Saying that search engines are not perfect is a bit like saying the sun rises every day: as if we didn't know. But there's oftentimes a tendency to think that Google is the be-all and end-all of search, when most content professionals know that oftentimes there are far better tools for specific purposes. A recent article by Mary Ellen Bates in EContent Magazine disses Google in favor of a new search engine from exalead, a French development team deeply steeped in search sciences. Exalead has a lot of very interesting state-of-the-art features to assist searchers. In honor of a French favorite I tried the term "baguette" as a search with interesting results: screen grab samples of pages, a "related terms" thesaurus that offers some interesting potential search refinements (French baguette? Toasted baguette? Baguette diamond? Sun-dried tomato?) as well as related categories, Web site locations and document types. Lots of "best practices" thinking is evidenced in the very friendly interface. But the search results themselves...?

First up out of 176,825 results on exalead is a site "baguette.com" that is just a domain name up for sale, second is an article in German on baguettes, third is an article on putting Stilton, an English cheese, on a baguette, fourth is an article on salade nicoise, fifth a baguette handbag, and so on. Why am I not impressed? On the other hand, Google's search on "baguette" turns up first out of 2,470,000 results a nifty article that defines and gives the history of baguettes with lots of facts and statistics. Second through fourth are popular restaurants and bakeries using the word, fifth is a site focusing on baguette diamonds. Google ads feature bakeries, diamond merchants and handbags. Try a similar search with the more American term "French bread" and exalead returns lots of recipes using baguettes, while Google returns great photos of French bread and a great article on French bread, along with the original article on baguettes down the page after a few recipes for baking and using French bread.

Exalead has a very well designed product whose strengths may become more evident once they crawl more content, but this little test is a simple example of how best practices in design don't always yield best results with search products. Search navigation tools remain very important, especially for finite collections of content that can be traversed more easily via a taxonomy. But in the relatively infinite environment of content on the open Web, search becomes less of a tool for researching possible answers from known high-quality sources and more of a tool to find out how other researchers have found the best "good enough" answer out of that infinite pool of possible answers. Put simply, most people don't like browsing and fiddling to get answers, no matter how sophisticated the interface. This is why the link evaluation methodology and other aspects of the Google search algorithm remain important factors for tuning us into the intelligent opinions of other searchers. No algorithm frozen in time is more intelligent than people who understand the context and use of content in its full dimensions that shift with the evolving human experience. It's a concept that powers everything from search engines to weblogs to stock rating services. Congratulations on exalead for a great Beta product - that hopefully becomes aligned with the content sets and users most suited for its strengths.