Editor’s Note: For WebmasterWorld’s PubCon 2006 conference in Las Vegas, a couple of our star performers here at Yahoo! Publisher Network, Senior Product Manager, Cody Simms, and Senior Product Manager, Amit Paunikar, put together a presentation on how publishers can optimize their web sites to help improve the relevancy of their contextual ads. It drew praise from all and sundry. For those of you who could not be at PubCon, we’ve turned the presentation into a four-part blog series. In the first chapter, we offer a little “Contextual Advertising 101″ to help get you grounded. In subsequent chapters, we’ll show you how you can optimize your pages to help improve relevancy.
 

What motivates publishers?

The anecdote is legendary. Orson Welles, while directing one of his films, told an actor to cross from one side of the set to the other. “What’s my motivation for this cross?” the actor wanted to know.

“Your paycheck on Friday!” growled Welles.

While running a network for publishers, we’ve learned that publishers come in all shapes and sizes, and are motivated to publish for all kinds of reasons—including their paychecks on Friday. Some of you enjoy the “work-at-home” lifestyle. Others like the community aspects of publishing. Some want to stay in touch with the technical innovations that online publishing is bringing to software, and enjoy the challenges of search engine optimization and monetization.

But whatever your motivation, two ultimate success measures stand out: revenue and traffic generation.

Keep your eyes on the prize, and optimize
Optimizing your site for contextual advertising can help you with both of these success measures. Contextual advertising crawlers and search engine crawlers are very similar, so some time and effort optimizing for a contextual crawler could help make your site more search engine-friendly, too.
 
So just what is contextual advertising?
But contextual advertising is not search advertising, and that’s the rub. In the absence of clear user intent—i.e., a user search for a specific product or service via a search engine—contextual advertising systems have to rely on secondary factors to glean what the user is looking for to serve up appropriate ad content. Users’ intentions can be derived from their behavior, their demographic information, or from the information available from the content of the pages they’re browsing. Demographic and behavioral information is hard to come by due to privacy and a host of other issues. Therefore, most contextual advertising networks rely heavily on semantic analysis of the content a given user is browsing—specifically, the content on your pages.

Contextual advertising is intended for content-oriented pages where users are passively consuming content (as opposed to search results pages where users have expressed a clear interest). Contextual advertising relies on one underlying major assumption: that a user is interested in the content of the page if he or she is reading that page. Contextual advertising therefore uses semantic analysis to determine the context of the page and then provide advertising of a similar topic.

Semantic analysis, you say?
Semantic analysis involves text mining, term/phrase extraction, SVMs, page vector analysis, associations and all that jazz. Um… what are these things, you may ask? In simpler terms, semantic analysis relies on good, clear, topic-oriented content on a well formatted page. We’ll come back to both of these ideas in the third post in this series.

The bottom line is that “content is the king” when it comes to getting relevant ads on a page. With a little attention and a bit of luck, you can improve your ads’ relevancy, along with that paycheck on Friday.

Next week: The anatomy of a webpage, site structure and bots.

—Cody Simms, Senior Product Manager, and Amit Paunikar, Senior Product Manager