Metcalfe's law is really about networks, but you can kind of extend it to the "network effect".
The network effect means that an active site with a lot of users will attract more users.
This is one of the reasons I think initiatives like OCLC WorldCat reviews will have a hard time gaining traction. Hasn't Amazon won the mindshare for book reviews?
It's a challenging issue.
For example, I see all these new Web 2.0 BMO (Buy-Me-Out) sites pop up.
"we do social networking bookmarks, with visual tagging!"
Um, yeah, but delicious has pretty much captured that space, so sad, too late is your fate.
There are only so many genres of applications, and once one choice dominates within a genre, it is very hard for any (even much better) choice to gain traction. Microsoft Word / PowerPoint / Excel etc. anyone?
If everyone is putting photos on Flickr, then am I going to put my photos on your new PhotoWizzBang site?
There is always a tension going on there between One Central Site and specialized sites.
For example, USENET was wonderful in that all the discussions were in one place. Got a question? Go to USENET, that's where everyone is.
Now discussions have exploded across the net. Got a digital photography question? Well, there's still rec.photo.digital hmm, but maybe DPreview Forums is better? Now imagine that you try to start your own digital photo discussion site. Your site: two forums, four postings. DPreview: dozens of forums, hundreds of thousands of postings. Where do you think people are going to go, even if your forum offers lots of cool features?
So there is a big tension around creating sticky portal sites, and using offsite resources, and how you build a community when you arrive late to the game, when other communities have already formed.
You're a research library?
How long does it take scientists to bookmark ScienceDirect rather than navigating through your site every time? How are you ever going to get them to stick to your library portal? Do you even want them to?
But on the other hand, if your site is just a list of pointers that say e.g. "use connotea for bookmarking", not only are your scientists gaining the benefit of bookmarking in a collaborative space with other scientists... but they're also taking a risk, as is your organization. Connotea blows up? Good bye local collective organization wisdom. (I'm just picking it as a random example. ANY site that is not controlled within my internal infrastructure I consider to be at risk of disappearing at any time.)
So maybe there is a compromise?
Store or at least copy the data locally, but use APIs so that your users are transparently integrated with the big networked communities.
For example, no one is going to use book tagging if your catalogue has like three tagged books.
But use an interface to Amazon to suck up book tags from there, plus build a local tagging interface that captures tags that your users add: none of your data lost if Amazon turns off tagging, while in the meantime you get thousands more tags than you could just from local users.
Stephen Abram's article The Shop Window: Compelling and Dynamic Library Portals got me thinking about this. Because I think a lot of people think "if only our library could be Amazon.com" -- the problem with this is, there already IS an Amazon.com so why should I come to your library for some lite version of the same functionality, with thousands? millions? fewer people contributing to the community.
I agree mainly with the idea that you have to create value around intensely LOCAL, community-specific concerns. That means your audience is either the local community (public library) or your researchers (research library) or your university or.... (Personally I'd be happy if libraries started making web sites for ordinary users instead of librarians, as a start.)
Darcy Quesnel inspired some of this thinking.
Recent Comments