Information Today
Volume 19, Issue 2 — February 2002
Table of Contents Previous Issues Subscribe Now! ITI Home
• Quint's Online •
The Return of the Reference Interview
Web search interfaces would benefit by implementing more user-oriented procedures
by Barbara Quint

You may not believe it, but many years ago I engaged in somewhat radical behavior. (Shaddup!) Well, at least it seemed pretty radical to me. This was back in the 1970s in the midst of the feminist revolution. (Does anyone out there still recall the origin—or meaning—of the acronym "MCP"?) I can remember a bumpersticker acquired at a woman-owned bookstore that read, "Women are like yeast, and they will rise." Following this strategic advice, some friends and I established a Women's Lobby at our corporation to advocate the welfare—and acquisition—of female employees.

This may not exactly rank with strapping oneself to the Washington Monument, but our campaign for advancing the hiring and promotion of women gave us a scare or two. In those early days of affirmative action plans, the addition of women to clauses protecting minorities was seen as an afterthought and even, by some, as a practical joke. (This was an era in which one of the most powerful House committee chairpersons informed his colleagues that "there is more difference between a woman and a man than between a horse chestnut and a chestnut horse." Suffice it to say his venerable seniority ended in the next election with a surprise female candidate's victory.)

At one of the earliest discussions between employees on our company's affirmative action committee and corporate executives, the issue of women didn't arise until the tail end of the meeting. When it did come up, one of the department heads popped out with the remark, "Oh, good. Women are cost-effective." He should know. He headed up the economics department. Seeing the look on the faces of the two women present, the president of the company quickly intervened with a nervous clarification, "You don't understand what he meant. He meant that women are cheaper."

Women's Lobby
Needless to say, the Women's Lobby came into being within 24 hours. The first battles we fought dealt mainly with the tendency for management (which was almost all male) to use their pre-existing conceptions of women's career goals instead of research or investigation. Turns out they had "fast tracks" and "slow tracks" in their minds,neither of which were communicated to individual employees, male or female (though it's amazing how ladylike those managers considered the more leisurely pace). Instead, managers conducted performance evaluations in silent solitude. Nothing went down on paper or, at least, the paper never appeared before the subject of the evaluation.

Groaning, weeping, and rending of garments accompanied our every success in trying to establish personnel procedures in which the personnel actually shared in the knowledge of their own career plans. But when the dust cleared, a personnel department executive admitted to me, reluctantly, that the cumulative impact of all our struggles was to establish competent and professional personnel policies and procedures companywide. Since this particular company's primary product was research and itsonly assets people, it would seem logical toprovide the same level of documented attention to the welfare and use of those human assets as one would assign to inventories of office equipment and company cars.

Lucky they had us Women's Lobbyists to help them out. Right? We gained advantage for our own members too. In the midst of the revolution, the nation was hit by a recession and the company by a major downsizing. We "troublemakers" braced ourselves for a Night of the Long Knives, but actually, when dawn broke it seemed that active participation in the lobby served as a kind of protection from layoff. In fact, among the bodies of the slain lay some of the very women who had publicly disparaged and distanced themselves from all connection with our cause. Some of us had warned them this might happen. Remember the wise words of Egyptian Nobelist Naguib Mahfouz (Echoes of an Autobiography): "When will the state of the country be sound?... When its people believe that the end result of cowardice is more disastrous than that of behaving with integrity."

Even when the threat of affirmative action consequences eased—partly due to a changing political environment, but also to the successful recruiting of women and minorities and their integration into the ebb and flow of corporate life—the company still maintained the policy and procedural changes installed in those years of struggle.

The Structure of Interaction
So how does all this relate to the information industry and its products and services? Well, it's just another example, on several levels, of the need to improve the structure of interaction. Employees and management at my old firm needed to listen to each other. They needed that process of listening and the information gleaned from it to affect decision making. And sometimes it took hard times and strong-minded opponents to teach that lesson. Today the traditional information industry certainly faces another case of hard times and strong opposition. Once again, however, the answer may lie in more listening.

For a long time I've advocated the introduction of effective feedback mechanisms into online search interfaces. Well, let's assume that all the vendors still reading this column have projects underway to add such features to their products. (Meet you in the bar when you're through work.) So let's move to the next step.

Search interfaces must inquire as to user needs and desires at the start of the search process as well as the conclusion. Equally important, they need to find out more about the searchers. They need background information on the searchers' knowledge of the subject area, their comfort with technical terminology or jargon, even their access to other sources of information. Any experienced intermediary searcher can tell you that this is the most critical phase of any search interaction, not just for the information it supplies the intermediary (or intermediating search system), but also for the clarification and education it often provides the end-user.

Expecting end-users to go to help screens or FAQs and learn special techniques or extensive details about sources is just unrealistic. I hereby challenge any vendor reading this column to go to his or her Webmaster and ask for a statistical analysis of the use made of help documentation. Little to none would be my prediction and, even when help documentation or training material is used, I'll bet it's only after the searchers have tried to find the information without paying a learning entrance fee. This eternal truth explains why we all know the meaning of the acronym "RTFM."

Chatting with users about themselves and their interests at the start of the search process can also help market a site. People like talking about themselves. They appreciate anyone—even a machine—expressinga sincere interest in their welfare. If your end-user clientele seeks an even warmer experience, vendors might add the ultimate in customer-relations management: live, online personnel. But even without them, intelligent interfacing can route users to different sections of databases. It can preset qualifiers' (e.g., date, language, etc.) matching fields within the databases.A chatty interface could also help interest users in the background of the data, making them more willing to learn details.

Straight to the Database
Some might fear that such interfaces would turn people off and discourage usage. For those in a hurry, the interface could clearly offer a quick route straight to the database, one that bypasses the interview. But I believe that some may underestimate the willingness of people to discuss their needs. Years ago, an experiment compared the success of early computer medical diagnostic software with diagnoses by physicians. The physicians won, but some interesting observations came out of the study. One of them was that the computer had more success in conducting the interview with patients. Not only did the computer takethe time to ask all the relevant questions (and a few irrelevant ones), but patients answered computers more truthfully.

Some might consider this phenomenon to be explained by the computer's willingness to give a patient all the time he or she needed to complete the interview. But I have always suspected that it might also stem from a psychological factor. Namely, that you only have "it" after a doctor tells you so. This can make one reluctant to trigger bad news with excessive information. On the other hand, computers aren't perceived as wise and authoritative, like live physicians. Heaven knows what they will diagnose if not given every specific detail!

Of course, one real problem remains. Once you've taken down all these details and processed them, you must be able to supply answers to the questions that you now assume the system understands. In fact, vendors might even find themselves having to learn to say "I don't know" or even—this is the Third Millennium still, isn't it?—how to send users to alternative sources where the answers can be found.

Why would vendors go to all this trouble? Because it's the right thing to do in an era of end-user searching in which more and more people have come to rely on the Internet and its Web as the basic source of all needed information. Because it's the safe thing to do in an era of a litigious public where standard marketing language could be construed as implied promises to deliver targeted truths. Because it's always profitable to learn about one's customers if one expects to develop products and services that do good and do well.

Up the rebels!

Barbara Quint is editor in chief of Searcher, a columnist for Information Today, and a longtime online searcher. Her e-mail address is

Table of Contents Previous Issues Subscribe Now! ITI Home
© 2002 Information Today, Inc. Home