Recently Robert Scoble presented a few videos on the topic of Why Mahalo, TechMeme, and Facebook are going to kick Google’s butt in four years. It was a controversial topic, and generated some signficant and emotional responses, such as the following two posts from Rand Fiskin and Danny Sullivan.
- I Used to Respect Robert Scoble’s Opinion…
- The Promise & Reality Of Mixing The Social Graph With Search Engines
Last week, Loren Baker asked me about patent applications which showed how Google and Yahoo might be using information from user annotations and social bookmarking and network sites to influence the rankings of search results. I sent back a list of ten patent applications from Google and Yahoo which Loren included in his excellent post on the topic – Social Media’s Direct Influence on Search Engine Ranking.
I did watch the Robert Scoble’s videos, but didn’t take any notes while doing so. I did see this snippet from Robert in Rand’s post after coming across a patent application from Google published last week on how user references in Orkut could be used to influence the rankings of search results for articles and web pages:
Google does not understand social behaviors… They have not demonstrated any proficiency at doing so. They even own Orkut and they are not using it in their search. Why is that? Because they can’t.
…Robert Scoble on Social Graph Based Search
The patent application was published after Robert uttered this statement, but it does clearly describe how references from members of a social network site such as Orkut could be used to influence the rankings of search results.
Methods and systems for endorsing search results
Invented by Misha Dynin
United States Patent Application 20070203887
Published August 30, 2007
Filed: June 30, 2004
Abstract
Methods and systems for improving user search experience with a search engine by providing a way for associated users to create and share personalized lists of article identifiers through endorsements of articles. Search endorsements can be used to personalize the search engine’s ranking of articles by offering a way for users to re-rank the article identifiers for themselves and for those who trust them.
The patent application describes a fairly detailed process where people could create personalized lists of articles and references on the Web and share them with their friends and acquaintances with the social network. It does cite Orkut as an example social network, but it’s possible that profiles and relationships in other social networks could also be used with the processes described.
Those referrals could then influence search results that people see in search results, and may allow searchers from those social networks to see the endorsements from other members of the social network, and rerank search results based upon those endorsements.
The search engines aren’t ignoring social networks by any means.
Slightly off topic. One idea that popped into my head a few days ago – that I’ve not heard anyone mention before – is another potential for social sites to be used in search. I’ve heard loose reference that state that social sites could be utilised, but the rarely give any specifics.
The idea is that a search engine could use a social bookmaking website, such as Digg (if they bought it) to basically be something like a giant DMOZ. You would have literally 1000s of free human editors to categorises not just home pages but articles and using tags as extra off-site metrics in rankings. Obviously the potential for abuse is here. I’m sure anyone who has used social bookmaking has seen the odd spam site. However, due to the huge Wikipediaesque thought police .. ehm .. I mean quality controllers that could be enlisted it may be more controllable than other metrics.
Yea right. They have one of the largest data sets on user behavior and they’re not paying attention? That borders on supreme ignorance. Also just because they haven’t said specifically doesn’t mean they aren’t testing some of this already in personalized search.
Contrats on the new gig by the way!
There are a number of academic papers that describe a search engine using DMOZ as a starting point for a crawl of the Web, because it has a set of categories that cover a wide range of topics, such as Topic Sensitive Page Rank. Some more recent academic papers described using Wikipedia to extract data about “named entities” like specific people and specific places. Using a site like Digg as a starting point for a crawl would be an interesting thing to try.
Agree completely on the dataset that Google has at their hands, TheMadHat. I wrote a post over at Search Engine Land that went into a lot of detail on how Google approaches User Behavior – Why Do People Google Google? Understanding User Data to Measure Searcher Intent
Comscore also had some interesting things to say about the growth of Orkut.
Google can certainly pick up on social networks and interactions. The inifluence will slowly become more apparent.
Social networking is great for businesses and individuals alike. From advertising your business to promoting your brand! If Google hasn’t yet picked up on behavioral statistics then it’s likely no one has effectively done so.
Hi Stan,
It is interesting to see all of the things that people are doing with social networking.
One of the problems that the major search engines face isn’t a lack of statistics regarding user behavior, but rather that there’s so much data that it might be hard to make sense of it all.
Its going to be very cool to see who comes up with the best algorithm to sort all the data on the internet. Google could certainly be improved upon. I would like to see more video networking also for the Iphone and such devices.
Hi Keegan,
I don’t know if there would be one “best” algorithm. The beauty of having multiple search engines all trying different things, and social networks with different approaches is that we can try more than one, and appreciate the ones that fit our tastes the best – and everyone has different tastes.
It’s funny how people speculated about this a few years back, and now google actually has made major changes to search. Robert Scoble doesn’t understand – google will do whatever it has to do to stay on top!
Hi jtpratt,
Thanks. It’s interesting to see how the search engine have been trying to incorporate information from social networking sites that become popular, such as Google’s recent inclusion of twitter postings into their search results. It’s not as easy to see how Google might be using sources such as twitter to rank web pages, though it’s possible that they might be.
Google’s Trust Rank patent described some of the ways that the search engine may try to use annotations from sources such as social networks, and annotations from applications like Google’s Sidewiki. Again, we don’t know for certain if Google is actually using the methods in that patent to rank pages in Web search, but they definitely aren’t ignoring the social web. As you note, it makes sense for Google to try to do what they can to stay on top. A large part of that appears to be creating the perception that they are innovative and timely.
I agree that it is interesting to see how search engines are changing. I think it only makes sense for them to incorporate social media into their sites. It seems to me that discussion on a topic in the social media space would be the most difficult to fake. Coming from the marketing world like I do, I have seen people do all sorts of crazy things to try to “game” the search engines. Writing umpteen million articles, posting videos across dozens of sites, etc. Social media, because of the way people are engaged in an actual conversation, whether on social networking sites like facebook and twitter, or even on forums, seems like it would be much harder to fake. Anyone else see the same thing?
Google is being attacked from all sides. I think that the changes coming are going to have serious ramifications on Google and on search as a whole. It is not just social media, it is mobile search and things like Siri, that will take the search out of google’s control. People are quickly going to need to find a way to get the placement needed through other avenues than traditional search.
Hi Cesar,
I’m not convinced that Google really needs to be too concerned that Siri will take away too much of their traffic, and they’ve been very active in attempting new approaches themselves to retain their strong lead in the search market.
As a site owner it’s always a wise thing to keep an eye out for other ways to get people to visit your pages beyond search engines and traditional search.