, which helps place ads for a number of major brands, revealed a partnership with ComScore to create a system that will measure which sites do the best at meeting the ad-delivery criteria.
Companies that perform so-called real-time bidding already run millions of calculations every time a Web page loads to try to figure out which ad is best suited for you. They base the price of ads on the strength of this match and can charge more when they get more people from a desired group to click on an ad or take some action. So the deal between ComScore and VivaKi points to a new bidding metric being introduced that will examine the ability of sites such as yahoo.com and nytimes.com to hold peoples’ attention for that magical 1 second. “Some might see fear and danger here, but we are all for any metric or methodology that makes us smarter about what drives an ad campaign’s performance and makes us a better steward of our advertisers’ dollars,” says Mike Zeman, a vice president in VivaKi’s research and development arm known as the Nerve Center.
For all the computer science behind ad technology, it’s somewhat surprising that the industry hasn’t already sorted out matters like these. After all, the big edge that online ads have over television and print ads is the data that can be brought to bear to prove their effectiveness. Yet here we are, still trying to figure out something as basic as whether or not an ad has actually been displayed and what you can charge for that extra bit of insight. All part of growing up, I suppose.