You Still Google? That is So Last Week

Whenever I give a workshop or presentation, I ask the people attending to raise their hands if Google is their default search engine. I am always somewhat surprised that at least 80% of the audience still start their search with Google, although it is heartening to see that at least half of librarians and info pros say they routinely use more than one search engine. (And, for the record, my default search engine this week is Yahoo!, although I use Firefox's pull-down menu to change my toolbar default search engine regularly.)

What I find distressing about this is that most people still consider Google to be the gold standard of search engines. Yes, I like Google—its Map feature works well, Google Scholar is sometimes a quick way to find a copy of an article, and its synonym feature using the tilde (~word) comes in handy. But the backbone of Google's search-results sorting is still Page Rank, and, frankly, it's not keeping up with the content available on the Web. Often, finding the most popular sites isn't what I want; I need to find the most authoritative sites or ones that were evaluated by experts, or I need to look at just one aspect of a topic.

This is one reason I'm enamored of Exalead, a still-in-beta search engine that has advanced search features to die for ("sounds like," approximate spelling, and proximity search, for example). Other Google competitors, such as Vivisimo and Ask Jeeves, offer clustering of search results, which helps conceptually narrow your search. And some non-Google search engines are much better at providing answers, not just organic search results, at the top of the results page. Type "Bush" in Ask Jeeves' search box, and the first result is a biography of President George W. Bush. Both Yahoo! and A9 search database content along with the open Web. Teoma identifies link-rich sites that are likely to be useful guides to the best sources on a topic.

But it's Google's linear presentation of results that I find so, well, 2001. And speaking of 2001, I've watched the movie with the same title every five years since 1968. In the early 1990s, it finally felt really dated: HAL9000 is a mainframe computer? I get that same sense as I look at Google's user interface, after having used some of the graphical tools available in other search engines. MSN's Search Builder provides sliders for you to specify how important it is to have "popular" pages and whether you want exact matches for all your search terms or a more approximate match. Yahoo! just rolled out its Mindset tool, which lets you indicate whether your inquiry is more related to shopping or to research (, and its delightful SmartSort has been available to shoppers for two years now. In addition, data visualization tools like Inxight, i2's Anacubis, and KartOO let us see search results in new ways, and detect relationships we would not have discerned from a plain linear display of information.

Because many of our clients use only Google, they are actually being left behind in terms of search technology. Google is dumb: it places so much trust on its relevance ranking in its presentation of search results as a simple list of Web sites. Users don't have access to suggestions of alternate concepts or terms or all the tools that other search engines provide. Our clients are placing too much reliance on the first ten results they get from what they consider to be the best search engine out there. It reminds me of what happened when we Baby Boomers, raised on Chef Boy-Ar-Dee Beefaroni and Jell-O corned beef salad loaf, finally encountered—you know—real food. "Wow, you mean we can fix food that has real taste and texture?"

What this means for us info pros is that, when we introduce our clients to all the sophistication of a high-end online service or enterprise search tool, we have to remember that they often do not have any context within which to evaluate it. They are accustomed to looking at search results that were cutting edge three years ago. The new search tools that are available do require more work for the user. Rather than just rely on the first page of search results, you are encouraged to look at some of the suggested modifications. Do you want to narrow your search of "solar energy" to "photovoltaic cells"? Would you like to limit your search to the appropriate category within the Open Directory Project? Would it be useful to just start with a few good Web-liographies compiled by nonprofit organizations? Or perhaps would you like to tweak just how crucial each aspect of your search is?

Of course, I am still waiting to see these features made available on the value-added online services, with the exception of Factiva's Start Page, which does quite a bit of behind-the-scenes filtering of results based on what it knows about the searcher. At least now the scales have fallen from our eyes. We know we can expect better. No more Beefaroni—we want slider bars!