Online KMWorld CRM Media Streaming Media Faulkner Speech Technology Unisphere/DBTA
Other ITI Websites
American Library Directory Boardwalk Empire Database Trends and Applications DestinationCRM Faulkner Information Services Fulltext Sources Online InfoToday Europe KMWorld Literary Market Place Plexus Publishing Smart Customer Service Speech Technology Streaming Media Streaming Media Europe Streaming Media Producer Unisphere Research



Magazines > Computers in Libraries > October 2003
Back Index Forward
 




SUBSCRIBE NOW!
Vol. 23 No. 9 — October 2003
Feature
The Librarian Is In and Online
by Jessamyn West

I am a librarian, first and foremost, but I always appreciate an opportunity to answer anyone's questions, especially if it pays. How many times have we wished we could put a tip jar out on the reference desk to supplement our income if we gave good reference? The plethora of fee-based question-answering services seem to be trying to do just that: take a service that is available for free and resell it "value added" to a public that is just looking for more things to click 'n buy. But are they really any sort of threat to traditional reference and librarianship in general? And if so, why and how? And what can librarians do about it?

Many of these nonlibrary services seem somewhat like Lucy Van Pelt's homemade booth in the "Peanuts" cartoon strip, where for 5 cents you could get "psychiatric advice." Sure, the price is right, but do you really want medical help from a 6-year-old? Such is the quandary of online commerce, where you never quite know who or what is behind that remote keyboard. Question-answering services didn't get a lot of attention until major player Google threw its hat into the ring with Google Answers, recently out of beta. It wasn't long before hand-wringers in the media began to once again bemoan the supplanting of library services with inferior fee-based alternatives.

After a lot of hubbub, the peanut gallery, librarians, and pundits alike started to settle down and arrive at a few similar conclusions, and they're not all bad. While I've always personally opted for a book over an e-book, and I still type letters on a manual typewriter, I think there are solid non-Luddite reasons to pick the library over what's behind terminal number three. They're not "talking points" exactly, but I feel that by highlighting the differences between the jobs of librarians and online question-answerers, we can perhaps put the matter to rest once and for all. Here, in my opinion, are the three big differences:

1. Google Answers researchers are generally not librarians and rarely pretend to be.

2. Question-answering is not the same as reference work, although it shares many similar trappings.

3. There are always people who prefer to pay for convenience.

In the past few years, many libraries have started offering their own "ask a librarian" online services, making use of real-time chat, e-mail, and other third-party software to provide their own free answers to patrons' questions. As someone who has been on both sides of these interactions—as a librarian and as a Google Answers researcher—I wanted to look a little more in-depth at what we are gaining, and losing, when we conflate answering a question with reference work, and what we can learn by observing the differences in the two systems.

1. The 'Doctor' Is In: Researchers Are Not the Same as Librarians

These days, I get fewer blank stares when I talk about library school than I did 8 years ago, but that doesn't mean people really "get" the reference thing. I explain it to my friends like this: This is a question, but not a reference question: "Do you guys have any information on caves?" And this is a reference question: "I am trying to find information on those sightless fish that live in caves. I would like a book for my 10th graders to read." It's the librarian's job to turn the first type of question into the second. The fact that we as librarians will also tell you what time it is or where the bathroom is does not mean that we're not doing some serious question alchemy to help you find most things. The best reference interactions are ones in which the patrons find what they want and are not even aware that the librarians have been giving them reference interviews the entire time.

Google Answers does not have a built-in reference interview mechanism in its Q&A technology. Once a user—let's call him Linus—asks a question, researchers vie to claim or "lock" the question, getting the sole opportunity to answer it, for 2 hours. Once it has been locked by a researcher, she can ask Linus clarifying questions. However, since a question can only be locked for 2 hours at a time, and Linus might go offline after asking his question, time often runs out. Linus has no personal information about the researcher and she has none about him except what he's given in the question. The question itself has a time limit—often a week—which can also lead to questions dying on the vine.

So Linus asks his question and crosses his fingers that it will be understood. Or, as I saw happen when I was a researcher, Linus asks his question and then gets indignant when the question that was answered was not the one he wanted answered. (He had anticipated a certain type of answer but was not clear enough about it in his question.) The debate then becomes whether Linus pays or doesn't pay, rates his researcher high or low, resubmits his question or doesn't, and so forth. Chances of an interaction with the same researcher in the future are slim to none. I secretly suspect this encourages rudeness; others have agreed. This contrasts strongly with the library, where you can be e-mailed, called, or even answered right on the spot, but if you insult the librarian, it is likely the last question you will get to ask her. What's more, you can provide instant feedback to your librarian if you don't get an answer you like, and you can generally get it addressed again for free and without hurt feelings.

There is one particular value-added service that Google Answers offers, and that's the ability for others with Google logins to comment on questions. These are not part of the fee structure—you can collect all the comments you want for free—though once your question gets an answer from a researcher, you'd better be prepared to pay up. When I was a researcher, this was a little annoying since there were lucrative questions that typically got all-but-answered in the comments section before a researcher could get to them. Commenters hope that they will be noticed by the Powers That Be at Google and maybe someday get asked to be a researcher. I likened it to having patrons in line at the reference desk answering each other's questions. And yet, on the reference desk, we want questions answered, and answered well. If the third person back in line happened to be an expert on blind cave fish, why shouldn't she help patron number one? The spirit is one of cooperation in the library, not competition in the marketplace.

Lessons I Learned

• Make sure people are aware of your skills and education.

• If you positively must answer "I don't know," explain why.

• Make every reference interaction a learning experience for the patron.

• If outside help is available, solicit it and wrap it into the answer you give.

• Have a feedback loop for improving your reference abilities as you help patrons improve their questions.

2. A Dictionary and a Handmade Sign Do Not a Library Make: Question-Answering Is Not the Same as Reference Work

So, Google Answers workers aren't librarians, at least mostly. What kinds of resources do they have available? When I started working at Google Answers in May 2002, I found that I would skip questions that involved phone calls because my long distance plan was expensive and because I wasn't allowed to tell people on the phone that I worked for Google, despite the fact that I clearly, in some capacity, did. Researchers willing to go the extra distance with phone calls were some of the earlier successes. Many of the questions we got required authoritative answers almost as much as they needed to be technically correct, so getting the right person or reference to support your answer was always important. Interestingly, saying that a librarian helped you answer a question was often enough to meet this "authority" standard, begging the question of why people weren't going to the library in the first place.

We were encouraged to cite Google in our answers, specifically searches that would help folks like Linus learn to use Google to get answers. While this was a nice idea, I found that only a small subset of questions could be answered just by doing a more sophisticated search on Google. Since Linus and the gang had come through Google to get to us, we had to presume they had already tried Google once. Of course, some people just wanted to get someone else to search Google for them, but during my tenure, a lot of the requests seemed to be somewhat more technical or businesslike in nature. When I locked a question asking "How many streams are there on the Internet?" (inquiring about real audio and MP3 streaming technology), it was clear that the best answer to this was a market research report that would cost a fair amount of money. I referred it to my friend who is a manager at Real Networks. He explained to me why the question was unanswerable and then I was in the awkward position of citing this e-mail as an "answer" to the question, which it sort of was, skipping the question, or dropping the e-mail into a comment. It was a $50 question; what would you have done?

Then there were those questions that are just not really questions, much less reference questions. My favorite question in an "I can't believe I got paid for this" way was someone asking us to tell him a joke he hadn't heard before. References used: none. Time taken to answer the question: the 15 minutes it took to type it. Money made: $3.18. Answering a question like this, or my other favorite, "I like girls, do you like girls?" required a specific kind of skill. I had to appear to answer an answerless question, possibly be charming and amusing, and all the while not let on that the question was ... sort of stupid. I know we librarians say that there are no stupid questions, but I believe there might be stupid question-askers. In the library, questions don't get answered when they're out of the scope of the library's mission, when they are unanswerable, or when there's not enough time to answer them, not just when they don't pay enough. There's an inherent equality to this that pleases me.

Lessons I Learned

• Use offline and online sources.

• Teach patrons about the invisible Web.

• Highlight other ways patrons can use the library resources besides in person (e-mail, phone, ask-a services).

• Don't always start with Google; there are many ways into the Web.

3. Why Pay 5 Cents When It's Free Down the Street? Some People Will Always Want to Pay for Convenience

The most important thing to remember is that your library can't be all things to all people, nor should it be. I wince at the library trade magazine ads that show a drive-through window and the phrase "Do you want fries with that?" If the patrons want fries, they should be in a restaurant. It's the librarian's job to tell them that. The rise of using market-driven decision making to replace educated long-range planning is one of the sorry legacies of the '90s. If libraries have issues because governments and voters do not continue to fund them, that's tragic, but it doesn't mean we should start serving ice cream and partnering with the local cigarette manufacturers.

Google Answers, now that it is out of beta, serves ads along with its answers. Google has never pretended to be anything other than market-driven and when money is involved, priorities can shift. While we were not supposed to do homework for school kids, we often did market research for businesspeople. When the lowest question-asking fee of $4 seemed too high, Google lowered it to $2. For a $2 question, a researcher will get $1.50. At my librarian rates, it would be hard for me to even type an answer fast enough to make that worth my while, much less "research" it. But at this price level, people are still asking complicated questions about workman's compensation, mortgages, and biology. Most of them are not getting answered, yet answerers are reluctant to tell the askers "You need to pay more for that kind of question." Google researchers are not supposed to get into discussions about costs with question-askers.

Google Answers also suffers from the two-for-one issue in which someone asks a question and needs more than one answer—in my case, "What beverages come in 8 and 10 oz containers?" for a home soda vending machine. Once I answered this $5 question, the questioner came back with a "clarifying question" saying essentially, "Nope, that isn't it, got any more ideas?" So, the question is "answered" and yet the questioner isn't satisfied. The debate goes on anew: Pay or don't pay, rate your researcher high or low.... At a library, there is no penalty for going back for more information, especially if you bring more of your own information to the table. At some theoretical level, Google's pricing structure is supposed to encourage competition which is supposed to result in better, quicker answers. In reality, a Linus gets one, maybe two shots at an answer, no matter how much he pays, though there will be more people clamoring to answer a higher-paying question.

Recently Google Answers implemented a tipping system where Linus can give extra money directly to his researcher if he is particularly pleased with an answer. This of course also means that any $2 question could potentially become a $10 question, and so ever-hopeful researchers keep at it, even when it might not pay.

Lessons I Learned

• Your library doesn't have to be all things to all people; simply promote and provide mission-driven service.

• Fee-based services are available for those who require more than the libraries can give.

• Don't sell your skills short even if you aren't paid by the question.

Instead of Worrying, Be Even More Resourceful

Libraries and Google are both here to stay, but let's be realistic about the scale here. The U.S. has 136,000 professional librarians and about 500 mostly part-time Google researchers. The librarians are generally people with master's degrees. The researchers have been "carefully screened" and have answered some questions. Google Answers is a tiny but well-promoted niche. Instead of worrying about whether Google Answers is in some way encroaching on their turf, librarians and other information professionals would do well to use it as any other fee-based resource, since Google Answers seems to show no qualms about doing the reverse.

At some point, I decided that I'd rather be an underemployed librarian than a scrambling-for-questions-and-tips question-answerer. The money wasn't that good—almost no researcher does the job as a full-time gig—and I felt that it detracted from my credibility as an information professional rather than added to it. I've done freelance research in the past and will continue to do it in the future. So I figured that if I'm going to sit behind a rinky-dink "questions answered" sign, it might as well be my own.

Further Reading You Might Enjoy

• Google Answers researcher interviews
http://blog.outer-court.com/interviews.html

• Google Answers question-pricing tips
http://answers.google.com/answers/pricing.html

• Google Answers discusses
the future of academic librarianship
http://answers.google.com/answers/
main?cmd=threadview&id=220145

• Google Meets eBay:
What Academic Librarians Can Learn from Alternative Information Providers
http://www.dlib.org/dlib/june03/kenney/06kenney.html

• List of questions I answered for Google Answers
http://www.librarian.net/ga.html

 

Jessamyn West is now an outreach librarian for the Rutland (Vt.) Free Library. She holds a master's in librarianship from the University of Washington and has worked for Google Answers as well as at the reference desk of Shoreline Community College and Seattle Public Library, both in Washington. You can reach West via her Web site at http://www.librarian.net.
       Back to top