http://redherring.com//ForumPage_092303-02.aspx
Dysfunctional relationships
Social networking systems promise ease and deliver irritability.
By Jerry Michalski
From Ryze and LinkedIn to Tribe.net and Friendster, social-networking Web
sites are all the rage. I count at least 18 active social-networking
services today, plus several that have vanished or morphed (SixDegrees, for
example, one of the earliest such services, is now being relaunched as an
upscale-dating site.) They are sprouting so quickly they practically
constitute a new Internet boomlet of non-revenue-generating startups.
To these offers you might add similar relationship-building and -mining
software, such as dating services (Match.com, SocialNet); webs of trust
(Advogato); reputation systems (Slashdot Karma, eBay Ratings); nifty
experiments (Friend of a Friend project); and ways to mine all the above
through analytic social-network, or "X ray," technology (Valdis Krebs's
Inflow).
I am a huge fan of social dynamics, relationship building, and online
collaboration. So why do so many of these systems give me the willies? The
problems I have with them range from the pragmatic to the moral.
Let me start with the pragmatic.
First, the services generate more complexity, rather than reducing it.
Helping groups meet and collaborate is a great cause, but why do we have to
use so many different, incompatible services? For example, to manage one
community I host, I juggle contact names in Outlook, Trillian (an
instant-messaging client), Excel, Yahoo Groups, PayPal, Web-site development
software, and other services. Now, in addition to the social-network system
invitations, many of which require me to register with yet more services and
fill out yet more profiles, I am also getting multiple "update your contact
information" requests from people using Plaxo, AccuCard (from CardScan), and
GoodContacts. Enough! Even Yahoo doesn't seem to know that the people in my
groups are the same as those I want to invite (oops, Yahoo canceled its
invitation service!), message, or pay using PayDirect.
The second pragmatic design problem is explicitness. Making relationships
explicit, available to any virtual passerby, creates subtle complications.
Long ago, when SixDegrees was in full swing, I wrote its CEO, Andrew
Weinreich, that people like me were unlikely to enter their important
first-degree contacts, because those contacts would be exposed to
solicitations from strangers pinging them, saying, "I'm five degrees from
Andrew, so we should talk." I take care when I recommend people to one
another; SixDegrees disrupts that process and devalues it completely.
Few of the systems I am referring to in this article have addressed this
problem successfully. From the perspective of explicitness, LinkedIn may be
the best-designed of these services because it masks your contacts'
contacts. Unfortunately, LinkedIn trips over the third problem: usability. I
can't really figure out what to do with the service. Since my contacts'
contacts are hidden (a virtue), I cannot really surf the network for my
edification or voyeuristic pleasure. And the request-forwarding process that
the site does offer is relatively clumsy to use.
Design problems can be solved by listening to complaints and improving
software. Moral issues, on the other hand, are not as easy to fix.
When the information from social networks becomes explicit, either because
participants voluntarily enter their information and upload their address
books or because management runs analyses on email traffic to see who is
talking to whom, the resulting data is catnip for engineers and MBAs. They
cannot wait to analyze, reorganize, and optimize the networks they can
control, including these new automated social networks.
Unfortunately, social networks are not like clockwork mechanisms or income
statements. They are full of human beings, with relationships, expectations,
and prejudices, and therefore require a gentle hand from management. They do
not teach enough of those skills in business schools, and they teach even
fewer of them in engineering schools. All the X-raying and mapping of social
networks and org charts and contact lists creates the urge to undertake
radical surgery on these systems. That may not be at all appropriate.
Finally, some of the social-network service providers and many third parties
are already busy mining the data in networks where individuals disclose
their expertise, goals, business links, and much more. In my experience,
when the issue arises of how this data is collected and used, the key
players all too often have lousy intent. Too few of them are completely
trustworthy. Microsoft has had enormous trouble getting people to buy into
HailStorm, Passport, and Palladium.
We may not trust Microsoft with this information, but should we trust, say,
the Emode Friend Network? Emode has earned its principal revenue for years
now by getting people to fill out funny, seemingly innocent personality
quizzes with questions like "what kind of dog are you?" It then sells the
information (aggregated, of course) to marketers. Tens of millions of
visitors have taken dozens of quizzes each, a data miner's mother lode.
I am thrilled that social systems are getting considerable play these days,
but I consider the trend as much a curse as a blessing. Here are three
things that would improve the situation markedly:
a.. openness and integration among all these tools, so services interact
smoothly and triple-, quadruple-, or even quintuple-entry of data vanishes.
b.. more training on how to manage social systems appropriately, so
productive relationships can be enhanced, not disrupted; and
c.. more emphasis on rethinking and improving the basic tools we use to
express ourselves, so we cease thinking in 7-bit ASCII email, HTML, and
PowerPoint, and start communicating better and building lasting resources
together.
Now that would launch many new relationships.