±±¾©Èü³µpk10

TechCrunch blogger, Michael Arrington, called it "a complete letdown", while Chris Sherman at Search Engine Land moaned that it was "yet another crappy search service that may, potentially, if all goes well, eventually turn into something useful".

In Wales' defence, he did point out at the launch that the community-oriented search engine was "like fine wine, in that it will get better and better as time goes by and more and more people contribute". That's because the search engine relies on user input - whether that's creating mini-articles on topics users search for, or rating results to make keyword searches more relevant.

The first feature carries definite overtones of Wales' other project, Wikipedia - although Wikipedia is owned by the not-for-profit organisation Wikimedia Foundation. Wikia Search, however, is a commercial venture from Wikia Inc - although the two are similar in that people will be able to edit and comment on these mini-articles. But Wikia Search combines this with several social networking features.

Users can share photos, add friends and detail their hobbies, as well as edit their own profiles. In Facebook-style, their friends can also see what they've been up to via the activity log and post messages on the white board. So Wikia Search is something of a hybrid of a social networking site and search engine.

Like other search engines, it uses an algorithm to generate results using technologies that crawl and index the web. But as the search engine gains momentum and more users start to rate sites based on the accuracy and relevance to their search keywords, user ratings will be incorporated into the search algorithm. This means that popular sites could secure a high place in the results even if they're not particularly well-known or an obvious contender for the number-one spot.

Whither innovation

In one sense, this is no different to how the pay-per-click (PPC) model works, argues Chrysi Philalithes, vice-president global marketing and communications at Miva. "Many PPC engines use yield-based ranking, where one of the factors that determines where a PPC ad appears in search results is how many times an ad is clicked on," she says.

At the moment, user ratings are based on a five-star system, and users allocate a score based on how relevant they think the site is. But Wales says the way results appear to users and also how much user ratings influence search results are still being refined. Over time, Wikia plans to build on the community features further, by adding social bookmarking tools, so that users can share URLs easily, "and it will generate good data on which websites people find popular," adds Wales.

The success of review sites, such as TripAdvisor, not to mention feedback features on Amazon and eBay, suggests that users generally trust what other users say. They trust even more what their own circles of friends on social networks such as Facebook have to say, and this is where Wikia Search could prove to be really powerful. "If you ask consumers where they go for research before they buy a product, in their top two or three answers will be friends and family, and a search engine," says Stephanie Carr, managing director of search engine marketing agency The Search Works. Wikia Search combines both.

One of Wikia Search's distinguishing features is its open-source infrastructure. While the big three search engines - Google, Yahoo! and Microsoft - view their search algorithms as commercial secrets, the technology behind Wikia Search is completely open to software developers and can be modified or adapted by anyone with programming knowledge. "I've said before that internet search must be more open and transparent, and today marks a major milestone in our mission to make it just that," announced Wales at the launch of Wikia Search.

That said, it's this same openness and transparency that many agencies and marketers believe will make Wikia Search a target for spammers and organisations that want to manipulate ratings to boost their site's ranking in the search results.

"A lot of black hat guys would love something like Wikia Search to become a success because they'd have a field day in playing that social algorithm," claims Nilhan Jayasinghe, head of natural search at search agency Spannerworks.

But Wales insists any spammers or hackers will suffer short shrift. "It's possible people might try to manipulate results, but we'd ban them," he says. He also believes that having a body of users rating search results will soon separate trusted from mistrusted sources. It's an argument that carries some weight. Walid Al Saqqaf, founder of restaurant and bars reviews site TrustedPlaces, believes its user base forms his site's first line of defence. "If a place has numerous reviews - whether negative or positive - from a member that has not given much information or contributed elsewhere, the community will flag them up as someone trying to beat the system," he says.

Community service

But the difference between the two sites is that TrustedPlaces already has a thriving community and Wikia Search doesn't. Nor does Jayasinghe think it will easily acquire one.

"Google's success is really about scalability, which is a huge, huge task," he says. "At the moment, Wikia Search is just not useful enough for anyone to use it and provide the user input. If you started with the relevancy that Google has and then asked people to add (their ratings), then I'm naturally going to use it because it has the relevancy and gives me this extra information. But I honestly think people aren't going to put the time in (with Wikia Search)."

Wikia acknowledges that the search experience on Wikia Search is currently poor; in the 'about us' section on the site, it states: "We are aware that the quality of the search results is low." But in a comment posted in response to Arrington's post on TechCrunch, Wales justified the performance of Wikia Search by saying: "It's a project to build a search engine, not a search engine."

He also made it clear at the launch of Wikia Search that it was only an 'alpha' release, so there are bound to be some glitches. "We expect it to be a minimum of two years before we have industry-standard quality search," he adds.

Philalithes, for one, admires Wales' candour. "What I thought was impressive was the honesty of the brand - on its front page it basically states that this product is in development and will require users' help to get optimal results. I found that very refreshing, and it also made it very inclusive," she says.

But there's no denying the uphill struggle Wikia Search faces. Although Lucy Allen, managing director of Netrank, thinks Wikia Search has a future, she worries that user ratings will be limited to popular search terms. "Where Wikia Search will struggle is in the long tail and the more obscure search terms," she says.

One also has to question how long it will take Wikia Search to build up a viable community of users to start delivering useful search results. "Even though Wales himself said, 'don't expect to log on to this site on day one and get a Google-type experience', the problem is you might log on in five years time and still not get a Google-type experience," warns Paul Mead, managing director of VCCP Search.

It's not as if the other search engines have been standing still either; Google and Yahoo!, in particular, have been actively pursuing social search opportunities. Google has created iGoogle, a personalised search engine that draws on a user's previous search histories to improve the relevancy of search results; Google Co-op, a customised version of Google, lets users rate relevant sites under different topics and add specific URLs in the search results; and, in December, it announced its Knol project.

A knol is defined by Google as "a unit of knowledge" and although not public yet, knols sound rather like the mini-articles that appear on Wikia Search, insofar as they feature content written by users (the aim is to start with articles written by authors and experts). People will be able to give feedback and, in a move that could have repercussions for Wikipedia, knols will feature in Google search results, where relevant.

Personalising search

Yahoo! has also been active in the area of personalised and social search.

"We see search moving from a purely keyword-based medium to an environment that is more social and contextual," says Jeff Revoy, European vice-president of Yahoo! search and social media. "We believe social search delivers a more comprehensive and satisfying search experience by complementing algorithmic web search with these human insights." An example of this is Yahoo! Answers, a knowledge-sharing community that allows people to share knowledge, experiences and opinions on a number of topics by posting and answering questions.

There are other smaller search engines too - such as Mahalo - which rely on human beings to improve the relevancy of results. But these also have nowhere near the traction of the big three search engines. Nathan Levi, head of search at digital agency Avenue A Razorfish, sums it up: "Google has desktop search, the Google taskbar and so many free applications that users keep coming back. So the argument about whether Wikia Search will catch up is redundant to start, as Google is completely unique."

Perhaps rather ironically, Google had this response when asked to comment on this article: "Search is a highly competitive industry and we welcome competition, which stimulates innovation and provides users with more choice."

Putting issues of size aside, Ed Foster, head of search for EMEA at Universal McCann, is unsure of the money-making potential behind social search engines. "I doubt Yahoo! makes a whole lot of money out of Yahoo! Answers - certainly they make more from paid search and use that to drive volumes - it does add to the user experience," he concedes. "However, would a company want to do nothing else but an answers service? I don't know where the revenue is going to come from."

Wales provides some enlightenment when outlining his future vision for the search engine. "There is a vague plan to do advertising around search - something similar to Google paid search," he reveals. There are also plans to add search functionality for different media - such as video and blogs - although he concedes news sites will probably be the first addition, "as it is reasonably straightforward", he explains.

Whether Wikia Search will be a Google killer or a Google victim remains to be seen, but there's a general consensus in the market that the concept of social search, which Wikia Search embraces, has long-term potential. "I'm a strong believer in people and communities determining relevancy on the internet, so if they can build an exceptional community, there is every possible chance that the Wikia Search results could be more qualified than Google. If this happens, they will attract a good chunk of my paid search and SEO attention," says Andrew Hunter, head of marketing at Qype, a reviews site with more than two million unique users each month.

The rising number of natural search results that incorporate social networking sites, such as MySpace and YouTube, is already having repercussions for the search engine marketing industry. Many now refer to reputation management instead of search engine optimisation (SEO), for instance, and the challenge for advertisers is monitoring what is happening to their brand both in those areas they can control (their website, paid search keywords and so on) and those they can't (blogs, videos and social networking sites).

"The ability for users to create and edit articles on Wikia Search is obviously an issue - however, the online environment is constantly evolving and brands have to accept that internet users can and will share their views," points out Richard Clark, marketing manager for DSG International, which owns the Dixons and PC World brands.

Whether that will be on Wikia Search or another site is uncertain, but there's no doubt that the hunger for user participation online is pushing the search market into new, uncharted waters and ensuring it remains a market to watch in 2008.