SUBSCRIBE NOW!

Vol. 21 No. 4 — April 2004

When love is shiny and new, you perceive your beloved to be surrounded by an otherworldly aura, devoid of imperfections. Forsaking all others, you cannot get enough of this dazzling newness. But as most of us know, once the affair is no longer fresh and dewy, there exists the distinct possibility for disillusionment. Maybe you start wondering how you didn't notice certain things—things that are now beginning to drive you crazy.

Breaking Up with Google?

All I have to say is be wary of falling in love with a search engine. You're likely setting yourself up for a fall. Of course, I'm talking about Google here. People get emotionally involved with Google.

In 1998, Google was still in beta and handling just 10,000 queries a day. Its innovative approach to link analysis was attracting the attention of not only potential investors, but also search professionals. A lot of us were totally sucked in. Hey, was it good for you too, baby?

Google came out of beta in September 1999. By the end of 2000, it was handling more than 100 million searches a day. Bells and whistles were added. APIs were released so creative souls could play. An innovative self-serve AdWords program began to revolutionize the way advertising was presented and sold on the Web.

Google now processes upward of 200 million queries a day. But is it still good for you, baby? Some of us are beginning to have our doubts. Yes, there's continuing Google hype in the mainstream media, and the average Internet user is still besotted. But those of us who search for a living are well into the process of perceiving and documenting Google's flaws—some minor, some relatively significant.

Certain issues are largely a function of Google's popularity. For way too many Internet users, if what they're looking for doesn't appear in the first page or two of results after a simple keyword search, it may as well not exist. It would never occur to many, if not most, of these seekers to take advantage of Google's advanced features or—heaven forbid—try their queries on another search engine.

Many search experts and pundits feel that Google may have hit the wall with its link-analysis technology. The search engine optimization industry has it pretty well figured out, which is why so many really mediocre and downright crummy links ooze their way to the top of your search results. You've run into these: spam pages that are basically nothing more than link farms. And then there's "Google bombing": manipulation, according to Danny Sullivan of SearchEngineWatch, "where Web sites are influencing Google's search results by controlling where they link to and what they say in their links."

And how many times has your search pulled up scores of duplicate results from the same site or database? Ugh!

Gary Price, my guru and colleague at ResourceShelf.com, has identified other problems with Google, including the following:

• Advanced search features that don't work or don't work properly, such as the Boolean OR

• Inaccurate page-count estimates

• The fact that Google indexes only the first 100 KB of any document (120 KB for Adobe PDFs). If what you need is beyond this arbitrary limit—which, according to Price, is never stated by Google—you won't find it.

• Announcements of new features that either don't work right when rolled out (airline flight searches) or more or less never get off the ground (Google crawling the abstracts of IEEE publications)

A librarian could help, of course. Most of us work diligently to educate our customers. We can show them how to use Google more effectively or introduce them to other search tools and—often better—specialized databases, either free or subscription-based. Many of the latter are paid for by our customers through their taxes but remain sorely underutilized due to lack of knowledge.

We all know that the big problem is getting them to come to us in the first place.

Back Together Again?

Every once in a while, however, we have a magic moment. Recently, a customer approached me with a citation for an article published in a journal that dealt largely with counterterrorism. (Such is life at a military base.) "I checked Google," she said, "but it wasn't in there."

I did not check Google—at least not right off the bat. I checked a customized military and intelligence database we get from Gale/InfoTrac. The good news: Yes, this publication was available there in full text. The bad news: In the database, there's a lag in the availability of full text from this publication. And the article the customer wanted, published in January 2004, was not online yet.

The standard document delivery channels we use for such requests would not get the article to the customer soon enough. (They always need it yesterday. Or is it just my customers?) Since I've indeed been around the block a few times, I've been known to engage in back-channel "guerrilla" interlibrary loan with colleagues who have what I don't. I told the customer I might be able to get the article the following day. This was OK with her.

As she walked away, I looked again at the citation she had left me. The author's last name was unusual—a long, Eastern European name. On a whim, I punched this into Google.

The article she was looking for had been presented as a conference paper. The entire conference proceeding had been placed online in full text by the City University of New York. (Thank you very much.) I bolted from behind my desk and tracked down the customer in our parking lot. I was able to catch her before she drove away. She was thrilled, especially since she could read other, related papers as well.

Sometimes, baby, it's still so-o-o-o-o good!