21 July 2008

Wikia Search and Google must both deal with the Established Site Effect.

I think one of the main reasons people feel hesitant about embracing Wikia Search is because it turns search into a popularity contest. While people are undoubtedly more effective than any algorithm at detecting spam and irrelevant content, in aggregate they are probably not going to be as effective at discovering and organizing new content. Human-edited search results are naturally going to reflect the Web as Internet users know it already -- established sites will rank higher than the rest simply because more people know about them. What if something new and super-relevant to a particular query comes along, though? Can it displace a less relevant but highly popular site in the search results for that query? I have some serious doubts about whether it can, and it's a big problem because search is one of the primary ways Internet users discover new content. If search just reflects the "same old thing," then that's all many people will ever see. A lot of new but very good content will just languish in obscurity.

It's not like purely algorithmic search engines don't have to deal with this same issue, though. Google attempts to tackle the problem by placing a premium on fresh content (for instance, new blog posts), allowing them to show up alongside the big, established sites. Indeed, the Google algorithm is more complicated than some people give it credit for; it's not ALL about links by any means. I'm not sure if this blog has ANY inbound links to a particular individual blog post, but people still occasionally find my posts when they search with Google. Whether that is a good thing or not is another issue! Still, Google undoubtedly lets established sites have a significant edge over their competition -- as time goes on and the big sites get more and more and more links, it may well become harder and harder for sites to start from scratch, utterly linkless, and still get noticed. Honestly, I found it easier to get people to read my stuff in 1997 as a 14 year old than I do now despite the fact that the Internet user base has grown so much bigger in the past 11 years. Obviously I need to get my infectious teenage energy back if I ever hope to make it big on the Web!

It's one thing to notice that there is a problem -- it's another to actually come up with a solution to the problem. I think Google has a better handle on this issue right now than Wikia Search does, which is understandable considering that WS is the new kid on the block. Google's fresh content and relevancy boosts let even sites lacking in links be seen. If the trend of more and more content being produced continues, though, I'm not sure it will be possible for every site to have its day in the sun. It may already be a necessity for webmasters to diversify beyond search (social media anyone? Sorry, just asking!). When it comes to Wikia Search, I think webmasters themselves are going to have to stake out a claim for their sites personally. The community will ultimately decide what sites should reign supreme for particular queries, but individual webmasters are probably going to be the ones who will be the first to suggest their own sites as being relevant for long-tail keywords. The worst thing Wikia Search could do right now is discourage people from promoting their own content even though self-promotion is another threat to the search engine's usefulness. Will people who have an aversion to self-promotion be able to get their content noticed on WS if their sites aren't already really popular? I rather doubt it, unfortunately, though I suppose Wikia Search could try to give an algorithmic boost to fresh content at the risk of upsetting the community.

I wonder if Wikia Search should really be considered as a continuation of the spirit of the Open Directory which is still an excellent resource. Organization problems have limited the Open Directory's growth -- it just doesn't index enough of the Web to be listed in the same sentence as Google. However, for general queries the results on Open Directory are often on par with or superior to that provided by Google: humans really are excellent at organizing relevant information. Because Wikia Search can have as many editors as it does users, perhaps it will have the manpower to keep up with the Web's ferocious growth, but Google's ability to intelligently index both established and non-established site alike with great speed gives it a definite edge over every human-powered Web index at the moment.