• by Chris Copeland , April 12, 2006

For ten years the search industry has been built on the concept of keywords. What began in organic search has multiplied many times over in paid search. To many, this means no list of terms is ever too long, and no word ever irrelevant. Why? Because the advertiser only pays when someone clicks, and by clicking, consumers declare that a listing is relevant to their query.

But what if their click was not a guarantee of relevancy? What if the consumer had declared marginal interest through the lure of minimal text listings, and not a burning desire to become a brand customer and loyalist? What if the notion of search being a free lunch with costs tied only to interested clickers was, in fact, a myth perpetuated by the engines to bolster revenues?

A new study from France notes that Google and Yahoo score virtually the same in the all- important metric of relevance. The study further states the relevancy of search today isn't very good.

Google built its reputation on the simplicity and relevancy of its results. Users of search are constantly seeking more relevant information in a faster delivery model. For a study to show that as the market continues to grow in size, the quality of its return product is not increasingly better, is reason alone to give pause.

What if the problem wasn't at all related to search algorithms, page rank or the pecking order of placement, but rather to the way we think about search programs?

Search marketers espouse the value of "the tail" ad nauseam. They contend the bigger the keyword list--20,000 terms, 100,000 terms or 1 million terms--the more success brands are likely to have.

But what if this is a fallacy perpetuated by bid engines and bid tool makers alike? What if the tail in and of itself was not the be-all or the end-all?

What if, in fact, the answer to relevant search results lies in the searcher? Instead of a search term being a billboard for a company, it becomes a conduit between user intent and advertiser content. If so, then a new world of behavioral learnings begins to emerge, whereby keyword searches begin to classify users and thesuddenly not-so-random way people behave in their searching.

What if it was from consumers' behavior that learnings could be derived, segmentation established, and keyword and content generated in order to be successful?

What if?