Designing Talkback Platforms for Different Dialogic Goals

"Where were you last night?"

If someone asked you that question, how would you answer? Answers will differ depending on who's asking, but they are also influenced by the designed environment in which questions are asked. People answer questions differently in harshly lit interrogation rooms than they do in welcoming therapists' offices or in the privacy of their own computer terminals. We have different conversations on the phone than we do in person or in internet chat rooms. The outcome of our conversations is dependent on the diversity of designed environments in which they occur.

If you want to design opportunities for visitors or users to respond to questions or engage in conversation, you need to think not only about what you want to ask visitors but how you will design conditions that are conducive to the types of answers that interest you. I'm not talking about guiding content; I'm talking about guiding form. If someone asks you a question on Twitter, you can only respond with 140 characters. We don't have the same limitations when designing talkback stations and other physical platforms for conversation, but that doesn't mean we shouldn't intentionally design the conversational tools offered. Many institutions do this unintentionally--by providing post-its or comment books, pens or crayons. Each design choice impacts the amount of thought and efforts visitors will put into their responses and the extent to which they will stay on-topic or proactively build on other visitors' arguments.

Here are a few design rules I use to think about what kinds of designed dialogue environments are right for different experience goals. I encourage you to share your own rules and thoughts on this in the comments.

If your goal is to encourage visitors to perceive themselves as partners in the content co-creation experience, make room for their thoughts sooner rather than later.

You don't need an entire gallery to frame a social question, but you do need to think about how the question or questions will be designed into the experience for maximum impact. The most common placement for questions is at the end of content labels and the end of exhibitions, but this location is by no means the most effective. Positioning questions at the end of labels accentuates the perception that they are rhetorical, or worse, afterthoughts. Similarly, making the only space for dialogue at the end of an exhibition ignores the thoughts that visitors brought with them into the experience or had along the way. If you are hoping for visitors to discuss their responses to questions with each other, or to share their answers with the institution, you can't end with the question; you need to provide several opportunities for questions and responses.

If your goal is to encourage visitors to share complex, personal responses to questions, consider offering private booths and progressive questions for visitor responses.

This technique was used in the Slavery in New York exhibition at the New-York Historical Society and continues in the popular StoryCorps project. When you want visitors to spend a long time reflecting and sharing their thoughts, you need to design spaces for response that are comfortable and minimize distractions. In the case of Slavery in New York, the end of the exhibition featured a story-capture station at which visitors could record video responses to a series of four questions about their reactions to the exhibition. The story capture experience averaged ten minutes, with visitors being given four minutes to respond to each personal, relatively imprecise question about how the exhibition affected the them. Richard Rabinowitz, curator of exhibition, noted that the progressive nature of the questions yielded increasingly complex responses, and that "it was typically in response to the third or fourth question that visitors, now warmed up, typically began relating the exhibition to their previous knowledge and experience." A lone "What do you think?" question station is not necessarily enough to elicit the rich personal reactions visitors might have to exhibitions. Rabinowitz commented that "as a 40-year veteran of history museum interpretation, I can say that I never learned so much from and about visitors." It was the lengthy progressive response process that turned what is often a series of brief and banal comments into a rich archive of visitor experience.

If you feel that your audience needs monitoring or social support, position the talkback stations in open settings.

This is the opposite situation of the previous design goal, one typical in science and children's museums. Placing feedback stations in the open lowers the probability of socially inappropriate behavior, and it also allows parents and teachers to help struggling visitors answer the questions at hand. There was a wonderful example at the Ontario Science Center in their Hot Zone area, which features several voting and commenting kiosks popular with teens. There was one kiosk in particular that was drawing several inappropriate comments, until it was moved from a corner into an open space close to the entrance to the women's bathroom. In its new location, under the watchful eyes of moms and other visitors, the inappropriate behavior diminished.

If your goal is to motivate dialogue between visitors and objects, questions and answer stations should be as proximate to the objects of interest as possible.

Visitors can speak more comfortably and richly about objects that they are looking at than objects they saw 30 minutes earlier in the exhibition. In many cases, visitors encounter talkback opportunities so infrequently throughout a visit that they seize on those opportunities to share many off-topic thoughts about their overall experience. This can frustrate museum staff, who wonder why the visitors are straying so far from the question posed. The more frequent explicit talkback opportunities are, and the more tightly and consistently connected to specific exhibits, the more visitors will focus on the experience at hand.

If you want to invite a wide range of visitors to respond to questions, it is best to design them into a context where visitor responses are of comparable aesthetics to the "official" museum content in the exhibition.

If a label is printed beautifully on plexiglass and visitors are expected to write responses in crayon on post-its, visitors may feel that their contributions are not valued or respected, and may respond accordingly. One of the things that makes the visitor stories contributed in the Denver Art Museum's Side Trip exhibition so compelling and on-topic is a design approach that elevates visitors' responses to comparable footing with the predesigned content. The vast majority of the signage in Side Trip was handwritten in pen on ripped cardboard, which meant that visitors' contributions (pen on paper) looked consistent in the context of the exhibition. The image at the top of this post is from one of their simple visitor feedback interactives which was built into a familiar, casual rolodex. By simplifying and personalizing the design technique used for the institutional voice, visitors felt like they were part of a natural conversation with the institution.

If you want visitors to answer questions collaboratively, whether in real-time or in a distributed manner, make sure your question and answer structure clearly supports visitors building on each other's ideas.

Unfortunately, most talk back walls don't support the grouping of visitor contributions or attempt to encourage conversational threads to develop. The Signtific game does this virtually by encouraging players to respond to each other by "following up" on other players' entries. But you could easily imagine doing something similar in physical space, either by using different color paper or pens for different types of questions and responses, or by explicitly encouraging visitors to comment on each other's responses or group their thoughts with like-minded (or opposing) visitor contributions.

If you want visitors to consume and enter in dialogue each other's responses, make sure that the visitors' answers are displayed in locations that makes them most useful to others.

A colleague recently called me to discuss an idea for a low-tech recommendation engine in which visitors could mark places on paper museum maps that might be of interest to other visitors like them. We talked about the fact that while visitors were most likely to be able to generate their maps of recommended spots as they walked through the institution, the completed maps would be of most value to subsequent visitors on their way into the museum rather than the way out. In this case, we talked about placing a large physical map of recommendations in the lobby rather than at the "end" of the experience, where visitor feedback often lives. This may sound obvious, but I think we often think about the creators and consumers of visitors' content as being the same people, whereas they are often visitors at different stages of their experiences with different needs.

What design techniques do you use to create successful visitor dialogue experiences? What have you seen work well, and what have you seen fail?