Individual Actions, Community Benefits

    The network effect is the backbone of many websites that provide individual services in a socially networked setting. Recall Librarything, the site which allows you to make a library-quality catalog of your own books. The more books you add to your personal catalog, the better a personal storage device it is. That's the individual action--the cataloging of personal books. The community benefit is an outcome of the networking of all of the individual libraries together. Librarything links you to potential books of interest by networking your books with those owned by other users of the site. Each book is effectively a node, and Librarything's system tracks connections among books in thousands of personal libraries. If you and I share several books in common, we are likely to be more interested in the other books in each other's personal libraries. Our personal networks of books overlap, or parts of our personal networks do. I may be sympatico with you when it comes to poetry but very different when it comes to science fiction. The unique value of Librarything isn't the cataloging function; it's the connection of lots of users with lots of books that generate valuable recommendations. The more people use Librarything to catalog their books, the more books and libraries are in the system, and the more likely any single user is to get quality recommendations.

    The same is true of Amazon, which provides recommendations for purchases based on what "people like you" bought, and Netflix, which offers movie ratings both in the aggregate across all Netflix users and weighted for "users like you." Wikipedia increases in knowledge value the more people contribute to it. Social networks like YouTube, Flickr, and Facebook, are as valuable as their user base is diverse, populous, and active. And perhaps most powerfully (and invisibly), the network effect guides the way that search engines like Google work. Google serves you responses to your queries based on what users across the web have linked to and found valuable. Every computer that is hooked up to the Web gives Google information every time a search is completed. Your experience of the Web is thus shaped on the most basic level by the concept of software that gets better the more people use it.

    This concept is fundamental to the definition of Web 2.0.  In 2006, Tim O'Reilly, who is credited with coining and defining Web 2.0, gave the commencement address at the UC Berkeley School of Information and said:  "A true Web 2.0 application is one that gets better the more people use it. Google gets smarter every time someone makes a link on the web. Google gets smarter every time someone makes a search. It gets smarter every time someone clicks on an ad. And it immediately acts on that information to improve the experience for everyone else. It's for this reason that I argue that the real heart of Web 2.0 is harnessing collective intelligence." (http://www.slideshare.net/GeorgeAppiah/tim-oreillys-commencement-speech-at-uc-berkeley-sims)

    "Harnessing collective intelligence" is another way to think of the networked effects of individuals. It's a deceptive phrase because we often think of collective intelligence rising from a group of people who are known to each other. But O'Reilly and other technologists use the phrase to describe a kind of mass collaboration in which the participants are not necessarily aware of or working in concert with each other. The system--in the case of the example above, Google--performs the harnessing or collaborative function, operating on the individual actions of many people to improve the overall value of the network. These kinds of systems are often called "platforms" because they are sets of rules and environments that support users sharing their own content (as if from a stage) rather than content distributors in their own right. When I use the word platform, I use it in this sense as a system that is designed to support users creating, sharing, and responding to each other's content in a networked way. This separates content producers or distributors (like movie studios) from content platforms (like YouTube and other video-sharing sites).

    Of course, not every contribution improves the value of a platform. A person who chooses to spray a community blog with hateful comments is adding to the content on the site but is likely not to be seen as a positive contributor. Systems like YouTube maintain internal algorithms that are highly likely to serve you popular videos but not necessarily the ones that will be most valuable to you. How do you build a system--physical or virtual--that harnesses collective intelligence in accordance with institutional values? To do this successfully, you need an architecture of participation that focuses the network effect of many users in a direction you deem positive.

    This may sound complicated. But remember the simple example of the Facing Mars exhibition introduced in Chapter 2. Visitors must walk through a set of two turnstiles to enter and exit the exhibit. Above the turnstiles is the question, "Would you go to Mars?" and the turnstiles are labeled "Yes" and "No." In the context of personalization, this experience primes you emotionally for the exhibit and contextualizes it relative to your own identity. But there's one more element to the turnstiles. Above each one is an LED display that shows the aggregate number of visitors who have selected each option to date. This is a very simple approach to harnessing collective intelligence. The intelligence in this case is the cumulative response of thousands of visitors to a hypothetical invitation to Mars. What is valuable about this intelligence? When the exhibition was open at the Ontario Science Centre, about 2/3 of visitors answered at the entrance that they would go to Mars. By the exit, the numbers had flipped, and only 1/3 were still confident about their desire to visit the red planet.

    The collective intelligence thus told visitors something very simple: lots of people think they want to go to Mars, but when they find out what's really involved, they change their minds. This insight is interesting and potentially surprising. Because it is based on visitor data, that statement could not be as convincingly offered in label text as it is via the LED displays (even if it is an underlying message of the whole exhibit).

    Providing the LED displays also does a host of things beyond offering a few simple data points with an interpolated narrative lesson. Each visitor can see the number tick up when she walks through the turnstile, which makes her feel that she has made an impression on the exhibit.  While it's debateable whether the exhibit has gotten "better" because she has entered, it certainly has been altered in an additive way. For visitors whose minds are changed by the exhibition, the LED displays offer confirmation of a shared social shift. For visitors who do not experience a change of heart from the beginning to the end of the exhibit, the LED displays provide a piece of information that may cause them to reconsider their own experience and reflect on what makes them unique from the majority who changed their mind. The LED displays create a social context for what is already a compelling personal experience by networking the individual selections of each visitor.

    The Facing Mars exhibit provides a limited but powerful form of networked intelligence. The platform highlights some underlying ideas: visitors’ voices matter, that exhibits can transform your thinking, and that visitors experience exhibits in a social context. Each of these statements is an institutional value channeled through a participatory platform.

    What stage of participation does the Facing Mars experience exhibit? The Facing Mars turnstiles are an elegant presentation of a familiar social device: the polling system. There are many interactive kiosks at many institutions that ask you a question, invite you to respond, and then show you your response relative to others who have engaged with the kiosk. Polling and voting systems are embodiments of stage three engagement, in which individual users' actions are networked and presented to each other in aggregate.

    Most user-generated content experiences in museums are also on stage three. Visitors can produce content (write their own labels, produce stop-motion videos, etc.) and other visitors can view them. While the visitor-created products in these cases are more personalized than answers to a poll, the social aspect of the experience is still limited. I can't respond specifically to the person who wrote the provocative message on the talkback board--I can only respond to the masses. My contribution can be accessed by others, but they can't talk directly to me. In most cases, visitor contributions are unsigned. I see the aggregate response to a poll. I read an anonymous comment on the wall.  I'm having a socially informed experience, but not an interpersonal one.

    This distinction highlights the "glass ceiling" between stage three and stage four in me-to-we design. Whether for reasons of privacy, technological complexity, or lack of understanding of its value, museums very rarely try to move to socially networked experiences in which visitors are explicitly and personally connected to each other through their actions. In many cases, this is appropriate. There is little value to seeing the names attached to each visitor who has answered yes or no to the Facing Mars question as they enter the exhibition. But it is worth investigating the relative merits of stage four experiences before settling for stage three.

Continue to the next section, or return to the outline.