CNET también está disponible en español.

Ir a español

Don't show this again


Be my friend? Only on my turf

Release 1.0 editor Esther Dyson examines the promise--and the confusion--around social networks.

    "Are you a member of LinkedIn?"

    I get asked that a lot, but it's the wrong question. The questioner is assuming that LinkedIn is a social network, when in fact it's a hosting service for its users' social networks, some of which overlap. The proper question would be: "Are you a member of Juan's social network on LinkedIn?"

    Yet somehow, people are confusing the two concepts. In part, that's because the message is blurred: It's not just "use our service to communicate with your own network," but (implicitly) "join our service to make new friends and valuable contacts."

    That may be a more compelling proposition, but it's one that will be hard to deliver, especially if too many people join and the value of a link goes down. In the end, the value isn't how many people you can link to, but rather how strong those links are. (There's a difference between business services such as LinkedIn, where the focus is mostly on increasing efficiency and limiting contacts to valuable ones, and the more social and would-be portal sites, such as Friendster or Orkut, where the focus is more on increasing the number of contacts. And, of course, some people use either kind of service for the opposite purpose, which only confuses things.)

    So what should LinkedIn be offering? First of all, LinkedIn should help me manage my own social network, from myself outward. I should be able to stay within my own environment of mostly e-mail, instead of segregating my friends--the people with whom I interact the most--off into a third-party environment.

    LinkedIn should acknowledge that each person is a member of a variety of networks, of which that person is the center. Networks overlap through individuals, not as groups.

    Take this typical, true story (fake names). I recently had lunch with Bill and decided to introduce him to another friend, Will. Like many people, Will has changed jobs recently, so I got his latest e-mail address from the cc: list of a message about a board meeting we both attended recently (forget the address book!).

    Now that's a social network--a small one, but a real one. I wouldn't declare that network to the world, though it's publicly disclosed, and I haven't asked all the people on that board to be my friends. I wouldn't consider passing on third-party requests to them, but I might--as with Will--introduce them to someone else I know personally, where I see a good fit. That is, they don't meet some stranger's search criteria; rather, they both meet my criteria.

    The fact is, most of my social networks happen in the context of communications about something or other.
    The fact is, most of my social networks happen in the context of communications about something or other; they happen in my regular mail, not via some social network platform. With many of my contacts, I share several activities, seamlessly.

    The tools I want from LinkedIn within my own context are a variety of processes: for example, to select two people--perhaps one from my calendar and another from my address book--and generate a customizable e-mail inviting them to meet. Of course, that mail could be an instant message, or it might contain some tool that would allow the recipients to declare their availability over some temporary IM-like presence manager.

    I do this all the time, using a basic e-mail function: the cc: field. But the social networking tool I'm dreaming of would automatically create a record of the introduction I have made so that someday, I could search on either of those two names and a process such as "intro," and find the e-mail with which I did the deed.

    That record would belong to me, not to LinkedIn. I don't mean to pick on LinkedIn in particular; it's just one that's good enough to be worth criticizing and improving, and it's the one I use. In fact, LinkedIn is about to add some client-side functionality of the sort I describe below, but unfortunately only for Outlook users, through plug-ins.

    Using this tool, I could easily find all the introductions I had made for a particular company, the reference checks with contacts I had made on a potential new hire, and so forth. I could also use and modify such social network process templates to define my own kinds of connections (business contact, acquaintance, secret admirer).

    Hundreds of people (according to my own unscientific surveys at industry gatherings) feel uncomfortable disappointing or insulting the people who reach out to them over these networks, but now I could develop my own templates for polite rejections such as "Unfortunately, I don't have time to give this the attention it deserves." My Meta-LinkedIn service might ask me if I want to designate someone as a good contact after I have sent them five e-mails in three days, or 10 e-mails over a month, or whatever metrics I set.

    The way we network now...
    So where does that leave LinkedIn? Right now, the social network platforms are generally monolithic server-based work flow applications: They manage interactions among users, such as declarations of a link, introductions of one person to another by a third, searches leading to double-blind introductions by the system and the like.

    They also allow users to share information in a common but selectively visible space--everything from resumes and job postings to photos of themselves or their cats, their personality profiles (true, fake or "aspirational"), and links to their "friends," according to various privacy strictures.

    The ratio of transactions, or work flow, to shared content varies, as does the ambience. But the basic formats are the same.

    Right now, the social network platforms are generally monolithic server-based work flow applications.
    Business models also vary: The shared-content sites, which really do want members to live there, focus more on advertising; the transaction-oriented sites, such as LinkedIn, will most likely charge users a variety of subscription or transaction-based fees.

    But the server-hosted databases are a useful complement to the user-side tools. For example, when I ask Juan whether he wants to be in my list of searchable contacts, his reply would get registered in my own contact database; if he said yes, a notification would go to LinkedIn, which is where the heavy-duty searching and matching and the cross-network link crawling would occur.

    LinkedIn would find its proper role--connecting me and my contacts to the social networks of other people--while giving me the tools to handle my own for myself.

    For sure, there would be issues about making indirect links from one network--and from one vendor's networks--to another's. One person's strong link is another's "met once"--but that reflects real life: Many more people "know" Bill Gates than Bill Gates knows back. In the long run, the various social networks might interlink directly--but only once proper privacy assurances and other social protocol standards are in place.

    Mingling an "open" community with one in which people prefer to be private--or one with strong links with one where people are open or eager to meet almost anyone--would disappoint both sides. Perhaps people will revert to private social networks--ones they manage locally--and more public ones, a la Plaxo, where being in someone's address book is sufficient evidence of a link to follow. One is a genuine social network, while the other is what you might call "human-aided search."

    Perhaps the law of networks--the strength of a tie degrades by the square of the number of links--would become more apparent...and perhaps that would be a good thing.

    I'm not sure how good that is as a business model, but it works as a social model.