This week we’ve got a fantastic interview on the future of social search with Bill Slawski of SEO by the Sea, where he runs an incredible blog analyzing Google’s latest patents.
With social’s growing influence on search, I figured it was about time we got a clear (as clear as it can be at this point) view of what might be ahead for the world of socialized search. So, without further blabbering from me, here’s the interview with Bill on the future of social search.
And, honestly, I was going to add in some of my own ideas, but Bill covered the topic so thoroughly that it seemed silly to build on it at this early point in the game. Onto the interview!
1. What are the strongest signals that the engines use to calculate author/social authority? Do you see any increasing or decreasing in influence in the near future? Any new ones? Why?
One signal that seems to be very important from Google’s perspective is authenticity, or the ability to know the identify of someone participating in a social network, and have it tied to some kind of digital ID. Google Profiles, connected to Google Plus accounts, and linking to other social networks like Twitter and Flickr provide Google with the ability to associate social media signals with individuals and with other websites. Since Google has access to information such as where someone might have logged into a Google Account from, when they logged in, what they posted on Google Plus, and whom they shared it with, Google can use that information to learn how active someone might be on the site, what they contribute in the form of new posts, how meaningful their interactions might be with others, as well as what kinds of topics they focus upon.
It’s quite possible that Google is generating a user rank based upon those types of contributions and interactions, as described in the Google patent application, Ranking User Generated Web Content. From the abstract to the patent filing, we’re told that it covers:
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for analyzing quality of user-generated content involve identifying interactions between users through an electronic network and assigning a weighting factor to each interaction representing a quality of the interaction. A user credential score is generated for each user based on the weighting factors for each interaction.
The Google whitepaper, Confucius and Its Intelligent Disciples: Integrating Social with Search (pdf) describes how a user ranking like that might be used with Google’s Q&A services (code named “Confucius”) to integrate social with search results, so that searchers can see whether social network connections have answered questions relevant to the queries they are performing.
There are some signs that Google might be working on adding in other signals from social networks outside of Google, such as Twitter, even though Google doesn’t have access to nearly as much collateral information about posts made on Twitter, such as the IP addresses that someone might have tweeted from. The Google patent application Identification of Message Recipients describes how Google might be using the “@” sign in tweets to identify conversations between individuals on twitter, for example. Google may also attempt to associate the web site listed in someone’s twitter account with their Twitter account as well.
When Google acquired Katango in November, they also acquired Katango’s pending patent application, Automated Agent for Social Media Systems, which in part may collect a fair amount of information about the activities of a person, their contacts, and interactions between them in multiple social networks:
A method to automatically process social media data includes capturing captured data, describing actions and/or context relating a user across multiple social media systems. The captured data is stored within a database. One or more interfaces are provided in order to provide access to the stored captured data. A rules database is configured to store multiple social media rules (e.g., behaviors) that may be associated with a user. A behavior engine is configured to perform autonomous activities, on behalf of a user with respect to multiple social media platforms, based on the social medial rules and/or the captured data.
So it’s possible that Google may use a number of approaches to generate a user rank for individuals across a number of social networks.
2. Authorship Profiles – The New Link Graph? How do you see the authority tied to author profiles evolving in the next few years? Will the body of social proof eventually rival the link graph for authority? Why?
Google’s Agent Rank patent was granted a couple of years ago, and describes how Google might use digital IDs to indicate who the author of something is, whether article or blog post or page, or even comments and forum posts. This system might even enable someone to use meta data to note where authorized syndicated copies of content they’ve created might be stored.
Google has introduced authorship markup which you can use to connect content that you publish as the primary author with your Google Account. They’ve also come up with publisher’s markup which publishers can use to identify content published on their sites, and connect it to them with a Google Account as well.
Google has also introduced syndication markup that news agencies can use in Google News to indicate which article is an original, and where it has been syndicated to. It’s possible that we might see that syndication markup move to Author and publishers on the Web as well, so that content shared on a syndication site might be identified in relation to the orignal publication of that content.
When you create a blog at Google’s Blogger these days, you can use your Google Account profile as your profile page, and you can also leave comments on Blogger blogs using your Google Account to sign in. It’s possible that other sites, including blogs and forums, might allow you to use your Google Account to sign in as well.
As more and more content on the Web becomes associated with digital signatures like unique user IDs from Google Accounts, it’s possible that we will have a user graph to rival the Web’s link graph. It’s also possible that this identity graph might integrate itself into the link graph.
Two continuation patent filings were published by Google in September and November of 2001, which look like integrate some aspects of Google’s social network into how they work.
The September, 2011 version of Agent Rank describes within its claims how the reputation score of an author whose digital signature appears with his or her content on a page can influence the quality score of that page.
The November, 2011, version of Agent Rank describes how endorsements from “trusted agents” on content found on a page might influence the ranking of the author of that content (as opposed to directly influencing a score of the document). The amount that reputation score might be boosted will depend upon the reputation score for a person leaving an endorsement.
3. What are the top three new developments in social’s effect on search that you can see in the next year? Which major social signals currently used by the engines are most likely to grow in influence? Why?
One of the differences between social rankings and link based rankings is that queries that are recency-based, such as natural disasters or newsworthy events are often best answered by looking at social signals rather than link-based signals. When people tweet earthquake, and point to a very new article about the disaster, those social signals are revealing a page that hasn’t had time to collect links otherwise.
Some queries are often a matter of opinion or tacit knowledge, where a webpage isn’t necessarily going to have the best answer but someone you know might. For example, someone might ask what the best winery in NOVA might be to get a great meal, and also avoid rush hour traffic traveling to from Arlington. Google Plus doesn’t have an explicit Q&A section yet, but it wouldn’t be a surprise if it added one. With Google deprecating Aardvark, we might see something very much like it return in Google plus, and integrate its way into Google search results.
I mentioned Katango above, and one of the things they developed was an iPhone application that helped you group your contacts in Facebook. I wouldn’t be surprised if a similar feature was added to Google Plus, and evolved into a “suggested contact” feature as well, based upon common interests and interactions. If effective, it could increase the number of people who you would see in Google’s new “Search Plus Your World” results.
4. What are the most interesting patents in the last couple years dealing with social search? Doesn’t necessarily have to be realistic or even in the remote future, just the most intriguing. And, you guessed it, why?
There are so many social networking patents coming out that it’s hard to keep up with all of them. I liked the following one because it’s something that I’d love to see, and think is something we are building towards, even thoughsociual networks like Twitter, Facebook, and MySpace seem to be upset these days that Google appears to be prioritizing Google Plus:
Service for Aggregating Event Information – Let’s say that you’re interested in attending a World Cup match, or just learning everything that you can about it. You search for “world cup” and you see tweets, Facebook status updates, Google Plus posts, blog posts, news articles, videos, maps and more about the event.
5. What’s your stance/feeling on the Bing+Twitter+Facebook/Google+GooglePlus+EveryoneElse debate? Will Bing use their access to the larger social networks to gain the edge on the growing social industry or will Google finally win over Twitter/Facebook with pressure/market share? Why? (Seeing a trend here?)
It makes sense for Google to rely upon Google Plus data at this point, since they have much more access to information associated with users of the network. It should be a lot easier for Google to identify sock puppets, fake profiles, and people attempting to manipulate any social signals coming out of Google Plus than it is for Google to access similar information from any of the other social networks, and Google has real time, or near real time access to Google Plus content. If the others want Google to feature their content more prominently, they may have to find ways to share that information both quickly, and with associated data that might help make that information more trustworthy and authentic. Seems like Google has been trying to do some of that with the Google Open Social API for a few years.
6. Anything else you feel impressed to share.*
I’m extremely grateful for Bill and his willingness to impart some of his considerable knowledge on Google patents to the topic of the future of social search. What do you think? How do you think social media is going to be influencing search results in the future? How awesome do you think Bill Slawski is?