When I follow conference events such as media140 online I’m always impressed by how quickly I’m able to get a grasp of what’s happening. It’s great to see people sharing information so freely and I probably learn more about Twitter and where it’s taking us in a few hours than if I’d spent an entire month reading and researching. It reminds me of the importance of bringing people together to discuss topics and it makes me think about how Twitter is revolutionizing the way in which we use the web, more specifically it causes me to reflect on how Twitter is affecting the area that we work in, social search.

For those not fully familiar with social search it’s about harnessing user-generated information to help people find what they want, it’s not new online and is age old off-line where for most of us asking for people’s advice is second nature. Approaches vary and for simplicity we can briefly split them in to three separate areas. First there’s reordering or amending algorithmic search results with user input, example proponents of this are Sproose and more recently Zakta. Secondly, there’s the option to incorporate relevant updates, tweets or reviews into your search results, something Google and Bing announced recently. Thirdly there’s the ability to find people as opposed to pages who can help you find what you’re looking for, this is the approach that Aardvark and we at Cofacio have chosen to adopt. The three approaches are complimentary so while the first and second method help us define the ‘what’, the third approach might be seen more as identifying the ‘who’ we need to know

So what’s the big deal with social search? Well when Eric Schmidt claimed recently at the Gartner Symposium/ITxpo Orlando 2009 that how to rank real-time social content is “the great challenge of the age”, in my opinion it’s not so much a sign of where he thinks things are going in the future but more a recognition that traditional search no longer works for an increasing amount of the web’s content, namely user generated information on Twitter and social networks. Finding stuff is one thing but where one search engine really wins over another is in relevancy, so while Google or Bing can incorporate tweets, updates and customer reviews, the problem they face, and as Schmidt points out, is how can they rank them, how can they break through the noise and make them relevant? It’s this problem that practitioners of social search are trying to resolve, and I would argue that it’s an area where we can take 3 key learnings from Twitter and tweeters.

1. It doesn’t all have to be about the quality of the search result

When Stephen Fry talks about Twitter being social and the almost inherent need people have to be social and interact with others, this is a very important lesson social search needs to take on board. Up to now the debate has focused on the quality of the results and while it’s important to look at which method produces the most relevant results, we shouldn’t forget that collaborative social search offers further benefits that go beyond simply finding an answer. Asking other people or searching jointly is social and can be fun, and by engaging with other people as we look for things opportunities arise for cooperation. Searching with other people also helps you think through problems in different ways and allows for distributed cognition which improves cognitive processing and concept formulation. Following this through, social search solutions need to think about how people want to interact when searching, it’s adding another layer on top of the standard search experience.

2. It doesn’t need to be behind registration, people will happily search for things on an open platform

When we first sat down to look at how people would share searches we assumed that people wouldn’t be prepared to search for things openly on the web, in my mind Twitter has proved this theory to be incorrect. There’s stuff people would be wise not to discuss in public but many of the things that we look for on an everyday basis we will happily share, what’s more, we’ll happily give advice in public as Twitter has also demonstrated. What this means for social search is that we are no longer limited to solutions that leverage the knowledge base of your network, but instead we can now look to leverage the knowledge of anyone on the web. Users are not used to, and shouldn’t need to, log in to perform a search, that’s the result we found when we launched a prototype of Cofacio back in April 2009. In a similar way, the way people search on Twitter would suggest that we don’t need to display lengthy profiles and give over personal information in order to interact with other people on simple topics such as which is the best fish and chip shop in London or which play is worth watching in the West End.

3. Ranking people or people’s comments requires a different approach to how you rank a web page

Choosing whose advice to take or who to listen to on a certain topic is something quite personal, every person will have their own criteria and method for doing this. Ranking people on the basis of the amount of other people who follow them, which is similar to ranking pages on the amount of other pages that link to them, seems overly simplistic. On Twitter I’d argue that it’s a person’s behaviour, what they tweet about, as well as our assessment of their credibility based on follower numbers and their reputation that’s important when deciding whether we should listen to their advice on a particular given subject. Social search then needs to come up with ways in which the community can recognize individuals as experts on specific areas by using both explicit and implicit ratings. Other people’s opinions are important but how that person behaves is also key. This recognition system would then guide users with regards who they should listen to without limiting their ability to judge for themselves.

Concluding, there is one thing that was said at the 140 conference in London which is key, Twitter is about people, and with the rise of social media the same can be said of the internet, it’s going to be more and more about people. The result is that search engines developed to find us information and products are no longer suitable tools for this growing social web. You cannot scrape people’s profiles as if they were goods on a shop shelf and you cannot rank people as more or less relevant as if they were information. In light of this, if search companies want to be useful in this new era of the people’s web they need to reinvent themselves to some extent, I would suggest that they need to think more about people’s behaviour online, and for me there’s no better place to start than with Twitter.