Holy crap, guys!  Am I here? Did I make it? Between canceled flights, digging cars out of ice, and more delays, my trip to San Jose has been a twisted version of The Amazing Race.   But I’m here. And that’s a very good thing because there is some serious liveblogging to get to – 15 sessions, in fact.   Up first, search and social signals.

Danny Sullivan is up on stage to moderate and kick off this whole SMX West 2011 experience. He makes a few really bad jokes and then assures us they’re only going to get worse from here.  Sweet. And with that, it’s time to start the show. Speaking we have Mike Cassidy and Paul Yiu to talk about how Google and Bing are approaching social search.

Danny calls social the Marcia Brady of search. The search engines are making more and more usage of social signals in order to give us a better search experience.

Paul Yiu from Bing is up first to talk about it. He promises he’s gonna crack better jokes than Danny but then goes back to using Danny’s Brady Bunch banter. So…apparently he lied. ;) Before social search, relevance was based upon pages linking to one another. But now we have a lot of people connected to each other thanks to these awesome social networks like Twitter and Facebook.  He talks a bit about how there used to be a startup that allowed people to surf at the same time so that you could look at the same pages your friends were looking at (doesn’t YouTube do that?) It was a natural marriage of things. What’s awesome about today is that you don’t have to come to search and make friends. He calls that unnatural. [I call that a search conference.] Now you can bring your friends with you when you search.  The premise here is that if you know people who are paying attention to certain things online, wouldn’t it be cool if that could blend into your search decisions?

At Bing, you can bring your friends with you.  If you go to Facebook, lots of cool things happen there.  He asks how many people monitor Facebook 24 hours are a day. No one raises their hand, thankfully.  But what happens if you miss something? If you marry that experience with search context, your friends can help you when they didn’t even mean to. What that looks like is Bing Social.  When you see an Italian restaurant in your search results, you can see how many of your friends ‘liked’ it on Facebook. He calls it similar to Back to the Future because your friends old behavior can help you in the future.  I like that analogy. I find it comforting. Thanks, friends!

He was in Hawaii recently and was looking for a place to eat. He says he’s ALWAYS looking for a place to eat. Dude, aren’t we all. ANYWAY – while he was searching, he found that many of his friends had ‘liked’ a certain place.  Now, instead of trusting solely how pages are linked to each other, you can friend-source your search relevance.  It’s all about shared attention.

They’ve also allowed social publishers to have their social presence and Web presence in one.  When you do a search for Netflix, you get their listing for their Web site, but underneath that there are tweets from the official Netflix Twitter account.   While it’s neat, I also wonder how they’re verifying they’re pulling in the official Twitter account and not one that’s been brandjacked.  How do we know people are who they say they are? I’m sure Bing is already on that.

So how does all this stuff happen behind the scenes? It looks something like this:

People Share and Like Content – the content goes to Facebook and Twitter – Public information comes to Bing, in real time — content process and indexing — ranking and annotation — user experience.

It’s more than just being followed by people. They look at User Attributes like friends, followers, follows and User Actions like retweets, replies, others that share the same link, and what you like. All of this together makes the secret sauce.  He says that “like farms”, similar to “link farms”, do not help.

Likes are still important. Plus, they are correlated with quality.  The more you try to tie a tweet to an existing piece of content online, the more quality it has. If the Charlie Sheen tweet that says “winning” doesn’t link to anything, it’s hard to get that tweet to rank really high.  Funny, that doesn’t seem to have been a problem for drug addict, women-abuser Charlie. What? He brought it up.

He shows the difference between how a naturally connected community looks compared to a spam marketing campaign.  He says that spam marketing campaign looks more like high school because there are big, blotchy groups, but none of them link to each other. They’re too busy  hanging out in cliques and smoking in the bathroom.

To rise above the noise, some thing to consider as search engines become more social.

  • Make it easy to Like and Share content. Include links in Tweets and Updates.
  • Trust-worthy people sharing your links or tweets, avoid spammy clumps.
  • # of people RTing or Liking what you said/shared in the last minute, hour, day week.
  • Be prepared to turn on a dime, and for the flash mob.

Next up is Mike Cassidy from Google.

Mike says it’s pretty fun for him to be up here. Years ago he was the CEO and CoFounder of Direct Hit and it was at a Danny-conference that he kicked off a deal with Microsoft that put their search results on MSN.  Direct Hit was a search engine based on click popularity.  He sold that company to Ask.com and started a new company, called XFire. It was an online gaming site that let you find your friends online and play games with them.   It was bought by MTV.  Then he started Friends Tip – recommendations based on what your friends suggest to you, which was acquired by Google. So, really, he’s come full circle.

Relevance is no longer based solely on the content of the page. It’s based on the searcher and the person that created/shared that content. Relevance is based upon recommendations. He believes that recommendations from friends are the most powerful recommendation in the world.  It’s a new way of organizing the world’s information.

Search Search in 2009:  You went to the bottom of the search results page and you’d see some social search results.

How does it work? All public Web information from the people you care about. It’s content being highlighted from two separate people – your Google chat buddies and people you’ve told Google you’re connected to through the links you shared in your Google Profile.

How do the social signals work? Social Search, Author Quality, Popularity

What’s new:

  1. Blending social results throughout the page – they’re no longer found at the bottom of the page or only when you filter. Now you can find them anywhere.
  2. Shared links – in the past Google focused only on content created by your friends, now they also include content shared by your friends.
  3. More control: connect sites privately.  If you don’t want your Twitter account to appear publicly, but still want to use it to get personalized content recommendations- you can do that now.
  4. Even easier to connect accounts.

There’s more to come!

They’re making improvements all the time. They view it as a constant project to improve the relevance of results. Social is just one of the many signals they use. They do more than 6,000 experiments a year on new things to try and 500 last year made it into the new algorithm.

Some highlights from Q&A :

Would gaining a lot of followers over a short period of time hurt your quality score?

Paul: We look at how many followers you have, what your network looks like, etc. Charlie Sheen’s networking is very one-directional, he’s not following a lot of people. The fact that you gain a lot of followers is okay because they look at user and content authority. If the tweet is one word, it doesn’t get a lot of love from a ranking perspective. Danny asks about “winning”. Hee.

Mike: They have all the same engineers who work on Web spam, working on social spam.

Is there a way to link a company’s Twitter account to get it showing in Bing?

Paul: There’s an algorithm right now that picks it up.  They focused on navigational queries to make the circle a bit clearer. It typically shows up when people are looking for you. [Would also be handy to tell Bing WHICH Twitter account you want showing up – company account vs employee account vs satire account. Also worth noting: Bing just takes your last tweet so may want to be careful about the final tweet you leave up for the day, as mentioned by Danny.]

Facebook released a comments plugin – do you consider those social signals? Do you like them?

Mike: As long as its publically crawlable. They want to use any signal they can

Paul: If it’s on the page, it’s interesting as a ranking signal.

Google can set you up with social results because you set up your profile? No profile no social results?

Mike: We do have social results, even if you don’t have a profile. If you have a public name and that matches a name on a third-party site like Twitter, then they can try and match those names and look for overlapping friends to see if there’s enough correlation.

Paul: Bing does most of its social stuff via Facebook and whoever you say you are on FB.

So, I could pay a high authority person to retweet and like everything I do and effectively game the system at this point.

Paul: Kim Kardashian gets paid $10k a tweet.  They also look at topical authority.  She may have millions of followers but she may not have topical authority on something like…science. [everyone laughs] So her authority wouldn’t come into play here.

 


About the Author

Lisa Barone

Lisa Barone co-founded Outspoken Media in 2009 and served as Chief Branding Officer until April 2012.


3 thoughts on “Social Signals & Search


Leave a Reply

Your email address will not be published.

*


*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Comments links could be nofollow free.