The idea of Google incorporating some kind of social network into their basic search functionality is nothing new, and the fact is that search relevance in 2011 and beyond needs to be defined by peer recommendation rather than external and uncontrolled factors such as artificial link building. Google Circles, accidentally announced (and then specifically denied) this weekend at SXSW seems to be the internal name for this product.
Google already have some pretty cool social search functionality for account holders built into their latest algo, and have also been very active in embracing social media both in terms of their relationship with users, and also via Orkut and Buzz. Arguably the collaborative ability of Google Docs and the ill fated Wave experiment also hint at a highly evolved social strategy.
Back in December, TechCrunch posted an article that included a screengrab of a confidential and altered Google interface that included the ability to restrict the search to Loops:
At the time, Google +1 was the name being used to describe the product extension, with a suggestion that it would be baked into the Chrome browser and other software like the toolbar.
For me, the use of the terms loops or Google Circles suggests the ability to compartmentalise relationships and restrict results to those groups:
We are all members of different groups throughout the day, and have different interests based on them. A single set of personalised results may not necessarily be enough to deliver the most relevance to users, whereas being able to confine a set of results to a particular set of websites through Google Circles would be much more useful.
The question is how Google Circles (or whatever it is ultimately called) might work to establish these loops or circles. There are massive challenges with launching their own network – they’ve learned this lesson well through Buzz and Orkut, and this approach for me is unlikely.
A toolbar based social functionality – not dissimilar to Del.Icio.us – could be more likely. This could take a number of forms, either passive tracking based on monitoring of behaviour during a session, looking at the websites that you spend a lot of time on, and those visited by other people within your different networks to determine what content is of most interest, or an active bookmarking system that overlays content that you and people within your networks have shared. I think a passive system is most likely.
Where Can Google Circles Data Come From
An awful lot of your social connections are freely available for Google to index once you have set up your profile with them (and even before you agree to share). Twitter followers, Linkedin profiles that are linked to by company pages and include co-worker lists. Google might not have access to the full social graph data from Facebook, but they are able to crawl open profile pages and incorporate some data from that into their search results. They also crawl your GMail emails and can recognise your interests based on who and what you are emailing.
In short, Google Circles are already well defined, thanks to publicly available data, and provided that you opted in, using that data effectively would be a possibility.
Google Circles – All About the Algorithm
Google’s strength has always been about their ability to mimic human understanding of relevance, and do this through the aggregation and analysis of data algorithmically. Using a passive system to overlay social metrics onto your wider search behaviour seems to me to be a more Google-ish experience for users than a clearly defined social network that competes directly with Facebook, which with 600+ million users is perhaps too large to go head to head with.
Thanks to Google’s existing range of socially integrated products, the company can already provide social search, taking a step further and integrating more content explicitly via Google Circles seems to be entirely logical.
Of course I could be wrong. We’ll find out in May when Google (allegedly won’t) launch Google Circles for real.