FLIRTing with the Crowds

According to Mashable, Wikia will be arriving shortly. I’ve said it before and I’ll say it again, Wikia will not be a real threat to Google any time soon. Here’s a critical assessment as seen through my FLIRT model:

Focus

Area

I see Wikia definitely as a complementary, not a substitute service to algorithm-based search engines. I myself use a lot of different search tools along with Google’s: wikipedia for definitions and references, del.icio.us and digg for recent hot links on a given topic, Jaiku for asking questions from my inner circle, etc. There’s definitely space for all these services as they answer a different questions - or the same questions a little differently. The big question is, does Wikia really bring something totally new to the table in addition to already existing services. Even if it does, they probably would be better off targeting smaller and more specific areas of search than taking on Google head on (although I understand the ambition after the success of wikipedia).

Scale

Wikipedia’s entries are written by a tiny minority (some 5%) of its users. In other services the number is even smaller. The way I see it, of the billions and billions of pages on the internet, only a small fraction can be reliably covered with people powered search. The web is expanding at an ever-increasing rate that only leaves automatic search technology as an alternative to really grasp the vastness of it.

Depth

There’s one major problem in the underlying logic of ceding ultimate control to the masses with open, people-powered search: Google and the likes devote massive efforts to avoid spammers and tricksters getting their links to the top of the charts, by closely guarding and continuously tweaking their complex search algorithms. With people powered search this process would be completely opened to the public and also to spammers. The effects can already be seen on people powered sites such as Digg.

Language

People don’t target their actions randomly; they congregate around social objects. While the links people find really important are good social objects that can inspire intentional action (as with Digg, del.icio.us), it is problematic to get them involved in contributing and organizing links that are slightly less relevant. If they’re not interesting, they’re not good social objects. If they’re not good social objects, it’s hard to build an active, wide-scale collaboration around them. This problem is not as elevated with algorithm-based search as it utilizes the passive contribution of the crowds: people are not (usually) intentionally contributing to Google’s search function when they e.g. post a link on their blog or elsewhere, but Google uses this contribution nonetheless.

Incentives

While Google is coming to the Wikipedia territory with Knol that emphasizes identity (fame) of the contributor and enables sharing ad revenue with the participants (money), Wikia Search is (at least to my current knowledge) anonymous and doesn’t deal in cash. How to motivate people to share their resources is going to be a major issue.

Rules

Related to the depth of control: more important than actually crafting the code of conduct will be how it is enforced. As algorithm based engines do through automatication , secrecy and continuous development, how will it be solved with people powered search? One answer will obviously be that the wisdom of the crowds will weed out bad content, but with the methods of spammers getting more sophisticated each day, will this suffice?

Tools

I haven’t had the opportunity to get to know the tools of the trade in depth but surely, in addition to more traditional web service, there needs to be very easy to install and use browser tools for storing, tagging and rating web content on the fly. However, getting masses of people to download the Grub client and running it on their computers - I’m not so sure that what works for medical and scientific (the common good) purposes could work well for business. This comes down to incentives and as I already stated I’m not convinced that they are right for going through the complex process of installing and continuously running client software. I for one won’t be lending my resources as they’re scarce already with my Windows ran Dell and I intend to keep my Mac as pristine as possible.

—

So there you go. Now don’t get me wrong, I’m definitely pro-people driven (you might have guessed that from the name of this blog, eh?), but as it stands now, there are still a lot of open questions when it comes to Wikia Search and this post exists simply to explore some of them. More on the topic here, here, and here.