Tuesday, December 06, 2005

 

what does the future hold for communities?

Search has, thanks to Google, dominated Internet innovation for years. But, while search has made our lives easier and put hundreds of gigabytes of useful data at our finger tips, it has also served to sterilize the Internet. Google isn't a club that you learn from and contribute to, it is an encyclopedia that you use when you know what you are looking for. But what if you don't know what you are looking for? How do you sort through all the news and information that gets added to the internet every day? That is where communities come in.

The entire Internet seems to be waiting with bated breath to see what will happen with communities on the Internet. Who's going to be big? Who's going to find something that works as well as eBay or Google? Almost every major new addition to the big three in recent months (be it Google base, local search, or Yahoo 360) has had the goal of better engaging users in the content creation process.

a simple search

Perhaps to best way to underpin this change and the need for communities is to try a search on Google. If you search for eBay on google you come up with stock information, spammy ads, and a ton of eBay sites. While this all looks good on paper it is based on an algorithm (pagerank) that gets less meaningful the more targeted the search. That is to say, it forces the less current results to the top. eBay is making news, but it is also making buzz, and you just can't find buzz through Google.

This is where communities come in. Communities provide a forum for users to decide what is important on any particular topic. They are also much more organic in their approach and quicker to let trash fall by the wayside as useful information rises to the top. They are also not in business to compete with search and, as such, feel no need to duplicate the reference nature of Google or Yahoo or Wikipedia or dmoz.org.

Consider a search for eBay on digg. What do you find? Well you find that people are selling fake xBoxs on eBay, you find an eBay Developers Challenge, and you find a coupon for %10 off your next winning auction. Now, the dynamic nature of digg means you may not find any of this information tomorrow, but you'll never have to sort through stock information, or a links to eBay, to see what the Internet is abuzz about.

But what makes sites like digg possible? And, more importantly, what keeps them useful? This is where community accountability comes in.

community accountability

There are a number of different models for accountability in online communities. Most exists with one common goal: to stop lame or spam posts and to encourage good posts. I've divided the varying types of accountability into four groups, represented by the largest communities using the approach.

The craigslist model

Users can post anything with only email verification. Posts go live instantly. Spam posts are taken care of by visitors who can flag posts as mis categorized, prohibited, or spam. The model as a whole is easy to use, doesn't require a signup, and is very flexible.

The fark model

Any new link (or topic) has to be greenlighted by an administrator. Offensive comments are blocked by an administrator (or, in some cases, an algorithm). The model encourages quality (at least in the topics) but isn't very flexible. The fark model has been adapted to TotalFark, a premium "anything goes" community that is a little more adaptive.

The slashdot model

Administrators oversee the creation and posting of topics, but the comments (where the real magic of slashdot can be found) are overseen by users who are randomly (in effect) assigned administration points. This lets them classify comments to make them more visible to readers or to mod them into oblivion if they aren't useful. The model encourages intelligent and entertaining discourse and rewards the best contributors with more exposure. The model is also very adaptive and flexible but still requires a large amount of administration oversight to keep the site and topics interesting.

The digg model

Users have full control over administration and posting, but every post starts out at the bottom and must be "dug" up by "power" users who choose to monitor new posts on their favorite topic. Most digg posts fade into the background, but those that succeed quickly build steam until they reach the main page. The model is the most flexible of the bunch and does the best job turning readers into community members. The downside to the digg model is that often some of the better posts and comments are never recognized or given exposure because they fade into the noise of all the new activity.

what is next?

There are other models and variations on these models out there, but this gives a good overview of the approaches currently being tried. The question I have is what will be the final model that will work? All of the above sites are targeted pretty tightly and so aren't useful unless you are looking specifically for the type of content or community they provide. What will it take to produce an adaptive model that will be like Google for community discussion and buzz?

Of all the sites I've looked at I think digg is the closest so far. Its difficultly is that it is primarily just a link and community news aggregator and provides a poor platform for posting your own content for discussion. It does nothing to replace fark for entertainment, slashdot for quality, or craigslist for classifieds and targeting. Likewise it is easy to see how changes, such as introducing user produced polls or photoshop threads, would detract instead of add to the community. Digg does, however, appear the best and combining low administration oversight with a high quality front page that is useful to the casual user and to the pro.

I realize no one has the perfect answer, but what winning features have you found in online communities and how would you scale them to a world wide market? Please post your thoughts.


Comments:
You didn't talk about wikipedia. Overall the things I like best in communities is where it is really easy (craigslist style) to start making posts to communities but where there is also a big reward (like digg) to sign up and become a full member. Its going to be hard to scale any community though and will probably require more oversight than digg, but less than slashdot.
 
I think Slashdot demands more admin effort than Fark, then again that's just me.

Besides, you should read the comments on Slashdot. They're more coherent, insightful, and you can generally LOL more than in digg, where you'll see typicall rants from uninformed users (I can't undigg!!!11).

Fark is by far the funniest. You just can't compare the three, I have Fark, /. & digg feeds in my Google homepage, mainly because I enjoy the unique aspects of each community.

Also, Google Desktop does have some sort of way to point out "what's hot" (or something like that).
 
I think there's something to be said about hobbyist communities, such as the uber-popular swing dancing forum at Yehoodi. There's no pre or post-modding of original posts or comments, but rather... threads that are perceived as particulary interesting or funny or otherwise valuable will end up constantly at the top of each sub-forum, since each new post brings a thread to the top (standard for most forum software).

So you have the mod-up/down style of slashdot, the voting style of Digg, the multi-moderation style of CraigsList, and the simple thread-float-to-the-top style of (often topical) community boards running standard forum software.
 
You didn't mention sites like Memeorandum which leverage the existing decentralized blogsophere community that and then algorithmically figuring out what are the most important and popular topics at any given moment.

While this idea of algorithmically figuring out what is substantive is essentially the basis for link popularity algorithms like Google PageRank, Memeorandum attempts to limit spam and manipulation by starting its search from only the elite bloggers and branching outward.
 
Post a Comment



<< Home

This page is powered by Blogger. Isn't yours?