4 Viral Social Software Tips From Enterprise 2.0
Maximize use of open social software, imitate what works for Facebook, and other expert advice from top social media maven.
When designing a social application, whether for the enterprise or public social media, you want it to be so engaging that it goes "viral," meaning that you don't have to promote it because the app's users do that for you.
At Enterprise 2.0 in Santa Clara, Jonathan LeBlanc shared a "a few basic principles you can use to socialize your applications and make them relevant to your users social and interest graphs." A developer evangelist for eBay's X.commerce business unit, who previously held a similar post at Yahoo, LeBlanc is the author of Programming Social Applications and a member of the board guiding the development of OpenSocial, which has both consumer and enterprise applications. He also participated in a Designing Social Applications panel discussion at the show.
For his talk on creating viral experiences, LeBlanc focused on the basics of what makes for an engaging social experience. It starts with the social graph, "a deeply interconnected web of people," he said. "There's a whole new realm emerging right now, building on the graph concept, which is the interest or entity graph. This is people's social connections to things--things they like, things they own, and things they are interested in. That can tell you so much more about them than their social graph ever could."
1. Map the Interest Graph
LeBlanc said one of the most effective ways of getting people to reveal their interests is by inviting them to join interest groups. Social websites have done this for years, but cutting edge sites are making it more automatic. With Google+ Circles, users categorize their contacts at the same time they form those online connections. This makes it relatively easy for users to share with subsets of their total audience of connections, although some critics say maintaining the circles quickly becomes a chore.
Facebook Smart Lists go even farther, inferring list associations from data Facebook knows about you and your contacts, for example grouping people who are local to each other, or who work at the same company.
"Facebook Smart Lists are just starting to come on the cusp of how the group model should be, because it is able to group people automatically," LeBlanc said. Anything your application can do to infer associations from user behavior, and help them form connections, will make your application more viral, LeBlanc said.
2. Model Identity
One of the basic decisions in social application design is how to identify people. "There is no be all and end all solution to identity out there, it's all about compromises," LeBlanc said.
The public Web is full of anonymous users and fake identities. Most common on social networks are identities that are presumably true (look legitimate) but have not been verified. When websites allow social login using OpenID, they save the user some of the hassle of registering on multiple websites but run the risk of pulling in fake identity data, LeBlanc said. What he called true identity is rarer, although he put in a brief plug for PayPal Access, a social login confirmed through commerce data like verified credit cards and bank accounts.
Some other lightweight identity mechanisms can also be useful. For example, the WebFinger project is reinventing the old Unix protocol that allowed you to look up basic contact information on someone based on their email address. In the Web incarnation, this can also include links to social profiles, providing a uniform way for people to identify themselves on the Web.
