Patri Friedman points
to a fascinating post by a Yahoo employee (speaking only for himself) speculating on the reasons Google is clobbering Yahoo
in the search ad market. In a nutshell, Google was a lot quicker to figure out the benefits of ranking ads by ad quality rather than simply auctioning off the top slot to the highest bidder. Given that online advertisers pay on a per-click basis, more relevant and useful ads can generate more clicks -- and therefore more revenue -- than lower-quality ads. Because ads with higher click-through rates became more highly ranked, advertisers began to compete on relevance as well as price. They began to optimize their ads to generate higher click-through rates. The average quality of ads on Google began to improve. And here's the really important point: as the quality of Google ads got better, users started to discover that Google ads were actually useful and relevant
, and they got in the habit of looking at them. This is an example of a principle Techdirt has been emphasizing for years: ads are content
, and they're a lot more effective if they contain information people actually want. Google's experience belies the conventional view that ads are a necessary evil users have to put up with as the cost of getting the content they want.
Nowhere is that conventional view of advertising more entrenched than in the television industry, which is constantly wringing its hands (and filing lawsuits) over the detrimental impact of devices like the TiVo and Replay TV that include ad-skipping technology. I think the Google example demonstrates how short-sighted that attitude is. With a little ingenuity, TV networks could be using devices like TiVo the same way Google uses click-through statistics: as a way to gather data on user attitudes toward different ads. If networks priced ad inventory the same way Google does, giving a discount to advertisers whose ads had lower skip rates, advertisers would respond by trying to make more entertaining and relevant ads. And as ads became more entertaining and useful, viewers would be less likely to pick up the remote and push the "30-second skip" button.
Even more radical, the networks could be using TiVo-like devices to distribute shows and ads directly over the Internet. In that case, the device could display a different set of ads to each viewer, with the ads chosen based on the individual viewer's show-watching and ad-skipping history as well as some basic demographic characteristics. For example, users who frequently skip car ads would be shown fewer car ads. Viewers under 40 would never be shown ads for adult diapers, and all-male households would never be shown ads for feminine hygiene products. Such a system would be a win-win for both advertisers and viewers: viewers would find ads more useful and less irritating, while advertisers would be willing to pay higher rates for ads that were precisely targeted at relevant subgroups. And that would solve the "TiVo problem" once and for all: not by forcing users to watch ads they'd rather avoid, but by finding ways to show users ads they actually find entertaining and useful.